Autonomous driving holds transformative potential but remains fundamentally constrained by the limited perception and isolated decision-making with standalone intelligence. While recent multi-agent approaches introduce cooperation, they often focus merely on perception-level tasks, overlooking the alignment with downstream planning and control, or fall short in leveraging the full capacity of the recent emerging end-to-end autonomous driving. In this paper, we present UniMM-V2X, a novel end-to-end multi-agent framework that enables hierarchical cooperation across perception, prediction, and planning. At the core of our framework is a multi-level fusion strategy that unifies perception and prediction cooperation, allowing agents to share queries and reason cooperatively for consistent and safe decision-making. To adapt to diverse downstream tasks and further enhance the quality of multi-level fusion, we incorporate a Mixture-of-Experts (MoE) architecture to dynamically enhance the BEV representations. We further extend MoE into the decoder to better capture diverse motion patterns. Extensive experiments on the DAIR-V2X dataset demonstrate our approach achieves state-of-the-art (SOTA) performance with a 39.7% improvement in perception accuracy, a 7.2% reduction in prediction error, and a 33.2% improvement in planning performance compared with UniV2X, showcasing the strength of our MoE-enhanced multi-level cooperative paradigm.

We present Ishtar, a power-efficient RISC-V baseband system-on-chip (SoC) tailored for multi-standard integrated sensing and communications (ISAC) in low-altitude wireless networks (LAWNs). Ishtar integrates a hierarchical scheduling scheme and a system-level power-gating architecture that dynamically controls power domains to balance performance and energy efficiency. It supports dynamic task scheduling across heterogeneous protocols using a domain-specific, graph-based representation. Implemented in 40 nm technology and running at 300 MHz, Ishtar achieves better normalized efficiency than state-of-the-art SDR SoCs, delivering real-time multi-standard sniffing under stringent power and area constraints.

Energy efficiency is crucial for Connected Autonomous Vehicles (CAVs), where real-time perception via Deep Neural Networks (DNNs) demands significant computing resources. Although techniques such as Dynamic Frequency Scaling (DFS) and block-skipping reduce energy usage, they may degrade accuracy or increase inference speed. Integrating these methods with Collaborative Perception (CP), which leverages data sharing among vehicles to improve perception performances, offers a potential trade-off between accuracy and energy usage. This paper introduces EC-PUBSE (Efficient Collaborative Perception Using Block-Skipping, DFS, and Environment awareness), an energy-efficient CP framework for CAVs. EC-PUBSE dynamically allocates computing resources using block-skipping and DFS, coupled with environment-aware response time budget to sustain accuracy and reduce energy. We formulate the problem as a mixed-integer nonlinear programming (MINLP) problem, addressing the trade-off between computation energy and accuracy, under communication and response time budget constraints. We leverage model diversity and redundancy to improve system performance compared to standalone and CP. The proposed framework adapts to traffic conditions to offer an adaptable solution for sustainable autonomous driving. Experiments show that EC-PUBSE reduces energy consumption by up to 22%; it improves energy efficiency by up to$2.7\times $and$2.3\times $vs. standalone and collaborative only, respectively, maintaining perception performance in dynamic conditions.

Federated learning (FL) is a promising paradigm for multiple devices to cooperatively train a model. When applied in wireless networks, two issues consistently affect the performance of FL, i.e., data heterogeneity of devices and limited bandwidth. Many papers have investigated device scheduling strategies considering the two issues. However, most of them recognize data heterogeneity as a property of individual devices. In this paper, we prove that the convergence speed of FL is affected by the sum of device-level and sample-level collective gradient divergence (CGD). Device-level CGD refers to the gradient divergence of the scheduled device group, instead of the sum of the individual device divergence. Sample-level CGD is statistically upper bounded by sampling variance, which is inversely proportional to the total number of samples scheduled for local update. To derive a tractable form of the device-level CGD, we further consider classification tasks and transform it into the weighted earth moving distance (WEMD) between the group distribution and the global distribution. Then we propose FedCGD algorithm to minimize the sum of sampling variance and WEMD on classification tasks by device scheduling and bandwidth allocation, within polynomial time. Simulation shows that the proposed strategy increases classification accuracy on the CIFAR-10 dataset by up to 4.2% while scheduling 41.8% fewer devices, and flexibly switches between reducing WEMD and reducing sampling variance.

In edge intelligence systems, deep neural network (DNN) partitioning and data offloading can provide real-time task inference for resource-constrained mobile devices. However, the inference time of DNNs is typically uncertain and cannot be precisely determined in advance, presenting significant challenges in ensuring timely task processing within deadlines. To address the uncertain inference time, we propose a robust optimization scheme to minimize the total energy consumption of mobile devices while meeting task probabilistic deadlines. The scheme only requires the mean and variance information of the inference time, without any prediction methods or distribution functions. The problem is formulated as a mixed-integer nonlinear programming (MINLP) that involves jointly optimizing the DNN model partitioning and the allocation of local CPU/GPU frequencies and uplink bandwidth. To tackle the problem, we first decompose the original problem into two subproblems: resource allocation and DNN model partitioning. Subsequently, the two subproblems with probability constraints are equivalently transformed into deterministic optimization problems using the chance-constrained programming (CCP) method. Finally, the convex optimization technique and the penalty convex-concave procedure (PCCP) technique are employed to obtain the optimal solution of the resource allocation subproblem and a stationary point of the DNN model partitioning subproblem, respectively. The proposed algorithm leverages real-world data from popular hardware platforms and is evaluated on widely used DNN models. Extensive simulations show that our proposed algorithm effectively addresses the inference time uncertainty with probabilistic deadline guarantees while minimizing the energy consumption of mobile devices.

Deep neural networks (DNNs) have been widely applied in diverse applications, but the problems of high latency and energy overhead are inevitable on resource-constrained devices. To address this challenge, most researchers focus on the dynamic voltage and frequency scaling (DVFS) technique to balance the latency and energy consumption by changing the computing frequency of processors. However, the adjustment of memory frequency is usually ignored and not fully utilized to achieve efficient DNN inference, which also plays a significant role in the inference time and energy consumption. In this paper, we first investigate the impact of joint memory frequency and computing frequency scaling on the inference time and energy consumption with a model-based and data-driven method. Then by combining with the fitting parameters of different DNN models, we give a preliminary analysis for the proposed model to see the effects of adjusting memory frequency and computing frequency simultaneously. Finally, simulation results in local inference and cooperative inference cases further validate the effectiveness of jointly scaling the memory frequency and computing frequency to reduce the energy consumption of devices.

The rapid development of deep neural networks (DNNs) is inherently accompanied by the problem of high computational costs. To tackle this challenge, dynamic voltage frequency scaling (DVFS) is emerging as a promising technology for balancing the latency and energy consumption of DNN inference by adjusting the computing frequency of processors. However, most existing models of DNN inference time are based on the CPU-DVFS technique, and directly applying the CPUDVFS model to DNN inference on GPUs will lead to significant errors in optimizing latency and energy consumption. In this paper, we propose a DVFS-aware latency model to precisely characterize DNN inference time on GPUs. We first formulate the DNN inference time based on extensive experiment results for different devices and analyze the impact of fitting parameters. Then by dividing DNNs into multiple blocks and obtaining the actual inference time, the proposed model is further verified. Finally, we compare our proposed model with the CPU-DVFS model in two specific cases. Evaluation results demonstrate that local inference optimization with our proposed model achieves a reduction of no less than 66% and 69% in inference time and energy consumption respectively. In addition, cooperative inference with our proposed model can improve the partition policy and reduce the energy consumption compared to the CPUDVFS model.

Future Vehicle-to-Everything (V2X) scenarios require high-speed, low-latency, and ultra-reliable communication services, particularly for applications such as autonomous driving and in-vehicle infotainment. Dense heterogeneous cellular networks, which incorporate both macro and micro base stations, can effectively address these demands. However, they introduce more frequent handovers and higher energy consumption. Proactive handover (PHO) mechanisms can significantly reduce handover delays and failure rates caused by frequent handovers, especially with the mobility prediction capability enhanced by artificial intelligence and machine learning (AI/ML) technologies. Nonetheless, the energy-efficient joint optimization of PHO and resource allocation (RA) remains underexplored. In this paper, we propose an AI/ML-based energy-efficient PHO and RA (AEPHORA) framework, which leverages AI/ML-based predictions of vehicular mobility to jointly optimize PHO and RA decisions. AEPHORA aims to minimize the time-averaged system transmit power while satisfying quality of service (QoS) constraints on communication delay and reliability. Simulation results demonstrate the effectiveness of the AEPHORA framework in balancing energy efficiency with QoS requirements in high-demand V2X environments.

With the growing integration of artificial intelligence in mobile applications, a substantial number of deep neural network (DNN) inference requests are generated daily by mobile devices. Serving these requests presents significant challenges due to limited device resources and strict latency requirements. Therefore, edge-device co-inference has emerged as an effective paradigm to address these issues. In this study, we focus on a scenario where multiple mobile devices offload inference tasks to an edge server equipped with a graphics processing unit (GPU). For finer control over offloading and scheduling, inference tasks are partitioned into smaller sub-tasks. Additionally, GPU batch processing is employed to boost throughput and improve energy efficiency. This work investigates the problem of minimizing total energy consumption while meeting hard latency constraints. We propose a low-complexity Joint DVFS, Offloading, and Batching strategy (J-DOB) to solve this problem. The effectiveness of the proposed algorithm is validated through extensive experiments across varying user numbers and deadline constraints. Results show that J-DOB can reduce energy consumption by up to 51.30% and 45.27 % under identical and different deadlines, respectively, compared to local computing.

Federated edge learning (FEEL) enables collaborative model training across distributed clients over wireless networks without exposing raw data. While most existing studies assume static datasets, in real-world scenarios, clients may continuously collect data with time-varying and non-independent and identically distributed (non-i.i.d.) characteristics. A critical challenge is how to adapt models in a timely yet efficient manner to such evolving data. In this paper, we propose FedTeddi, a temporal-drift-and-divergence-aware scheduling algorithm that facilitates fast convergence of FEEL under dynamic data evolution and communication resource limits. We first quantify the temporal dynamics and non-i.i.d. characteristics of data using temporal drift and collective divergence, respectively, and represent them as the Earth Mover's Distance (EMD) of class distributions for classification tasks. We then propose a novel optimization objective and develop a joint scheduling and bandwidth allocation algorithm, enabling the FEEL system to learn from new data quickly without forgetting previous knowledge. Experimental results show that our algorithm achieves higher test accuracy and faster convergence compared to benchmark methods, improving the rate of convergence by 58.4% on CIFAR10 and 49.2% on CIFAR-100 compared to random scheduling.

Collaborative perception (CP) is emerging as a promising solution to the inherent limitations of stand-alone intelligence. However, current wireless communication systems are unable to support feature-level and raw-level collaborative algorithms due to their enormous bandwidth demands. In this paper, we propose DiffCP, a novel CP paradigm that utilizes a diffusion model to efficiently compress the sensing information of collaborators. By incorporating both geometric and semantic conditions into the generative model, DiffCP enables feature-level collaboration with an ultra-low communication cost, advancing the practical implementation of CP systems. This paradigm can be seamlessly integrated into existing CP algorithms to enhance a wide range of downstream tasks. Through extensive experimentation, we investigate the tradeoffs between communication, computation, and performance. Numerical results demonstrate that DiffCP can significantly reduce communication costs by 14.5-fold while maintaining the same performance as the state-of-the-art algorithm.

Cooperative perception research is hindered by the limited availability of datasets that capture the complexity of real-world Vehicle-to-Everything (V2X) interactions, particularly under dynamic communication constraints. To address this gap, we introduce WHALES (Wireless enHanced Autonomous vehicles with Large number of Engaged agentS), the first large-scale V2X dataset explicitly designed to benchmark communication-aware agent scheduling and scalable cooperative perception. WHALES introduces a new benchmark that enables state-of-the-art (SOTA) research in communication-aware cooperative perception, featuring an average of 8.4 cooperative agents per scene and 2.01 million annotated 3D objects across diverse traffic scenarios. It incorporates detailed communication metadata to emulate real-world communication bottlenecks, enabling rigorous evaluation of scheduling strategies. To further advance the field, we propose the Coverage-Aware Historical Scheduler (CAHS), a novel scheduling baseline that selects agents based on historical viewpoint coverage, improving perception performance over existing SOTA methods. WHALES bridges the gap between simulated and real-world V2X challenges, providing a robust framework for exploring perception-scheduling co-design, cross-data generalization, and scalability limits. The WHALES dataset and code are available at https://github.com/chensiweiTHU/WHALES.

Collaborative perception technology enhances the perception range and accuracy in complex scenarios by enabling vehicles to share their sensor data. However, due to communication bandwidth limitations, it is important to prioritize the transmission of the most critical sensor data. In the context of multi-target tracking, this paper proposes a multi-level sensor scheduling framework for collaborative perception. The framework quantifies the collaborative gain from object-level, featurelevel, and track-level data across different collaborative vehicles, and employs covariance intersection to address the issue of unknown correlations in multi-vehicle collaborative perception and tracking. By introducing auxiliary variables and employing relaxation techniques, the original scheduling problem is transformed into a convex optimization problem, resulting in a low-complexity sensor selection algorithm. Simulation results demonstrate that the proposed algorithm outperforms the benchmark algorithms. Ablation experiments further reveal that adding feature-level and track-level fusion on top of object-level fusion yields significant collaborative gains.

Autonomous driving technologies are swiftly evolving, characterized by two main strategies: Single-Vehicle Autonomous Driving (SVAD) and Vehicle-Infrastructure Cooperative Autonomous Driving (VICAD). SVAD depends entirely on the vehicle’s internal sensors and processing capabilities, whereas VICAD benefits from a synergistic network combining roadside infrastructure, connected vehicles, and cloud services to boost safety and efficiency. Nevertheless, VICAD encounters challenges with high-bandwidth data transmission and perception latency. To mitigate these concerns, we introduce an innovative intelligent roadside unit (I-RSU) platform integrating perception, computing, and communication into one cohesive system. The platform features dual neural processing units (NPUs) for the effective extraction of images and LiDAR features, alongside a C-V2X communication module, all realized on a Field-Programmable Gate Array (FPGA). This setup minimizes latency and expenses by enabling computation near the sensors and facilitating selective data transmission. Our system also supports multi-modal fusion, enhancing overall perception and safety. Through extensive real-world trials and simulations, our system demonstrates a substantial reduction in end-to-end latency, providing a scalable solution for VICAD scenarios.

This paper studies an internet of things (IoT) network where a fusion center relies on multi-view and correlated information generated by multiple sources to monitor various regions. Each region possesses hard age of correlated information (AoCI) constraints for information update, and accordingly we propose a scheduling policy to satisfy such needs and minimize the required wireless resources. We first approximate the problem to a dual bin-packing problem. Secondly, efficient scheduling policies are identified when the age constraints possess special mathematical properties, where the number of channels at most required is analyzed. Optimality conditions of the proposed policies are presented. For general constraints, a two-step grouping algorithm for multi-view (TGAM) is proposed to establish scheduling policies. Under TGAM, the constraints are mapped into a combination of the special constraints. To quickly identify an optimized mapping from a vast solution space, TGAM heuristically groups the regions according to their constraints and then searches for the optimal mapping for each group. Numerical results demonstrate that, compared to a derived lower bound, the proposed TGAM requires only 1.07% more channels. Additionally, the number of regions that can be served by TGAM is significantly larger than the state-of-the art algorithm, given the number of channels.

In edge intelligence systems, task offloading and resource allocation policies critically depend on the required computing capacity of the task, which can only be accurately measured after execution, presenting significant design challenges. In this paper, we address the problem of robust task offloading and resource allocation under imperfect computing capacity information, where the exact value as well as distribution knowledge of the required computing capacity cannot be obtained in advance. Specifically, we formulate theenergy-time cost(ETC) minimization problem using min-max robust optimization. To tackle this challenging issue, we propose a decoupling method. This method first assumes the offloading policy is predetermined and derives two independent subproblems: local ETC and edge ETC. Then, we provide a closed-form optimal solution for the local ETC problem. The edge ETC problem is equivalently transformed into a geometric programming (GP) problem, and we introduce an effective iterative algorithm to obtain a stationary point, utilizing successive convex approximation (SCA). Finally, we design a coordinate descent (CD)-based algorithm to optimize the offloading policy effectively. Extensive simulations demonstrate that the proposed policy significantly outperforms other benchmark methods, achieving near-optimal performance even in the presence of high estimation errors in computing capacity.

For servers incorporating parallel computing resources, batching is a pivotal technique for providing efficient and economical services at scale. Parallel computing resources exhibit heightened computational and energy efficiency when operating with larger batch sizes. However, in the realm of online services, the adoption of a larger batch size may lead to longer response times. This paper aims to provide a dynamic batching scheme that delicately balances latency and efficiency. The system is modeled as a batch service queue with size-dependent service times. Then, the design of dynamic batching is formulated as a semi-Markov decision process (SMDP) problem, with the objective of minimizing the weighted sum of average response time and average power consumption. A method is proposed to derive an approximate optimal SMDP solution, representing the chosen dynamic batching policy. By introducing an abstract cost to reflect the impact of “tail” states, the space complexity and the time complexity of the procedure can decrease by 63.5% and 98%, respectively. Numerical results showcase the superiority of SMDP-based batching policies across various parameter setups. Additionally, the proposed scheme exhibits noteworthy flexibility in balancing power consumption and latency.

Wireless baseband processing (WBP) serves as an ideal scenario for utilizing vector processing, which excels in managing data-parallel operations due to its parallel structure. However, conventional vector architectures face certain constraints such as limited vector register sizes, reliance on power-of-two vector length (VL) multipliers, and vector permutation capabilities tied to specific architectures. To address these challenges, we have introduced an instruction set extension (ISE) based on RISC-V known as unlimited vector processing (UVP). This extension enhances both the flexibility and efficiency of vector computations. UVP employs a novel programming model that supports non-power-of-two register groupings (RGs) and hardware strip mining, thus enabling smooth handling of vectors of varying lengths while reducing the software strip-mining burden. Vector instructions are categorized into symmetric and asymmetric classes, complemented by specialized load/store strategies to optimize execution. Moreover, we present a hardware implementation of UVP featuring sophisticated hazard detection mechanisms, optimized pipelines for symmetric tasks such as fixed-point multiplication and division, and a robust permutation engine for effective asymmetric operations. Comprehensive evaluations demonstrate that UVP significantly enhances performance, achieving up to$3.0\times $and$2.1\times $speedups in matrix multiplication and fast Fourier transform (FFT) tasks, respectively, when measured against lane-based vector architectures. Our synthesized register transfer level (RTL) for a 16-lane configuration using SMIC 40-nm technology spans 0.94 mm2and achieves an area efficiency of 21.2 GOPS/mm2.

Leveraging the computing and sensing capabilities of vehicles, vehicular federated learning (VFL) has been applied to edge training for connected vehicles. The dynamic and inter-connected nature of vehicular networks presents unique opportunities to harness direct vehicle-to-vehicle (V2V) communications, enhancing VFL training efficiency. In this paper, we formulate a stochastic optimization problem to optimize the VFL training performance, considering the energy constraints and mobility of vehicles, and propose a V2V-enhanced dynamic scheduling (VEDS) algorithm to solve it. The model aggregation requirements of VFL and the limited transmission time due to mobility result in a stepwise objective function, which presents challenges in solving the problem. We thus propose a derivative-based drift-plus-penalty method to convert the long-term stochastic optimization problem to an online mixed integer nonlinear programming (MINLP) problem, and provide a theoretical analysis to bound the performance gap between the online solution and the offline optimal solution. Further analysis of the scheduling priority reduces the original problem into a set of convex optimization problems, which are efficiently solved using the interior-point method. Experimental results demonstrate that compared with the state-of-the-art benchmarks, the proposed algorithm enhances the image classification accuracy on the CIFAR-10 dataset by 4.20% and reduces the average displacement errors on the Argoverse trajectory prediction dataset by 9.82%.

In wireless body area networks (WBANs), the deep channel fading between the nodes and the hub significantly impairs the reliability of end-to-end signal transmission. However, some nodes in WBANs necessitate high-priority data transmission with stringent latency and accuracy requirements. Retransmission is ineffective against channel fading and can result in increased communication overhead and extended transmission delays. This paper proposes an opportunistic relay strategy tailored to the characteristics of WBAN nodes with varying priorities. This strategy converts suitable low-priority nodes as relays during high-priority time slots to forward high-priority data to the hub when deep fading occurs. The relay can be opportunistically selected based on the channel condition and the power usage. With Lyanupov optimization, we maximize the transmission reliability while ensuring the extra power consumption of low-priority nodes are acceptable. Subsequently, simulations are conducted in the Network Simulator 3 (NS3) to validate the proposed relay selection strategy, showing that the proposed strategy effectively improves the reliability from 90.2 % to 99.1 % with an additional power consumption of 30%.

Environmental perception in Automated Valet Parking (AVP) has been a challenging task due to severe occlusions in parking garages. Although Collaborative Perception (CP) can be applied to broaden the field of view of connected vehicles, the limited bandwidth of vehicular communications restricts its application. In this work, we propose a BEV feature-based CP network architecture for infrastructure-assisted AVP systems. The model takes the roadside camera and LiDAR as optional inputs and adaptively fuses them with onboard sensors in a unified BEV representation. Autoencoder and downsampling are applied for channel-wise and spatial-wise dimension reduction, while sparsification and quantization further compress the feature map with little loss in data precision. Combining these techniques, the size of a BEV feature map is effectively compressed to fit in the feasible data rate of the NR-V2X network. With the synthetic AVP dataset, we observe that CP can effectively increase perception performance, especially for pedestrians. Moreover, the advantage of infrastructure-assisted CP is demonstrated in two typical safety-critical scenarios in the AVP setting, increasing the maximum safe cruising speed by up to 3m/s in both scenarios.

In recent years, the concept of collaborative perception (CP) in self-driving vehicles has emerged as a new paradigm for augmenting the safety and efficiency of connected autonomous vehicles (CAVs). However, CP's energy consumption remains a major concern, due to their computation- and transmission-intensive characteristics. To address this issue, this paper first presents a theoretical definition of CP coverage along with a 2-dimensional CP model, followed by a novel framework that leverages roadside units (RSU) to facilitate CP, namely the RSU-Aided Energy-Efficient Sensing, Computation, and Communication (RE2SCC). Through a mix of centralized scheduling and a decentralized data-sharing approach, RE2SCC improves perception performance and energy efficiency. The core of RE2SCC is a novel approach for reducing the overall computation load and energy-efficient CP by scheduling computation and transmission depending on CAVs topology while maintaining the perception performance. The centralized scheduling exploits CP capabilities via sensing data selection, avoiding redundant computation, and direct transmission of perception object data to CAVs, enabling extended perception while minimizing the transmission power. Simulations show the efficiency of the RE2SCC framework for energy savings along with increased perception performance by up to 51% in a given scenario.

In next-generation wireless networks, the integration of localization and communications techniques are regarded as a paradigmatic shift for enhancing spectrum and hardware utilizations. The channel sensing, encompassing localization and channel estimation, plays a pivotal role in various aspects such as beamforming, precoding and high-quality data transmission. In this paper, we present a method for optimizing the frame structure in the localization and communication integration system, to reveal the intricate relationship among channel estimation, user's localization and communication throughput in terms of the spectral efficiency (SE). Specifically, we initially derive the error bounds for channel estimation and location prediction in dynamic point-to-point communication scenario. Leveraging these bounds, we optimize the sensing and communication duration together with beamwidth design, to maximize the SE while ensuring communication requirements. An efficient iterative algorithm is employed to tackle this non-convex problem. Numerical results demonstrate that our proposed method can achieve a nearoptimal SE performance with significantly lower complexity compared to exhaustive search method. Furthermore, our results underscore the critical role of localization in optimizing sensing and communication durations for SE, particularly in high dynamic scenarios.

With the rapid advancement of intelligent vehicle technology, the demand for vehicular computing power is increasing. To alleviate computing loads on the on-board computer, this paper proposes an in-vehicle distributed computing system that leverages in-vehicle devices, such as smartphones and tablet computers, for cooperative task execution. Different from previous works that focused on the impact of CPU frequency management for computation offloading, we consider the scenario where the CPU frequency of devices cannot be adjusted, and study a more practical approach to make offloading and scheduling decisions by considering the CPU utilization of devices. Therefore, we first derive an analytical relationship between CPU utilization and computation latency, and then propose a CPU Utilization-Aware Scheduling (CUAS) policy to minimize response latency consisting of computation and communication latency. Simulations conducted on Simgrid show that our proposed CUAS policy can reduce the response latency by up to 20.60% compared with the benchmarks. Additionally, we established a real-world testbed to validate our system's practicality. Experimental results indicate that our proposed policy can reduce response latency by up to 20.75% compared with the benchmarks.

To improve the performance of multi-agent reinforcement learning under the constraint of wireless resources, we propose a message importance metric and design an importance-aware scheduling policy to effectively exchange messages. The key insight is spending the precious communication resources on important messages. The message importance depends not only on the messages themselves, but also on the needs of agents who receive them. Accordingly, we propose a query-message-based architecture, called QMNet. Agents generate queries and messages with the environment observation. Sharing queries can help calculate message importance. Exchanging messages can help agents cooperate better. Besides, we exploit the message importance to deal with random access collisions in decentralized systems. Furthermore, a message prediction mechanism is proposed to compensate for messages that are not transmitted. Finally, we evaluate the proposed schemes in a traffic junction environment, where only a fraction of agents can send messages due to limited wireless resources. Results show that QMNet can extract valuable information to guarantee the system performance even when only 30% of agents can share messages. By exploiting message prediction, the system can further save 40% of wireless resources. The importance-aware decentralized multi-access mechanism can effectively avoid collisions, achieving almost the same performance as centralized scheduling.

Network slicing has been envisioned to play a crucial role in supporting various vehicular applications with diverse performance requirements in dynamic Vehicle-to-Everything (V2X) communications systems. However, time-varying Service Level Agreements (SLAs) of slices and fast-changing network topologies in V2X scenarios may introduce new challenges for enabling efficient inter-slice resource provisioning to guarantee the Quality of Service (QoS) while avoiding both resource over-provisioning and under-provisioning. Moreover, the conventional centralized resource allocation schemes requiring global slice information may degrade the data privacy provided by dedicated resource provisioning. To address these challenges, in this paper, we propose a two-timescale resource management mechanism for providing diverse V2X slices with customized resources. In the long timescale, we propose a Proximal Policy Optimization-based multi-agent deep reinforcement learning algorithm for dynamically allocating bandwidth resources to different slices for guaranteeing their SLAs. Under the coordination of agents, each agent only observes its partial state space rather than the global information to adjust the resource requests, which can enhance the privacy protection. Moreover, an expert demonstration mechanism is proposed to guide the action policy for reducing the invalid action exploration and accelerating the convergence of agents. In the short-term time slot, with our proposed Cross Entropy and Successive Convex Approximation algorithm, each slice allocates its available physical resource blocks and optimizes its transmit power to meet the QoS. Simulation results show our proposed two-timescale resource allocation scheme for network slicing can achieve maximum 8.4% performance gains in terms of spectral efficiency while guaranteeing the QoS requirements of users compared to the baseline approaches.

Open radio access network (O-RAN) is considered as a viable method for reducing the cost and enhancing the energy efficiency of cellular networks, due to its native incorporation of intelligence and open interfaces. However, the processing delay of software-based wireless protocol stacks has hindered its development. This paper presents the implementation of parallel computing acceleration for an LTE baseband system based on single program multiple data (SPMD) methodology, and proposes detailed optimization strategies for the time-consuming synchronization and OFDM modulation modules in the system. Experiment results based on the implicit SPMD program compiler (ISPC) show that continuous memory access has a significant impact on the final acceleration effect. Moreover, the processing speed of software-based physical layer can be increased up to 10–30 times through SPMD.

Neural interfaces play a crucial role in biomedical engineering since they serve as pathways of communication linking the brain and computers. Integrating wireless neural interfaces with a Media Access Control (MAC) protocol presents an efficient method for recording and analyzing neural activities simultaneously from multiple sensor nodes. However, current wireless networks, such as Wireless Sensor Networks (WSNs) and Wireless Body Area Networks (WBANs), are not directly applicable to neural interface networks, due to the high complexity and weak priority access support. This paper proposes a MAC protocol called NI-MAC that not only includes priority access but also simplifies the superframe structure and frame formats. Both random access and managed access are included in the superframe design. Analysis of transmission latency and energy consumption shows the improvement over IEEE 802.15.4 and IEEE 802.15.6. With NS3 simulator, we compare the proposed NI-MAC with existing protocols and demonstrate that: transmission latency of high-priority data can achieve 12ms, while under IEEE 802.15.4 it is 22ms; as the number of access nodes increases, energy consumption is reduced by up to 50% compared to IEEE 802.15.6.

Federated learning (FL) is a promising approach to enable the future Internet of vehicles consisting of intelligent connected vehicles (ICVs) with powerful sensing, computing and communication capabilities. We consider a base station (BS) coordinating nearby ICVs to train a neural network in a collaborative yet distributed manner, in order to limit data traffic and privacy leakage. However, due to the mobility of vehicles, the connections between the BS and ICVs are short-lived, which affects the resource utilization of ICVs, and thus, the convergence speed of the training process. In this paper, we propose an accelerated FL-ICV framework, by optimizing the duration of each training round and the number of local iterations, for better convergence performance of FL. We propose a mobility-aware optimization algorithm called MOB-FL, which aims at maximizing the resource utilization of ICVs under short-lived wireless connections, so as to increase the convergence speed. Simulation results based on the beam selection and the trajectory prediction tasks verify the effectiveness of the proposed solution.

In up-to-date machine learning (ML) applications on cloud or edge computing platforms, batching is an important technique for providing efficient and economical services at scale. In particular, parallel computing resources on the platforms, such as graphics processing units (GPUs), have higher computational and energy efficiency with larger batch sizes. However, larger batch sizes may also result in longer response time, and thus it requires a judicious design. This paper aims to provide a dynamic batching policy that strikes a balance between efficiency and latency. The GPU-based inference service is modeled as a batch service queue with batch-size dependent processing time. Then, the design of dynamic batching is a continuous-time average-cost problem, and is formulated as a semi-Markov decision process (SMDP) with the objective of minimizing the weighted sum of average response time and average power consumption. The optimal policy is acquired by solving an associated discretetime Markov decision process (MDP) problem with finite state approximation and “discretization”. By introducing an abstract cost to reflect the impact of “tail” states, the space complexity and the time complexity of the procedure can decrease by 63.5% and 98 %, respectively. Our results show that the optimal policies potentially possess a control limit structure. Numerical results also show that SMDP-based batching policies can adapt to different traffic intensities and outperform other benchmark policies. Furthermore, the proposed solution has notable flexibility in balancing power consumption and latency.

The introduction of 5G has changed the wireless communication industry. Whereas previous generations of cellular technology are mainly based on communication for people, the wireless industry is discovering that 5G may be an era of communications that is mainly focused on machine-to-machine communication. The application of Ultra Reliable Low Latency Communication in factory automation is an area of great interest as it unlocks potential applications that traditional wired communications did not allow. In particular, the decrease in the inter-device distance has led to the discussion of coding schemes for these interference-filled channels. To meet the latency and accuracy requirements of URLLC, Non-orthogonal multiple access has been proposed but it comes with associated challenges. In order to combat the issue of interference, an enhanced version of Sliding window superposition coding has been proposed as a method of coding that yields performance gains in scenarios with high interference. This paper examines the abilities of this coding scheme in a broadcast network in 5G to evaluate its robustness in situations where interference is treated as noise in a factory automation setting. This work shows the improvements of enhanced sliding window superposition coding over benchmark protocols in the high-reliability requirement regions of block error rates $\approx 10^{-6}$

Vehicle-to-Everything network enabled connected autonomous driving has been regarded as a promising solution to realize advanced autonomous driving. However, non-ideal factors in wireless communication and localization severely limit the development. In this work, we propose MoRFF, a Mobility-robust Regional Features Fusion framework for multi-terminal multi-view object detection to realize wireless cooperative perception. To conquer the limited communication bandwidth, stochastic latency, and inaccurate positioning caused by wireless links and mobility, our method features a universal two-stage detection paradigm with deep metric learning, matching the same object from different viewpoints directly on the regional feature maps, and thus helps to greatly reduce the data size to transmit. Our proposed architecture only requires image data without any additional information such as geo-positions, sensor poses, or point clouds from LiDAR, and thus conducive to the promotion of connected autonomous driving. Experimental evaluations show that the proposed algorithm successfully benefits from other viewpoints, increases the detection precision of barely visible objects by 13.42%, and achieves tenfold promotion in communication bandwidth requirements. Furthermore, the proposed algorithm is robust under various communication delays.

Neural interfaces are essential in biomedical engineering as they establish communication pathways between the brain and computers. Integration of wireless neural interfaces with random access protocols offers an efficient method for recording and analyzing neural activities. However, current access protocols, such as IEEE 802.15.4 and IEEE 802.15.6, do not appropriately prioritize data types in neural interface networks. This paper proposes a random access protocol that distinguishes access priorities for different data types. We provide analytical models for the proposed random access algorithm, and accordingly by optimizing parameters, such as the Exclusive Access Phase (EAP) duration in the superframe, the reliability and latency performance is improved. The performance of the proposed random access protocol is evaluated using NS3 simulator. Results indicate that the theoretical analysis matches experimental results, with a transmission successful rate over 95% and a latency below 50ms when there are 15 nodes with average arrival interval no less than 100ms.

In this paper, a novel time-triggered reservation scheme is proposed to improve the performance of IEEE 802.11 medium access control (MAC) protocol. The proposed Reservation-based Distributed Channel Access (RDCA) protocol leverages the overhearing feature of the shared wireless channel and allows reservation-based data transmission with collision resolution. The reservation is realized in a time-triggered manner by setting extra timers on mobile stations. Through time-triggered reservation, a collision-free transmission scheduling can be achieved in a collaborative way without the assistance from the access point (AP). RDCA substantially reduces the delay and can function as the MAC layer for the wireless extension of time-sensitive networking (TSN). Meanwhile, the throughput of the network is increased due to the reduction of collisions. Moreover, RDCA also maintains backward compatibility with the legacy IEEE 802.11 terminals. The analysis of saturated throughput is provided and the performance gain in terms of delay and throughput is evaluated by extensive simulations. Compared to legacy IEEE 802.11, the throughput in RDCA increases up to 50% and the delay reduces up to 70%.

We consider collaborative perception (CP) systems where a fusion center monitors various regions by multiple sources. The center has different age of information (AoI) constraints for different regions. Multi-view sensing data for a region generated by sources can be fused by the center for a reliable representation of the region. To ensure accurate perception, differences between generation time of asynchronous status updates for CP fusion should not exceed a certain threshold. An algorithm named scheduling for CP with asynchronous status updates (SCPA) is proposed to minimize the number of required channels and subject to AoI constraints with asynchronous status updates. SCPA first identifies a set of sources that can satisfy the constraints with minimum updating rates. It then chooses scheduling intervals and offsets for the sources such that the number of required channels is optimized. According to numerical results, the number of channels required by SCPA can reach only 12% more than a derived lower bound.

The sensing and communication layers are both integral parts of the Internet-of-Things. Most of recent studies on sensory status update treat the information sensing and sensory data communication problems separately (i.e., a decoupled approach) and optimize specific latency metrics such as age of information relying on simplified models of communication networks or sensory traffic. In this paper, we propose a deeply integrated sensing and communication scheduling (S2) framework based on status-error-triggered update, focusing specifically on multiaccess wireless networks. We first analyze a motivating example consisting of two-state Markov sensors, showing that when both optimized, S2 outperforms the decoupled approach significantly. For sensors with random-walk state transitions, the closed-form Whittle's index with arbitrary status tracking error functions is presented and the indexability is established. Furthermore, a mean-field approach is applied such that the decentralized status update medium access control design is solved explicitly, for both homogeneous nodes and heterogeneous nodes in terms of status transition behaviors. According to the numerical results, the performance of the proposed S2 scheme is close to the optimum and better than the decoupled approach. In addition, a potential application of dynamic Channel State Information (CSI) update is presented, with CSI generated by a commercial ray-tracing simulator.

Cloud gaming is promising yet poses big challenges to wireless communications, due to its stringent requirements for low response delay and high reliability. In this paper, we propose a predictive frame transmission scheme (PFT) in cloud gaming, to predict and pre-transmit future game frames to users. The PFT scheme takes full advantage of good network states to transmit the predicted frames, which consequently reduces the frame loss rate (FLR) against the network dynamics. We first model a FLR minimization problem in the single-user system with the PFT scheme, which allocates packets to carry the predicted frames. The upper and lower bounds of FLR are derived, respectively. Then, we study the system with Markovian property, and derive the optimal packet allocation policy via Markov Decision Process. A near-optimal policy is also proposed with low-complexity. The PFT scheme is further extended to the multi-server multi-user scenario, in which the users are adaptively scheduled to multiple servers based on their different requirements. Finally, we extend the policy to fit the scenario without direct knowledge of the network state by exploiting the packet loss rate estimation. We set up a practical testbed to evaluate the proposed PFT scheme, showing the capability of decreasing the mean FLR from$7\%$to$1\%$.

In this paper, we study the problem of joint Task offloading and resource Allocation for vehicular edge computing with Result Feedback Delay (TARFD). Specifically, we consider a typical roadside unit (RSU) and vehicles within its coverage area, and optimize the task offloading decisions of vehicles as well as the uplink bandwidth allocation and the computation resources allocation on the RSU. The TARFD problem is formulated as a non-convex mixed integer nonlinear programming (MINLP) to minimize the average delay consisting of task offloading delay, task computation delay, and result feedback delay. We derive a lower bound of the optimum to the TARFD problem, based on which we propose an approximate algorithm of the TARFD problem, called A-TARFD. The A-TARFD algorithm can effectively deliver solutions for small-scale scenarios. To tackle large-scale scenarios, a low-complexity algorithm for the TARFD problem, called L-TARFD, is developed by constructing an iteratively updated sequence of locally tight approximate geometric programming (GP) problems. The L-TARFD algorithm can converge to a Karush-Kuhn-Tucker (KKT) point and forces the offloading decisions arbitrarily close to binary values. By comparison with the lower bound, simulation results show that the proposed two algorithms have near-optimal performance over a wide range of parameter settings.

Graphics processing units (GPUs) can improve deep neural network inference throughput via batch processing, where multiple tasks are concurrently processed. We focus on novel scenarios that the energy-constrained mobile devices offload inference tasks to an edge server with GPU. The inference task is partitioned into sub-tasks for a finer granularity of offloading and scheduling, and the user energy consumption minimization problem under inference latency constraints is investigated. To deal with the coupled offloading and scheduling introduced by concurrent batch processing, we first consider an offline problem with a constant edge inference latency and the same latency constraint. It is proven that optimizing the offloading policy of each user independently and aggregating all the same sub-tasks in one batch is optimal, and thus the independent partitioning and same sub-task aggregating (IP-SSA) algorithm is inspired. Further, the optimal grouping (OG) algorithm is proposed to optimally group tasks when the latency constraints are different. Finally, when future task arrivals cannot be precisely predicted, a deep deterministic policy gradient (DDPG) agent is trained to call OG. Experiments show that IP-SSA reduces up to 94.9% user energy consumption in the offline setting, while DDPG-OG outperforms DDPG-IP-SSA by up to 8.92% in the online setting.

The cooperation among intelligent agents is the key to build smarter real-world intelligent systems. Recent researches have shown that wireless communication plays a vital role in multi-agent cooperation. However, limited wireless resources become the bottleneck of large-scale multi-agent cooperation. Therefore, we focus on how to improve the performance of multi-agent reinforcement learning under the constraint of communication resources. To share more valuable information among agents under resource constraint, we formulate the message importance and design a decentralized scheduling policy with query mechanism, so that agents can effectively exchange messages according to their message importance. We further design a message prediction mechanism to compensate for those messages that are not scheduled for transmission in the current round. Finally, we evaluate the performance of the proposed schemes in a traffic junction environment, where only a fraction of agents can broadcast their messages in each round due to limited wireless resources. Results show that the importance-aware message exchange can extract valuable information to guarantee the system performance even when less than half of agents can share their states. By exploiting message prediction, the system can further save 40% of communication resources while guaranteeing the system performance.

Cooperative perception of connected vehicles comes to the rescue when the field of view restricts stand-alone intelligence.While raw-level cooperative perception preserves most information to guarantee accuracy, it is demanding in communication bandwidth and computation power.Therefore, it is important to schedule the most beneficial vehicle to share its sensor in terms of supplementary view and stable network connection.In this paper, we present a model of raw-level cooperative perception and formulate the energy minimization problem of sensor sharing scheduling as a variant of the Multi-Armed Bandit (MAB) problem.Specifically, volatility of the neighboring vehicles, heterogeneity of V2X channels, and the time-varying traffic context are taken into consideration.Then we propose an online learning-based algorithm with logarithmic performance loss, achieving a decent trade-off between exploration and exploitation.Simulation results under different scenarios indicate that the proposed algorithm quickly learns to schedule the optimal cooperative vehicle and saves more energy as compared to baseline algorithms.

Federated edge learning (FEEL) is a promising distributed machine learning (ML) framework to drive edge intelligence applications. However, due to the dynamic wireless environments and the resource limitations of edge devices, communication becomes a major bottleneck. In this work, we propose time-correlated sparsification with hybrid aggregation (TCS-H) for communication-efficient FEEL, which exploits jointly the power of model compression and over-the-air computation. By exploiting the temporal correlations among model parameters, we construct a global sparsification mask, which is identical across devices, and thus enables efficient model aggregation over-the-air. Each device further constructs a local sparse vector to explore its own important parameters, which are aggregated via digital communication with orthogonal multiple access. We further design device scheduling and power allocation algorithms for TCS-H. Experiment results show that, under limited communication resources, TCS-H can achieve significantly higher accuracy compared to the conventional top-K sparsification with orthogonal model aggregation, with both i.i.d. and non-i.i.d. data distributions.

In a system with energy-constrained sensors, each transmitted observation comes at a price. The price is the energy the sensor expends to obtain and send a new measurement. The system has to ensure that sensors’ updates are timely, i.e., their updates represent the observed phenomenon accurately, enabling services to make informed decisions based on the information provided. If there are multiple sensors observing the same physical phenomenon, it is likely that their measurements are correlated in time and space. To take advantage of this correlation to reduce the energy use of sensors, in this paper we consider a system in which a gateway sets the intervals at which each sensor broadcasts its readings. We consider the presence of battery-powered sensors as well as sensors that rely on Energy Harvesting (EH) to replenish their energy. We propose a Deep Reinforcement Learning (DRL)-based scheduling mechanism that learns the appropriate update interval for each sensor, by considering the timeliness of the information collected measured through the Age of Information (AoI) metric, the spatial and temporal correlation between readings, and the energy capabilities of each sensor. We show that our proposed scheduler can achieve near-optimal performance in terms of the expected network lifetime.

Machine learning and wireless communication technologies are jointly facilitating an intelligent edge, where federated edge learning (FEEL) is emerging as a promising training framework. As wireless devices involved in FEEL are resource limited in terms of communication bandwidth, computing power and battery capacity, it is important to carefully schedule them to optimize the training performance. In this work, we consider an over-the-air FEEL system with analog gradient aggregation, and propose an energy-aware dynamic device scheduling algorithm to optimize the training performance within the energy constraints of devices, where both communication energy for gradient aggregation and computation energy for local training are considered. The consideration of computation energy makes dynamic scheduling challenging, as devices are scheduled before local training, but the communication energy for over-the-air aggregation depends on the$l_{2}$-norm of local gradient, which is known only after local training. We thus incorporate estimation methods into scheduling to predict the gradient norm. Taking the estimation error into account, we characterize the performance gap between the proposed algorithm and its offline counterpart. Experimental results show that, under a highly unbalanced local data distribution, the proposed algorithm can increase the accuracy by 4.9% on CIFAR-10 dataset compared with the myopic benchmark, while satisfying the energy constraints.

A typical Internet-of-Things (IoT) system consists of three major layers: 1) sensing; 2) communication; and 3) application (i.e., actuation and control) layers. The co-design of these layers has been studied for over two decades, dating back to the concept of communication, computing, and control, i.e., 3C, convergence in the 1990s. Nowadays, with the emergence of wireless-networked machine-type applications, such as connected autonomous driving and factory automation, this co-design is more urgently desired than ever to meet the stringent quality-of-service requirements thereof. To realize this goal, the 5G wireless network of today has mainly focused on the communication part and strived to reliably achieve low air-interface communication delay, i.e., ultra-reliable and low-latency communications (uRLLC). However, more and more wireless communications in IoT are based on status updates instead of general content delivery. The current uRLLC design is insufficient to characterize the status update quality, and thus is unable to optimize for timely status update with constrained wireless resources. Therefore, the performance of computing and control in IoT networks that rely highly on wireless communications is suboptimal.

We consider a base station supporting two types of traffics, i.e., status update traffic and timely throughput traffic. The goal is to improve the information freshness of status update traffic while satisfying timely throughput constraints. Age of Information (AoI) is adopted as a metric for information freshness. We first propose an age-aware policy that makes scheduling decisions based on the current value of AoI directly. Given timely throughput constraint, an upper bound of the weighted average AoI under this policy is provided. To evaluate policy performance, it is important to obtain the minimum weighted average AoI achievable given timely throughput constraint. A low complexity method is proposed to estimate a lower bound of this value. Furthermore, inspired by the estimation procedure, we design an age-oblivious policy that does not rely on the current AoI to make scheduling decisions. Surprisingly, simulation results show that the weighted average AoI of the age-oblivious policy is comparable to that of the age-aware policy, and both are close to the lower bound.

In federated learning (FL), devices contribute to the global training by uploading their local model updates via wireless channels. Due to limited computation and communication resources, device scheduling is crucial to the convergence rate of FL. In this paper, we propose a joint device scheduling and resource allocation policy to maximize the model accuracy within a given total training time budget for latency constrained wireless FL. A lower bound on the reciprocal of the training performance loss, in terms of the number of training rounds and the number of scheduled devices per round, is derived. Based on the bound, the accuracy maximization problem is solved by decoupling it into two sub-problems. First, given the scheduled devices, the optimal bandwidth allocation suggests allocating more bandwidth to the devices with worse channel conditions or weaker computation capabilities. Then, a greedy device scheduling algorithm is introduced, which selects the device consuming the least updating time obtained by the optimal bandwidth allocation in each step, until the lower bound begins to increase, meaning that scheduling more devices will degrade the model accuracy. Experiments show that the proposed policy outperforms state-of-the-art scheduling policies under extensive settings of data distributions and cell radius.

In a vehicular edge computing (VEC) system, vehicles can share their surplus computation resources to provide cloud computing services. The highly dynamic environment of the vehicular network makes it challenging to guarantee the task offloading delay. To this end, we introduce task replication to the VEC system, where the replicas of a task are offloaded to multiple vehicles at the same time, and the task is completed upon the first response among replicas. First, the impact of the number of task replicas on the offloading delay is characterized, and the optimal number of task replicas is approximated in closed-form. Based on the analytical result, we design a learning-based task replication algorithm (LTRA) with combinatorial multi-armed bandit theory, which works in a distributed manner and can automatically adapt itself to the dynamics of the VEC system. A realistic traffic scenario is used to evaluate the delay performance of the proposed algorithm. Results show that, under our simulation settings, LTRA with an optimized number of task replicas can reduce the average offloading delay by over 30% compared to the benchmark without task replication, and at the same time can improve the task completion ratio from 97% to 99.6%.

This paper studies the status update performance in wireless sensor networks when status, describing the physical reality that is being sensed, is temporally and spatially correlated. The status is modeled as a time-varying Gauss-Markov Random Field (GMRF), whereby the estimation error of status update at the fusion center is analyzed. The transmission latency introduced by wireless networks is modeled as exponentially distributed random variables. We extend the existing queuing analysis results for Age of Information (AoI) with uncorrelated sources to GMRF in the considered scenario. Closed-form expressions of average remote estimation error are obtained for both one- and two-dimensional GMRFs assuming the exponential time-correlation function, both First-Come First-Served (FCFS) and Last-Come First-Served (LCFS) service disciplines, and a single wireless link. The analytical results are then extended to scenarios wherein multi-packet reception, i.e., multiple concurrent wireless links, is enabled; the difficulty of analyzing obsolete updates in this case is addressed leveraging a reasonable approximation validated by theoretical analysis in the regime where the number of sensors is far more than that of wireless links. Monte-Carlo simulation results are also presented which agree with our theoretical analysis. Based on the results, optimal time and spatial domain sampling rates (e.g., sensor density) can be obtained, providing helpful guidance to wireless sensor deployment.

Mobile edge caching can effectively reduce service delay but may introduce information staleness, calling for timely content refreshing. However, content refreshing consumes additional transmission resources and may degrade the delay performance of mobile systems. In this work, we propose a freshness-aware refreshing scheme to balance the service delay and content freshness measured by Age of Information (AoI). Specifically, the cached content items will be refreshed to the up-to-date version upon user requests if the AoI exceeds a certain threshold (named as refreshing window). The average AoI and service delay are derived in closed forms approximately, which reveals an AoI-delay tradeoff relationship with respect to the refreshing window. In addition, the refreshing window is optimized to minimize the average delay while meeting the AoI requirements, and the results indicate to set a smaller refreshing window for the popular content items. Extensive simulations are conducted on the OMNeT++ platform to validate the analytical results. The results indicate that the proposed scheme can restrain frequent refreshing as the request arrival rate increases, whereby the average delay can be reduced by around 80% while maintaining the AoI below one second in heavily-loaded scenarios.

With the large-scale deployment of intelligent Internet of things (IoT) devices and the increasing need for computation support in wireless access networks, edge computing plays a vital role in satisfying these merging needs. The deployment of machine learning algorithms as one of the key applications at the network edge requires efficient training, in order to adapt themselves in the changing environment. However, the transmission of the training dataset collected by edge devices requires huge wireless communication resources. To address this issue, we exploit the fact that data samples have different importance for training, and use the training loss of the target machine learning model to represent the importance of data samples. Based on the importance metric, we propose a data upload scheme combining data compression that removes unimportant information and data filtering for data selection. As a result, the number of data samples as well as the size of every data sample to be transmitted can be substantially reduced while keeping the good training performance. Experiments show that training a machine learning model via the proposed scheme can enjoy faster convergence under limited wireless resources, specifically, we get almost the same training performance with only 2.5% of communication resources in the experiments training on MNIST dataset.

Owing to the increasing need for massive data analysis and model training at the network edge, as well as the rising concerns about the data privacy, a new distributed training framework called federated learning (FL) has emerged. In each iteration of FL (called round), the edge devices update local models based on their own data and contribute to the global training by uploading the model updates via wireless channels. Due to the limited spectrum resources, only a portion of the devices can be scheduled in each round. While most of the existing work on scheduling focuses on the convergence of FL w.r.t. rounds, the convergence performance under a total training time budget is not yet explored. In this paper, a joint bandwidth allocation and scheduling problem is formulated to capture the long-term convergence performance of FL, and is solved by being decoupled into two sub-problems. For the bandwidth allocation sub-problem, the derived optimal solution suggests to allocate more bandwidth to the devices with worse channel conditions or weaker computation capabilities. For the device scheduling sub-problem, by revealing the trade-off between the number of rounds required to attain a certain model accuracy and the latency per round, a greedy policy is inspired, that continuously selects the device that consumes the least time in model updating until achieving a good trade-off between the learning efficiency and latency per round. The experiments show that the proposed policy outperforms other state-of-the-art scheduling policies, with the best achievable model accuracy under training time budgets.

Federated learning (FL) enables workers to learn a model collaboratively by using their local data, with the help of a parameter server (PS) for global model aggregation. The high communication cost for periodic model updates and the nonindependent and identically distributed (i.i.d.) data become major bottlenecks for FL. In this work, we consider analog aggregation to scale down the communication cost with respect to the number of workers, and introduce data redundancy to the system to deal with non-i.i.d. data. We propose an online energy-aware dynamic worker scheduling policy, which maximizes the average number of workers scheduled for gradient update at each iteration under a long-term energy constraint, and analyze its performance based on Lyapunov optimization. Experiments using MNIST dataset show that, for non-i.i.d. data, doubling data storage can improve the accuracy by 9.8% under a stringent energy budget, while the proposed policy can achieve close-to-optimal accuracy without violating the energy constraint.

In this paper, we consider a relay selection scheme and analyze the corresponding cooperative diversity for over-the-air computation (AirComp) systems, where multiple source nodes transmit their signals over a wireless multi-access channel to achieve fast data aggregation. We first formulate the power control problems to minimize the computation mean square error (MSE) at the fusion center, and introduce the concept of MSE outage probability and diversity order in the context of AirComp. When there are no relays but multiple receive antennas at the fusion center, we propose an antenna selection scheme that selects the best antenna for reception. We then characterize its outage performance and prove that the AirComp diversity order is equal to the number of receive antennas. Motivated by the analogy between multiple-relay systems and multiple-antenna systems, we develop a relay selection scheme in relay-aided AirComp where only the best relay is chosen to amplify and forward its received signal to the fusion center. We show that the relay selection scheme can achieve the full diversity order, which is equal to the number of relays, and an outage performance comparable to AirComp with the same number of receive antennas.

With the development of intelligent Internet of things (IoT) devices, implementing machine learning algorithms at the network edge has become essential to many applications, such as autonomous driving, environment monitoring. However, the limited computation capability and energy constraint results in difficulties of running complex machine learning algorithms on edge devices subject to latency requirements, and one solution is to offload the computation tasks to the edge server. However, the wireless transmission of raw data from devices to the server is time consuming and may violate the latency requirement. To this end, lossy data compression can be helpful, but the information loss may lead to erroneous learning result, e.g., wrong classification. In this paper, we propose a transmission scheme with compression ratio selection for inference tasks with task completion latency guarantee. By dynamically selecting the optimal compression ratio with the awareness of the remaining latency budget, more tasks can be timely completed and get the correct inference results under the communication resource constraint. Furthermore, retransmitting less compressed data of tasks with erroneous inference results can potentially enhance the average accuracy. However, it is often hard to know whether the inference result is correct or not. We therefore use uncertainty to estimate the confidence of the results, and based on that, jointly optimize the retransmission and compression ratio selection.

Implementing machine learning algorithms on the Internet-of-Things (IoT) devices has become essential for emerging applications, such as autonomous driving and environment monitoring. But the limitations of computation capability and energy consumption make it difficult to run complex machine learning algorithms on IoT devices, especially when the latency deadline exists. One solution is to offload the computation-intensive tasks to the edge server. However, the wireless uploading of the raw data is time consuming and may lead to deadline violation. To reduce the communication cost, lossy data compression can be exploited for inference tasks but may bring more erroneous inference results. In this article, we propose a dynamic compression ratio selection scheme for edge inference system with hard deadlines. The key idea is to balance the tradeoff between the communication cost and inference accuracy. By dynamically selecting the optimal compression ratio with the remaining deadline budgets for queued tasks, more tasks can be timely completed with correct inference under limited communication resources. Furthermore, information augmentation that retransmits less compressed data of task with erroneous inference, is proposed to enhance the accuracy performance. While it is often hard to know the correctness of inference, we use uncertainty to estimate the confidence of the inference and based on that, jointly optimize the information augmentation and compression ratio selection. Finally, considering the wireless transmission errors, we further design a retransmission scheme to reduce performance degradation due to packet losses. The simulation results show the performance of the proposed schemes under different deadlines and task arrival rates.

The blockchain technology has achieved tremendous success in open (permissionless) decentralized consensus by employing Proof of Work (PoW) or its variants, whereby unauthorized nodes cannot gain a disproportionate impact on consensus beyond their computational power. However, PoW-based systems incur a high delay and low throughput, making them ineffective in dealing with the real-time Internet-of-Things (IoT) applications. On the other hand, the Byzantine fault-tolerant (BFT) consensus algorithms with better delay and throughput performance cannot be employed in permissionless settings due to vulnerability to Sybil attacks. In this article, we present a Sybil-proof wireless network coordinate-based Byzantine consensus (SENATE), which has the merits of both real-time consensus reaching and Sybil-proof, i.e., it is based on the conventional BFT consensus framework yet works in open systems of wireless devices where faulty nodes may launch Sybil attacks. As in a Senate, in the legislature, where the quota of senators per state (district) is a constant irrespective with the population of the state, “senators” in SENATE are selected from participating distributed nodes based on their wireless network coordinates (WNCs) with a fixed number of nodes per district in the WNC space. Elected senators then participate in the subsequent consensus reaching process and broadcast the result. Thereby, the SENATE is a proof against Sybil attacks since pseudonyms of a faulty node are likely to be adjacent in the WNC space and hence fail to be elected. The simulation results reveal that the SENATE can achieve real-time consensus (consensus delay under one second) in a network of hundreds of nodes.

Space-air-ground integrated networks (SAGIN) extend the capability of wireless networks and will be the essential building block for many advanced applications, like autonomous driving, earth monitoring, and etc. However, coordinating heterogeneous physical resources is very challenging in such a large-scale dynamic network. In this paper, we propose a reconfigurable service provisioning framework based on service function chaining (SFC) for SAGIN. In SFC, the network functions are virtualized and the service data needs to flow through specific network functions in a predefined sequence. The inherent issue is how to plan the service function chains over large-scale heterogeneous networks, subject to the resource limitations of both communication and computation. Specifically, we must jointly consider the virtual network functions (VNFs) embedding and service data routing. We formulate the SFC planning problem as an integer non-linear programming problem, which is NP-hard. Then, a heuristic greedy algorithm is proposed, which concentrates on leveraging different features of aerial and ground nodes and balancing the resource consumptions. Furthermore, a new metric, aggregation ratio (AR) is proposed to elaborate the communication-computation tradeoff. Extensive simulations shows that our proposed algorithm achieves near-optimal performance. We also find that the SAGIN significantly reduces the service blockage probability and improves the efficiency of resource utilization. Finally, a case study on multiple intersection traffic scheduling is provided to demonstrate the effectiveness of our proposed SFC-based service provisioning framework.

We consider a star-topology wireless network for status update where a central node collects status data from a large number of distributed machine-type terminals that share a wireless medium. The Age of Information (AoI) minimization scheduling problem is formulated by the restless multi-armed bandit. A widely-proven near-optimal solution, i.e., the Whittle's index, is derived in closed-form and the corresponding indexability is established. The index is then generalized to incorporate stochastic, periodic packet arrivals and unreliable channels. Inspired by the index scheduling policies which achieve near-optimal AoI but require heavy signaling overhead, a contention-based random access scheme, namely Index-Prioritized Random Access (IPRA), is further proposed. Based on IPRA, terminals that are not urgent to update, indicated by their indices, are barred access to the wireless medium, thus improving the access timeliness. A computer-based simulation shows that IPRA's performance is close to the optimal AoI in this setting and outperforms standard random access schemes. Also, for applications with hard AoI deadlines, we provide reliable deadline guarantee analysis. Closed-form achievable AoI stationary distributions under Bernoulli packet arrivals are derived such that AoI deadline with high reliability can be ensured by calculating the maximum number of supportable terminals and allocating system resources proportionally.

Multiple-input multiple-output (MIMO) and sparse code multiple access (SCMA) can be combined to achieve higher spectrum efficiency and more access for users, which also introduces more difficulties in signal detection. This paper explores low-complexity and low-latency iterative algorithms for soft symbol detection in an uplink MIMO-SCMA system over Rayleigh flat-fading channels. An expectation propagation framework (EPA) based on the extended factor graph is developed for MIMO-SCMA with multiantenna users. A new initialization method is proposed to accelerate convergence. Moreover, the SC-EPA with lower complexity is proposed by introducing QR decomposition and RE cluster-based decentralized factor node (FN) processing. Furthermore, new approaches for message passing between variable nodes (VNs) and FNs are proposed to improve the parallelism and reduce the complexity of the algorithm. The complexity of SC-EPA scales linearly with constellation size Ω (Ω <; M) and is independent of the receiving antenna Nr without any performance penalties. The robustness of the proposed algorithm in imperfect channels is evaluated, and the state evolution (SE) of the SC-EPA is derived. The link-level simulation results demonstrate that the EPA and SC-EPA receivers can achieve nearly the same performance as state of-the-art methods but with much lower complexity.

Functional split is a promising technique to flexibly balance the processing cost at remote ends and the fronthaul rate in cloud radio access networks (C-RAN). By harvesting renewable energy, remote radio units (RRUs) can save grid power and be flexibly deployed. However, the randomness of energy arrival poses a major design challenge. To maximize the throughput under the average fronthaul rate constraint in C-RAN with renewable powered RRUs, we first study the offline problem of selecting the optimal functional split modes and the corresponding durations, jointly with the transmission power. We find that between successive energy arrivals, at most two functional split modes should be selected. Then the optimal online problem is formulated as an Markov decision process (MDP). To deal with the curse of dimensionality of solving MDP, we further analyze the special case with one instance of energy arrival and two candidate functional split modes as inspired by the offline solution, and then a heuristic online policy is proposed. Numerical results show that with flexible functional split, the throughput can be significantly improved compared with fixed functional split. Also, the proposed heuristic online policy has similar performance with the optimal online one, as validated by simulations.

As 5G and Internet-of-Things (IoT) are deeply integrated into vertical industries such as autonomous driving and industrial robotics, timely status update is crucial for remote monitoring and control. In this regard, Age of Information (AoI) has been proposed to measure the freshness of status updates. However, it is just a metric changing linearly with time and irrelevant of context-awareness. We propose a context-based metric, named as Urgency of Information (UoI), to measure the nonlinear time-varying importance and the non-uniform context-dependence of the status information. This paper first establishes a theoretical framework for UoI characterization and then provides UoI-optimal status updating and user scheduling schemes in both single-terminal and multi-terminal cases. Specifically, an update-index-based scheme is proposed for a single-terminal system, where the terminal always updates and transmits when its update index is larger than a threshold. For the multi-terminal case, the UoI of the proposed scheduling scheme is proven to be upper-bounded and its decentralized implementation by Carrier Sensing Multiple Access with Collision Avoidance (CSMA/CA) is also provided. In the simulations, the proposed updating and scheduling schemes notably outperform the existing ones such as round robin and AoI-optimal schemes in terms of UoI, error-bound violation and control system stability.

Coded distributed computing framework enables large-scale machine learning (ML) models to be trained efficiently in a distributed manner, while mitigating the straggler effect. In this work, we consider a multi-task assignment problem in a coded distributed computing system, where multiple masters, each with a different matrix multiplication task, assign computation tasks to workers with heterogeneous computing capabilities. Both dedicated and probabilistic worker assignment models are considered, with the objective of minimizing the average completion time of all tasks. For dedicated worker assignment, greedy algorithms are proposed and the corresponding optimal load allocation is derived based on the Lagrange multiplier method. For probabilistic assignment, successive convex approximation method is used to solve the non-convex optimization problem. Simulation results show that the proposed algorithms reduce the completion time by 80% over uncoded scheme, and 49% over an unbalanced coded scheme.

Space-air-ground integrated network (SAGIN) brings great potentials to extend the terrestrial networks and satisfy the diverse service demands from many emerging applications. The major challenge is the coordination of large-scale networks with heterogeneous communication and computation resources. In this paper, flexible and reconfigurable service provisioning based on service function chaining (SFC) is exploited to address the challenge, where the traffic flow of the network services need to pass through specified virtual network functions (VNFs) in a given order. Our main target is to optimize the planning of the service function chains under limited heterogeneous resources and to map them on physical networks, considering the balance of resource utilization of both communication and computation. The SFC planning problem is formulated as an integer non-linear programming problem, which is NP-hard. Then, we propose a heuristic SFC planning algorithm (HSP) to reduce the computational complexity. Moreover, we propose a new metric, aggregation ratio (AR), to observe the tradeoff between communication and computation resource consumptions. The simulations results demonstrate that the HSP achieves near-optimal performance and the communication and computation resources can be well tradeoffed via tuning AR. The service blockage probability is significantly decreased and the efficiency of resource utilization is improved by integrating SAGIN based on SFC.

Mobile edge caching is an effective way to reduce the service delay of content delivery, where the popular contents can be pro-actively stored in proximity to users. In practice, the cached contents should be updated timely to avoid information staleness, in case that the information of a content changes with time and environment. However, cache update consumes additional transmission resources, which can degrade the delay performance. This work studies the fundamental tradeoff relationship between the content freshness (depicted by the age of information (AoI)) and service delay in mobile edge caching networks, and proposes a freshness-aware cache update scheme to achieve the AoI-delay balance. In specific, the base station will fetch the latest version of a content before delivery, if the AoI is larger than a certain threshold (i.e., update window size). The average AoI and service delay are derived in closed forms through approximated analysis of queueing systems, revealing a tradeoff relationship with respect to the update window size. Extensive simulations are conducted on the OMNeT++ platform, which validates the analytical results. Both the analytical and simulation results show that the proposed scheme can flexibly balance the average AoI and delay on demand, by tuning the update window size. Furthermore, the proposed scheme can also avoid frequent update in case of heavy content requests, whereby the AoI and delay are regulated by setting the appropriate update window size.

Emerging applications in Internet of Things (IoT), such as remote monitoring and control, extensively rely on timely status updates. Age of Information (AoI) has been proposed to characterize the freshness of information in status update systems. However, it only considers the time elapsed since the generation of the latest packet, and is incapable of capturing other critical information in remote control systems, such as the stochastic evolution and the importance of the source status. In order to evaluate the timeliness of status updates in remote control systems, we propose a context-aware metric, namely the context-aware information lapse. The context-aware information lapse characterizes both the stochastic evolution of the source status and the context-aware importance of the status. In this paper, the minimization of average context-aware lapse in a multi-user system is considered, and a corresponding user scheduling policy is proposed based on Lyapunov optimization. Numerical results show that compared to AoI-based policy, the context-aware-lapse-based policy can achieve a substantial improvement in terms of error- threshold violation probability and control performance.

Future wireless access networks need to support diversified quality of service (QoS) metrics required by various types of Internet-of-Things (IoT) devices, e.g., age of information (AoI) for status generating sources and ultra low latency for safety information in vehicular networks. In this paper, a novel inner-state driven random access (ISDA) framework is proposed based on distributed policy learning, in particular a cross-entropy method. Conventional random access schemes, e.g., p-CSMA, assume state-less terminals, and thus assigning equal priorities to all. In ISDA, the inner-states of terminals are described by a time-varying state vector, and the transmission probabilities of terminals in the contention period are determined by their respective inner-states. Neural networks are leveraged to approximate the function mappings from inner-states to transmission probabilities, and an iterative approach is adopted to improve these mappings in a distributed manner. Experiment results show that ISDA can improve the QoS of heterogeneous terminals simultaneously compared to conventional CSMA schemes.

Information source sampling and update scheduling have been treated separately in the context of real-time status update for age of information optimization. In this paper, a unified sampling and scheduling (S2) approach is proposed, focusing on decentralized updates in multiaccess wireless networks. To gain some insights, we first analyze an example consisting of two-state Markov sources, showing that when both optimized, the unified approach outperforms the separate approach significantly in terms of status tracking error by capturing the key status variation. We then generalize to source nodes with random-walk state transitions whose scaling limit is Wiener processes, the closed-form Whittle's index with arbitrary status tracking error functions is obtained and indexability established. Furthermore, a mean-field approach is applied to solve for the decentralized status update design explicitly. In addition to simulation results which validate the optimality of the proposed S2scheme and its advantage over the separate approach, a use case of dynamic channel state information (CSI) update is investigated, with CSI generated by a ray-tracing electromagnetic software.

Motivated by the increasingly urgent demands for delivering fresh information, the age-of-information (AoI) has recently been introduced as an important metric for evaluating the timeliness performance of information update systems and has shed light on a number of research studies. Nevertheless, the most common goal of the existing works does not characterize the value of information freshness from the users' perspective. In this paper, we introduce the concept of effective AoI (EAoI) to quantify the freshness of the information users utilize for decision-making. We consider a general request-response model, which captures both proactive information update and timely information delivery, for investigating the scheduling problem with respect to EAoI minimization. By decomposing the scheduling problem into multiple computationally tractable subproblems, we propose request-aware scheduling policies for static and dynamic request models, respectively. The numerical results show that serving users requests proactively can reduce time-average EAoI in both scenarios.

In many applications, status information of a general spatial process, in contrast to a point information source, is of interest. In this paper, we consider a system where status information is drawn from a random field and transmitted to a fusion center through a wireless multiaccess channel. The optimal density of spatial sampling points to minimize the remote status estimation error is investigated. Assuming a one-dimensional Gauss Markov random field and an exponential correlation function, closed-form expressions of remote estimation error are obtained for First-Come First-Served (FCFS) and Last-Come First-Served (LCFS) service disciplines. The optimal spatial sampling density for the LCFS case is given explicitly. Simulation results are presented which agree with our analysis.

In a typical Internet of Things (IoT) application where a central controller collects status updates from multiple terminals, e.g., sensors and monitors, through a wireless multiaccess uplink, an important problem is how to attain timely status updates autonomously. In this paper, the timeliness of the status is measured by the recently proposed age-of-information (AoI) metric; both the theoretical and practical aspects of the problem are investigated: we aim to obtain a scheduling policy with minimum AoI and, meanwhile, requires little signaling exchange overhead. Toward this end, we first consider the set of arrival-independent and renewal policies; the optimal policy thereof to minimize the time-average AoI is proved to be a round-robin policy with one-packet (latest packet only and others are dropped) buffers (RR-ONE). The optimality is established based on a generalized Poisson-arrival-see-time-average theorem. It is further proved that RR-ONE is asymptotically optimal among all policies in the massive IoT regime. The AoI steady-state stationary distribution under RR-ONE is also derived. An implementation scheme of RR-ONE is proposed which can accommodate dynamic terminal appearances with little overhead. In addition, considering scenarios where packets cannot be dropped, a Lyapunov optimization-based max-AoI-weight policy is proposed which achieves better performance compared with state-of-the-art.

Mobile device-to-device (D2D) network has now become a standardized feature in many mobile devices, by which mobile devices can communicate with each other even when commercial Internet access is not available. Because D2D network is expected to be an intrinsic part of the Internet of Things (IoT) and mobile device is the smartest and the most advanced commercial device in everyday usage, the D2D feature and related security protocols it adopts influences the design and implementation of many other IoT devices. While D2D network provides tangible benefits to users, it also raises the security risks of information leaking. This paper presents an in-depth empirical security analysis on mobile D2D network among Android devices. Android apps could establish a mobile D2D network in various ways, including Wi-Fi hotspot, Wi-Fi Direct, and Bluetooth. Those mobile D2D protocols normally take different protection mechanisms, which makes security investigation considerably challenging. In this paper, we focus on most popular apps in the Google Play Store, with aggregated downloads more than 500 million. Our analysis reveals some critical vulnerabilities. The key findings are bi-fold. First, the current mobile D2D network framework enabled by Android has significant flaw of overprivilege issue. Second, we have identified that most data transfer over mobile D2D network is unencrypted. Furthermore, we exploit the identified Android framework flaws to construct three proof-of-concept attacks and we conclude this paper with security lessons and suggestions of possible solutions against the identified security issues.

The high density and heterogeneous mobility of users in many applications pose challenges for the channel acquisition in massive multiple-input-multiple-output (MIMO) systems. For such scenarios, we propose an intermittent channel estimation (ICE) scheme to save pilot resources, which utilizes the aged channel state information (CSI) based on the temporal correlations of user channels. The optimal CSI update pattern to maximize the achievable sum rate is obtained by solving a formulated multichain Markov decision process (MDP), which is denoted by ICE-MDP. Furthermore, to reduce the computational complexity of the MDP, we relax the constraint of the CSI update pattern design problem and convert it into a convex optimization problem, whose solution is denoted by ICE-CVX. The simulations validate the close-to-optimal performance and the computational efficiency of ICE-CVX and show that the ICE scheme can significantly outperform a conventional scheme which persistently updates the CSI of all users.

The energy efficiency of future networks is becoming a significant and urgent issue, calling for greener network designs. However, the increasing complexity in network structure and resource space lead to growing problem scales and coupled resource dimensions, which bring great challenges in obtaining a joint solution in optimizing the energy efficiency. In this paper, we develop a multi-dimensional network model on the basis of tuple-links associated with transmission patterns (TPs) and formulate the optimization problem as a TP based scheduling problem which jointly solves transmission scheduling, routing, power control, radio, and channel assignment. In order to tackle the complexity issues, we propose a novel algorithm by exploiting the delay column generation technique to decompose the coupled problem into recursively solving a master problem for scheduling and a sub-problem for power allocation. Further, we theoretically prove that the performance gap between the proposed algorithm and the optimum is upper bounded by that for the sub-problem solution, where the latter is derived by solving a relaxed version of the sub-problem. Numerical results demonstrate the effectiveness of the multi-dimensional framework and the benefit of the proposed joint optimization in improving network energy efficiency.

Channel state information (CSI) plays a vital role in wireless communication systems. However, the CSI acquisition overhead is an enormous obstacle to realize the system performance improvements promised by massive connectivity and massive multiple-input-multiple-output (MIMO). To alleviate this overhead, this paper proposes a remote channel inference framework by probing the channels occupied by a source base station (BS) and inferring the channels of target BSs at geographically separated sites. The work generalizes existing literature which mainly focuses on utilizing the CSI linear correlations of adjacent antennas, by adopting a model-free deep learning framework to investigate non-linear dependence among remote CSI. The existence of such cross-BS CSI dependence is first shown by calculating the mutual information between remote channels, and the Cramér-Rao lower bound of remote CSI inference performance based on a one-ring channel model. Inspired by this finding, modern deep learning approaches are leveraged to perform remote channel inference in heterogeneous networks for both single user and multi-user scenarios. The simulation results based on ray tracing data show evident performance advantages over conventional methods, under both homogeneous and heterogeneous frequency coverage. The proposed framework achieves beamformer inference accuracy within 4.6% of the genie-aided optimum at the cost of sweeping only two beams.

Timely status updates are crucial to enabling applications in the massive Internet of Things (IoT). This paper measures the data-freshness performance of a status update system with an energy-harvesting transmitter, considering the randomness in information generation, transmission, and energy harvesting. The performance is evaluated by a non-linear function of age of information (AoI) that is defined as the time elapsed since the generation of the most up-to-date status information at the receiver. The system is formulated as two queues with status packet generation and energy arrivals both assumed to be Poisson processes. With negligible service time, both first-come-first-served (FCFS) and last-come-first-served (LCFS) disciplines for arbitrary buffer and battery capacities are considered, and a method for calculating the average penalty with non-linear penalty functions is proposed. The average AoI, the average penalty under exponential penalty function, and the AoI's threshold violation probability are obtained in a closed form. When the service time is assumed to follow exponential distribution, a matrix geometric method is used to obtain the average peak AoI. The results illustrate that under the FCFS discipline, the status update frequency needs to be carefully chosen according to the service rate and energy arrival rate in order to minimize the average penalty.

Timely and accurate knowledge of channel state information (CSI) is necessary to support scheduling operations at both physical and network layers. In order to support pilot-free channel estimation in cell sleeping scenarios, we propose to adopt a channel database that stores the CSI as a function of geographic locations. Such a channel database is generated from historical user records, which usually can not cover all the locations in the cell. Therefore, we develop a two-step interpolation method to infer the channels at the uncovered locations. The method firstly applies the K-nearest-neighbor method to form a coarse database and then refines it with a deep convolutional neural network. When applied to the channel data generated by ray tracing software, our method shows a great advantage in performance over the conventional interpolation methods.

Channel state information (CSI) is of vital importance in wireless communication systems. Existing CSI acquisition methods usually rely on pilot transmissions, and geographically separated base stations (BSs) with non-correlated CSI need to be assigned with orthogonal pilots which occupy excessive system resources. Our previous work adopts a data-driven deep learning based approach which leverages the CSI at a local BS to infer the CSI remotely, however the relevance of CSI between separated BSs is not specified explicitly. In this paper, we exploit a model-based methodology to derive the Cramer-Ran lower bound (CRLB) of remote CSI inference given the local CSI. Although the model is simplified, the derived CRLB explicitly illustrates the relationship between the inference performance and several key system parameters, e.g., terminal distance and antenna array size. In particular, it shows that by leveraging multiple local BSs, the inference error exhibits a larger power-law decay rate (w.r.t. number of antennas), compared with a single local BS; this explains and validates our findings in evaluating the deep-neural-network-based (DNN-based) CSI inference. We further improve on the DNN-based method by employing dropout and deeper networks, and show an inference performance of approximately 90% accuracy in a realistic scenario with CSI generated by a ray-tracing simulator.

In this paper, the wireless uplink is considered for status update with a large number of terminals. Thy key problem we address is that whether spatial multiplexing of multiple terminals, enabled by the massive multiple-input multiple-output technology, can help to improve the scaling law of age-of-information versus the number of terminals, on account of the mandatory pilot overhead. Based on a queuing theory analysis, we show that the proposed elastic spatial multiplexing scheme, which assigns an optimized pilot length that is smaller than the number of transmitting terminals on account of random packet arrivals, can indeed improve the scaling law compared with the optimal scaling law without spatial multiplexing, by a factor that is related to the packet lengths and arrival rates. Simulation results are provided to validate our findings.

Distributing popular contents, e.g., high precision digital maps and latest road conditions, through roadside units (RSUs) is a promising way to provide better driving safety and support for autonomous driving. Successful content download probability can be improved by allocating cache to RSUs and caching popular contents therein. In this paper, we consider a joint cache allocation and content placement problem in vehicular networks to maximize the overall successful content download probability, exploiting the moving information of vehicles. We first prove that the original problem is NP-hard, and then propose a low-complexity approximate algorithm which performs within a bounded gap (a multiplicity factor of 1 -1/ e) to the optimum. The expression of the cache size allocated to each RSU is derived for the proposed algorithm when the content popularity obeys Zipf distribution. Extensive numerical experiments show that, the proposed strategy can significantly increase the successful content download probability as compared to existing solutions.

In a vehicular edge computing (VEC) system, some vehicles with surplus computing resources can provide computation task offloading opportunities for other vehicles or pedestrians. However, the vehicular network is highly dynamic, with fast varying channel states and computation loads. These dynamics are difficult to model or to predict, but they have a major impact on the quality of service (QoS) of task offloading, including delay performance and service reliability. Meanwhile, the computing resources in VEC are often redundant due to the high density of vehicles. To improve the QoS of VEC and exploit the abundant computing resources on vehicles, we propose a learning-based task replication algorithm (LTRA) based on combinatorial multi-armed bandit (CMAB) theory, in order to minimize the average offloading delay. LTRA enables multiple vehicles to process the replicas of the same task simultaneously, and vehicles that require computing services can learn the delay performance of other vehicles while offloading tasks. We take the occurrence time of vehicles into consideration, and redesign the utility function of existing CMAB algorithm, so that LTRA can adapt to the time varying network topology of VEC. We use a realistic highway scenario to evaluate the delay performance and service reliability of LTRA through simulations, and show that compared with single task offloading, LTRA can improve the task completion ratio with deadline 0.6s from 80% to 98%.

One of the most challenging issues in deploying massive multiple-input multiple-output (MIMO) systems is the significant radio-frequency (RF) front-end complexity, hardware cost and power consumption. Towards this end, the beamspace-MIMO based approach is a promising solution. In this paper, we first show that traditional beamspace-MIMO approaches suffer from spatial power leakage and imperfect channel statistics estimation. A beam combination module is hence proposed, which consists of a small number (compared with the number of antenna elements) of low-resolution (possibly one- bit) digital (discrete) phase shifters after beamspace transformation to further compress the beamspace signal dimensionality, such that the number of RF chains can be reduced beyond beamspace transformation and beam selection. The optimum discrete beam combination weights are obtained based on the branch-and-bound (BB) approach. The key to the BB-based solution is to solve the embodied sub- problem, whose solution is derived in a closed-form. Link-level simulation results based on realistic channel models and LTE parameters are presented which show that the proposed schemes can reduce the number of RF chains by up to 25% with a one-bit phase-shifter-network.

Vehicular cloud computing (VCC) is proposed to effectively utilize and share the computing and storage resources on vehicles. However, due to the mobility of vehicles, the network topology, the wireless channel states and the available computing resources vary rapidly and are difficult to predict. In this work, we develop a learning-based task offloading framework using the multi-armed bandit (MAB) theory, which enables vehicles to learn the potential task offloading performance of its neighboring vehicles with excessive computing resources, namely service vehicles (SeVs), and minimizes the average offloading delay. We propose an adaptive volatile upper confidence bound (AVUCB) algorithm and augment it with load-awareness and occurrence-awareness, by redesigning the utility function of the classic MAB algorithms. The proposed AVUCB algorithm can effectively adapt to the dynamic vehicular environment, balance the tradeoff between exploration and exploitation in the learning process, and converge fast to the optimal SeV with theoretical performance guarantee. Simulations under both synthetic scenario and a realistic highway scenario are carried out, showing that the proposed algorithm achieves close-to- optimal delay performance.

We consider a system where multiple terminals transmit their randomly generated status updates to a base station (BS) sharing a wireless multiaccess uplink channel. The problem of interest, especially in massive Internet-of-Things systems, is that how to schedule the terminals to minimize the time-average age-of-information in a decentralized manner, namely terminals transmit autonomously without signalling exchange (overhead) with the BS or other terminals. Towards this end, the round-robin with one-packet buffers (the newest packet at each terminal only) policy (RR-ONE) is proposed and proved optimal among arrival-independent renewal (AIR) policies. In addition to its simple structure which is instrumental for decentralized implementation, RR-ONE is further proved asymptotically (massive terminals) optimal among all policies, including centralized and non-causal policies.

In vehicular cloud computing (VCC) systems, the computational resources of moving vehicles are exploited and managed by infrastructures, e.g., roadside units, to provide computational services. The offloading of computational tasks and collection of results rely on successful transmissions between vehicles and infrastructures during encounters. In this paper, we investigate how to provide timely computational services in VCC systems. In particular, we seek to minimize the deadline violation probability given a set of tasks to be executed in vehicular clouds. Due to the uncertainty of vehicle movements, the task replication methodology is leveraged which allows one task to be executed by several vehicles, and thus trading computational resources for delay reduction. The optimal task replication policy is of key interest. We first formulate the problem as a finite-horizon sampled-time Markov decision problem and obtain the optimal policy by value iterations. To conquer the complexity issue, we propose the balanced-task-assignment (BETA) policy which is proved optimal and has a clear structure: it always assigns the task with the minimum number of replicas. Moreover, a tight closed-form performance upper bound for the BETA policy is derived, which indicates that the deadline violation probability follows the Rayleigh distribution approximately. Applying the vehicle speed-density relationship in the traffic flow theory, we find that vehicle mobility benefits VCC systems more compared with road traffic systems, by showing that the optimum vehicle speed to minimize the deadline violation probability is larger than the critical vehicle speed in traffic theory which maximizes traffic flow efficiency.

Base station (BS) sleeping is an effective way to reduce the energy consumption of mobile networks. Previous efforts to design sleeping control algorithms mainly rely on stochastic traffic models and analytical derivation. However, the tractability of models often conflicts with the complexity of real-world traffic, making it difficult to apply in reality. In this paper, we propose a data-driven algorithm for dynamic sleeping control called DeepNap. This algorithm uses a deep Q-network (DQN) to learn effective sleeping policies from high-dimensional raw observations or un-quantized systems state vectors. We propose to enhance the original DQN algorithm with action-wise experience replay and adaptive reward scaling to deal with the challenges in nonstationary traffic. We also provide a model-assisted variant of DeepNap through the Dyna framework for inferring and simulating system dynamics. Periodical traffic modeling makes it possible to capture the nonstationarity in real-world traffic and the incorporation with DQN allows for feature learning and generalization from model outputs. Experiments show that both the end-to-end and the model-assisted version of DeepNap outperform table-based${Q}$-learning algorithm and the nonstationarity enhancements improve the stability of vanilla DQN.

In beam-based massive multiple-input multiple-output systems, signals are processed spatially in the radio-frequency (RF) front end and thereby the number of RF chains can be reduced to save hardware cost, power consumptions, and pilot overhead. Most existing work focuses on how to select or design analog beams to achieve performance close to full digital systems. However, since beams are strongly correlated (directed) to certain users, the selection of beams and scheduling of users should be jointly considered. In this paper, we formulate the joint user scheduling and beam selection problem based on the Lyapunov-drift optimization framework and obtain the optimal scheduling policy in a closed form. For reduced overhead and computational cost, the proposed scheduling schemes are based only upon statistical channel state information. Towards this end, asymptotic expressions of the downlink broadcast channel capacity are derived. To address the weighted sum rate maximization problem in the Lyapunov optimization, an algorithm based on block coordinated update is proposed and proved to converge to the optimum of the relaxed problem. To further reduce the complexity, an incremental greedy scheduling algorithm is also proposed, whose performance is proved to be bounded within a constant multiplicative factor. Simulation results based on widely-used spatial channel models are given. It is shown that the proposed schemes are close to optimal and outperform several state-of-the-art schemes.

In this paper, the joint user and beam scheduling problem in beam-based massive multiple-input multiple-output (MIMO) systems is formulated based on the Lyapunov-drift optimization framework and the optimal scheduling policy is given in a closed-form. To address the weighted sum rate maximization problem (mixed integer programming) arisen in the Lyapunov-drift maximization, an algorithm based on the block coordinated update is proposed and proved to converge to the global optimum of the relaxed convex problem. In order to make the scheduling decisions based only upon statistical channel state information (CSI), asymptotic expressions of the downlink broadcast channel capacity are derived. Simulation results based on widely-adopted spatial channel models are given, which show that the proposed scheme is close to the optimal scheduling scheme, and outperforms the state-of-the-art beam selection schemes.

In this paper, we propose a learning-based low-overhead channel estimation method for coordinated beamforming in ultra-dense networks. We first show through simulation that the channel state information (CSI) of geographically separated base stations (BSs) exhibits strong non-linear correlations in terms of mutual information. This finding enables us to adopt a novel learning-based approach to remotely infer the quality of different beamforming patterns at a dense-layer BS based on the CSI of an umbrella control-layer BS. The proposed scheme can reduce channel acquisition overhead by replacing pilot-aided channel estimation with the online inference from an artificial neural network, which is fitted offline. Moreover, we propose to use more anchor points and more candidate beam patterns to obtain better performance. Simulation results based on stochastic ray-tracing channel models show that the proposed scheme can reach an accuracy of 99.74\% in settings with 20 beamforming patterns.

The dense deployment of small-cell base stations (SBSs) endowed with cloud-like computing capabilities paves the way for pervasive mobile edge computing (MEC), enabling ultra-low latency and location-awareness for emerging mobile applications. To handle spatially imbalanced computation workloads in the network, cooperation among SBSs via peer offloading is essential to avoid large latency at overloaded SBSs and provide high quality of service to end users. However, performing effective peer offloading faces many challenges due to uncertainties of the system dynamics, limited energy budget committed by SBS owners and co- provisioning of radio access and computing services. This paper develops a novel online SBS peer offloading framework, called OPEN, by leveraging the Lyapunov technique, in order to maximize the long-term system performance while keeping the energy consumption of SBSs below individual long-term energy budget. OPEN works online without requiring future information of system dynamics, yet provides provably near-optimal performance compared to the oracle solution with complete future information. Extensive simulations are carried out and show that proposed algorithm dramatically improves the performance of edge computing system.

Multiuser multiple-input multiple-output (MIMO) systems suffer from a huge overhead of channel estimation for the application of the Internet of things (IoT), which demands the systems to support massive connectivity of users. An intermittent estimation scheme is proposed to ease the burden of channel acquisition. In the scheme, we exploit the temporal correlation of MIMO channels and analyze the influence of the age of CSI on the downlink transmission rate using linear precoders. We show the CSI updating interval of each user should follow a quasi-periodic distribution. The CSI updating frequency is optimized to balance between the accuracy of CSI estimation and the overhead of CSI acquisition.

Merging Mobile Edge Computing (MEC), which is an emerging paradigm to meet the increasing computation demands from mobile devices, with the dense deployment of Base Stations (BSs), is foreseen as a key step towards the next generation mobile networks. However, new challenges arise for designing energy efficient networks since radio access resources and computing resources of BSs have to be jointly managed, and yet they are complexly coupled with traffic in both spatial and temporal domains. In this paper, we address the challenge of incorporating MEC into dense cellular networks, and propose an efficient online algorithm, called ENGINE (ENerGy constrained offloadINg and slEeping) which makes joint computation offloading and BS sleeping decisions in order to maximize the quality of service while keeping the energy consumption low. Our algorithm leverages Lyapunov optimization technique, works online and achieves a close-to-optimal performance without using future information. Our simulation results show that our algorithm can effectively reduce energy consumption while guaranteeing quality of service for users.

In wireless communication systems powered by harvested energy, besides the channel fading, there is another dimension of dynamics induced by energy arrival variations, which makes the design of wireless transmission policies nontrivial. In this paper, we propose a framework for analyzing the energy harvesting powered wireless transmissions where the channel fading and the energy arrival variations are of different timescales. We define the duration between two consecutive changes of energy arrival rate as an energy harvesting frame, which consists of N channel fading slots. The power allocation problem can be formulated as an Markov decision process (MDP), and can be decoupled into two sub-problems. The inner problem deals with the power allocation in channel fading timescale in every N slots where the energy arrival rate keeps constant, and the outer problem deals with the energy management in energy harvesting timescale among frames. The two sub-problems can be solved by finite horizon dynamic programming (DP) and infinite horizon DP, respectively. Numerical simulations show that the average rate decreases slightly as N increases, and the rate under i.i.d. channel is higher than that under Markov channel.

With the development of wireless networks, the scale of network optimization problems is growing correspondingly. While algorithms have been designed to reduce complexity in solving these problems under given size, the approach of directly reducing the size of problem has not received much attention. This motivates us to investigate an innovative approach to reduce problem scale while maintaining the optimality of solution. Through analysis on the optimization solutions, we discover that part of the elements may not be involved in the solution, such as unscheduled links in the flow constrained optimization problem. The observation indicates that it is possible to reduce problem scale without affecting the solution by excluding the unused links from problem formulation. In order to identify the link usage before solving the problem, we exploit deep learning to find the latent relationship between flow information and link usage in optimal solution. Based on this, we further predict whether a link will be scheduled through link evaluation and eliminate unused link from formulation to reduce problem size. Numerical results demonstrate that the proposed method can reduce computation cost by at least 50% without affecting optimality, thus greatly improve the efficiency of solving large scale network optimization problems.

To deal with the huge traffic demand in wireless networks, small base stations (SBSs) are introduced in cellular networks, not only with dense deployment, but also with caching capabilities. Due to the short distances between users and SBSs, files can be downloaded at higher transmission rate from the cache of SBSs than being downloaded over backhaul links. However, user mobility makes the file allocation in SBS caches more challenging, mostly because the user association to SBSs dynamically changes. Moreover, the amount of data that users can download depends on the sojourn time within the coverage of SBSs. In this paper, based on the assumption that the user sojourn time obeys exponential distribution, we get the upper bound of the mean download time of files. A file allocation strategy is then proposed based on the derived bounds of mean sojourn time. Optimal file allocation strategies under low and high mobility are further obtained. Simulations show that the strategy performs notably better than the popularity-based allocation strategy.

Merging mobile edge computing with the dense deployment of small cell base stations promises enormous benefits such as a real proximity, ultra-low latency access to cloud functionalities. However, the envisioned integration creates many new challenges and one of the most significant is mobility management, which is becoming a key bottleneck to the overall system performance. Simply applying existing solutions leads to poor performance due to the highly overlapped coverage areas of multiple base stations in the proximity of the user and the co-provisioning of radio access and computing services. In this paper, we develop a novel user-centric mobility management scheme, leveraging Lyapunov optimization and multi-armed bandits theories, in order to maximize the edge computation performance for the user while keeping the user's communication energy consumption below a constraint. The proposed scheme effectively handles the uncertainties present at multiple levels in the system and provides both short-term and long-term performance guarantee. Simulation results show that our proposed scheme can significantly improve the computation performance (compared to state of the art) while satisfying the communication energy constraint.

Mobile edge computing is a novel technique in which mobile devices offload computation-intensive tasks with stringent delay requirements to the edge cloud. However, the limited computational resource in the edge cloud may result in the Quality of Service degradation. In this paper, we address this issue by coordinating the heterogeneous cloud which includes the edge cloud and the remote cloud. Considering the offloading of delay-bounded tasks, we study into the scheduling of heterogeneous cloud in order to maximize the probability that tasks can have the delay requirements met. The problem formulation is proved to be concave, and an optimal algorithm is proposed accordingly. The optimal policy with heterogeneous cloud is notably different from the policy merely using the edge cloud. With only the edge cloud, the system serves tasks with loose delay bounds and drops tasks with stringent delay bounds when the traffic load is heavy. However, with the heterogeneous cloud, tasks with stringent delay bounds are offloaded to the edge cloud and tasks with loose delay bounds are offloaded to the remote cloud. In numerical results, the probability that the delay bounds of tasks are satisfied can be improved by about 40% with the assistance of the remote cloud.

Energy harvesting is an emerging technology providing clean energy for wireless communication systems. Due to the randomness in energy arrivals, wireless service process needs to be matched with energy provision to avoid energy waste or shortage. Other than passively adjusting energy usage according to traffic and energy profiles, a framework, namely, GreenDelivery has been proposed to proactively push popular contents to users in advance, such that harvested energy can be utilized more efficiently. In this paper, a heterogeneous network with multiple GreenDelivery small cells is considered. Due to spatial proximity, adjacent small cells might conflict with each other due to simultaneous transmissions or repeated pushes of identical contents to the same user, which calls for a more sophisticated design of push scheme in a multi-cell scenario. To tackle the interference, small base station (SBS) scheduling is proposed, exploiting the intermittent nature of renewable energy, to temporally separate the transmission of adjacent small cells. Heuristic push schemes are then proposed to further reduce the user requests handled by macro base stations (MBS) with centralized and distributed realizations. Simulations show that the proposed push schemes outperform the baseline scheme in which contents are pushed in the descending order of their popularities, especially when content popularity is more uniformly distributed.

Merging mobile edge computing (MEC) functionality with the dense deployment of base stations (BSs) provides enormous benefits such as a real proximity, low latency access to computing resources. However, the envisioned integration creates many new challenges, among which mobility management (MM) is a critical one. Simply applying existing radio access-oriented MM schemes leads to poor performance mainly due to the co-provisioning of radio access and computing services of the MEC-enabled BSs. In this paper, we develop a novel user-centric energy-aware mobility management (EMM) scheme, in order to optimize the delay due to both radio access and computation, under the long-term energy consumption constraint of the user. Based on Lyapunov optimization and multi-armed bandit theories, EMM works in an online fashion without future system state information, and effectively handles the imperfect system state information. Theoretical analysis explicitly takes radio handover and computation migration cost into consideration and proves a bounded deviation on both the delay performance and energy consumption compared with the oracle solution with exact and complete future system information. The proposed algorithm also effectively handles the scenario in which candidate BSs randomly switch ON/OFF during the offloading process of a task. Simulations show that the proposed algorithms can achieve close-to-optimal delay performance while satisfying the user energy consumption constraint.

Motivated by the rapid development of energy harvesting technology and content-aware communication in access networks, this paper considers the push mechanism design in small-cell base stations (SBSs) powered by renewable energy. A user request can be satisfied by either push or unicast from the SBS. If the SBS cannot handle the request, the user is blocked by the SBS and is served by the macro-cell BS instead, which typically consumes more energy. We aim to minimize the ratio of user requests blocked by the SBS to total number of user requests. With finite battery capacity, Markov decision process-based problem is formulated, and the optimal policy is found by dynamic programming (DP). Two threshold-based policies are proposed: the push-only threshold-based policy and the energy-efficient threshold-based policy, and the closed-form blocking probabilities with infinite battery capacity are derived. Numerical results show that the proposed policies outperform the conventional non-push policy if the content popularity changes slowly or the content request generating rate is high, and can achieve the performance of the greedy optimal threshold-based policy. In addition, the performance gap between the threshold-based policies and the DP optimal policy is small when the energy arrival rate is low or the request generating rate is high.

In this paper, we explore the content information to design the joint caching and push mechanism in the small-cell base stations (SBSs) powered by renewable energy. The problem is formulated as a Markov decision process by exploring the features of content popularity and renewal and by taking into consideration the energy consumption for both content fetch from core network and push to the users. The objective is to minimize the number of requests which cannot be met by the SBSs. We adopt the policy iteration algorithm to obtain the optimal caching and push policy. According to the numerical results, the performance gain with large SBS cache size is marginal due to the limited energy. We also find that the optimal policy reveals noticeable performance gain compared with the greedy fetch policy and the non-push policy. In addition, simulations shows the tradeoff between the number of cached contents in the SBS and the available energy for content push.

This paper considers a single-transmitter energy harvesting system and looks into the problem of completion time minimization from the worst-case point of view. In offline study, an optimal algorithm [1] is given to yield the minimum completion time of transmission. However, for online algorithms, the randomness of future energy arrivals adds to the difficulty of scheduling, thus the offline minimum completion time cannot always be reached. This leads to the question "What is the deterministic performance bound of online algorithms compared to the offline optimum". By a game-theoretic method, this paper shows that with an infinite-sized battery, there exist several algorithms that guarantee a completion time no more than the twice of its offline counterpart for all possible energy arrivals, and more importantly that the ratio of two cannot be further reduced. This property is of great significance especially when reliability is valued in the system.

Base station (BS) sleeping is an effective way to improve the energy-efficiency of cellular networks. However, it may bring extra user-perceived delay. We conduct a theoretical study into the impact of BS sleeping on both energy-efficiency and user-perceived delay. We consider hysteresis sleep and three typical wake-up schemes, namely single sleep, multiple sleep, and N-limited schemes. We model the system as an M/G/1 vacation queue, which captures the setup time, the mode-changing cost, as well as the counting or detection cost during the sleep mode. Closed-form expressions for the average power and the Laplace-Stieltjes transform of delay distribution are obtained. The impacts of system parameters on these expressions are analyzed. We then formulate an optimization problem to design delay-constrained energy-optimal BS sleeping policies. We show that the optimal solutions possess a special structure, thereby allowing us to obtain them explicitly or numerically by simple bisection search. In addition, the relationship between the optimal power consumption and the mean delay constraint is analyzed, so as to answer the fundamental question: how much energy can be saved by trading off a certain amount of delay? It is shown that this optimal relationship is linear only when the delay constraint is lower than a threshold. Numerical studies are also conducted, where the impact of detection or counting cost during the sleep mode is explored, and the delay distribution under the optimal policy is obtained.

With small cell base stations (SBSs) densely deployed in addition to conventional macro base stations (MBSs), the heterogeneous cellular network (HCN) architecture can effectively boost network capacity. To support the huge power demand of HCNs, renewable energy harvesting technologies can be leveraged. In this paper, we aim to make efficient use of the harvested energy for on-grid power saving while satisfying the quality of service (QoS) requirement. To this end, energy-aware traffic offloading schemes are proposed, whereby user associations, ON-OFF states of SBSs, and power control are jointly optimized according to the statistical information of energy arrival and traffic load. Specifically, for the single SBS case, the power saving gain achieved by activating the SBS is derived in closed form, based on which the SBS activation condition and optimal traffic offloading amount are obtained. Furthermore, a two-stage energy-aware traffic offloading (TEATO) scheme is proposed for the multiple-SBS case, considering various operating characteristics of SBSs with different power sources. Simulation results demonstrate that the proposed scheme can achieve more than 50% power saving gain for typical daily traffic and solar energy profiles, compared with the conventional traffic offloading schemes.

This paper considers two base stations (BSs) powered by renewable energy serving two users cooperatively. With different BS energy arrival rates, a fractional joint transmission (JT) strategy is proposed, which divides each transmission frame into two subframes. In the first subframe, one BS keeps silent to store energy, while the other transmits data, and then, they perform zero-forcing JT (ZF-JT) in the second subframe. We consider the average sum-rate maximization problem by optimizing the energy allocation and the time fraction of ZF-JT separately. First, the sum-rate maximization for given energy budgets in each frame is analyzed. We prove that the optimal transmit power can be derived in closed form, and the optimal time fraction can be found via bi-section search. Second, an approximate dynamic programming algorithm is introduced to determine the energy allocation among frames. We adopt a linear approximation with the features associated with system states and determine the weights of features by simulation. We also operate the approximation several times with random initial policy, named policy exploration, to broaden the policy search range. Numerical results show that the proposed fractional JT greatly improves the performance. In addition, appropriate policy exploration is shown to perform close to the optimal.

Cloud radio access network (C-RAN) was proposed recently to reduce network cost, enable cooperative communications, and increase system flexibility through centralized baseband processing. By pooling multiple virtual base stations (VBSs) and consolidating their stochastic computational tasks, the overall computational resource can be reduced, achieving the so-called statistical multiplexing gain. In this paper, we evaluate the statistical multiplexing gain of VBS pools using a multi-dimensional Markov model, which captures the session-level dynamics and the constraints imposed by both radio and computational resources. Based on this model, we derive a recursive formula for the blocking probability and also a closed-form approximation for it in large pools. These formulas are then used to derive the session-level statistical multiplexing gain of both real-time and delay-tolerant traffic. Numerical results show that VBS pools can achieve more than 75% of the maximum pooling gain with 50 VBSs, but further convergence to the upper bound (large-pool limit) is slow because of the quickly diminishing marginal pooling gain, which is inversely proportional to a factor between the one-half and three-fourth power of the pool size. We also find that the pooling gain is more evident under light traffic load and stringent quality of service requirement.

User behaviour analysis based on traffic log in wireless networks can be beneficial to many fields in real life: not only for commercial purposes, but also for improving network service quality and social management. We cluster users into groups marked by the most frequently visited websites to find their preferences. In this paper, we propose a user behaviour model based on Topic Model from document classification problems. We use the logarithmic TF-IDF (term frequency - inverse document frequency) weighing to form a high-dimensional sparse feature matrix. Then we apply LSA (Latent semantic analysis) to deduce the latent topic distribution and generate a low-dimensional dense feature matrix. K-means++, which is a classic clustering algorithm, is then applied to the dense feature matrix and several interpretable user clusters are found. Moreover, by combining the clustering results with additional demographical information, including age, gender, and financial information, we are able to uncover more realistic implications from the clustering results.

To provide the paradigm shift of green cellular communications, Hyper Cellular Architecture (HCA), has been proposed, in which the common control functionalities are decoupled from the data service functionalities at base station (BS) level so that the traffic BSs can be more adaptive to the temporal and spatial traffic fluctuations. In this paper, we develop a system level simulator (SLS) for HCA to evaluate the HCA performance under temporal and spatial traffic fluctuations. The SLS enjoys low complexity, open interface and completed functions through the carefully tuned modeling on long-term large-scale traffic model, the separation architecture and the resource allocation strategies. Simulation results show that even with some basic BS sleeping algorithms, HCA can achieve up to 45% energy efficiency (EE) gain over conventional cellular architecture with macro BSs only or heterogeneous network during the low traffic period, and about 36% EE gain on average for a typical daily traffic pattern.

Wireless communication networks rely heavily on channel state information (CSI) to make informed decision for signal processing and network operations. However, the traditional CSI acquisition methods is facing many difficulties: pilot-aided channel training consumes a great deal of channel resources and reduces the opportunities for energy saving, while location-aided channel estimation suffers from inaccurate and insufficient location information. In this paper, we propose a novel channel learning framework, which can tackle these difficulties by inferring unobservable CSI from the observable one. We formulate this framework theoretically and illustrate a special case in which the learnability of the unobservable CSI can be guaranteed. Possible applications of channel learning are then described, including cell selection in multi- tier networks, device discovery for device-to-device (D2D) communications, as well as end-to-end user association for load balancing. We also propose a neuron-network-based algorithm for the cell selection problem in multi-tier networks. The performance of this algorithm is evaluated using geometry-based stochastic channel model (GSCM). In settings with 5 small cells, the average cell-selection accuracy is 73% - only an 3.9% loss compared with a location-aided algorithm which requires genuine location information.

Energy harvesting (EH), which explores renewable energy as a supplementary power source, is a promising 5G technology to support the huge energy demand of heterogeneous cellular networks (HCN). However, the random arrival of renewable energy brings great challenges to network management. By adjusting the distribution of traffic load in spatial domain, traffic shaping helps to balance the cell-level power demand and supply, and thus improves the utilization of renewable energy. In this paper, we investigate the power saving performance of traffic shaping in an analytical way, based on the statistic information of energy arrival and traffic load. Specifically, an energy-optimal traffic shaping scheme (EOTS) is devised for HCNs with EH, whereby the on-off state of the off-grid small cell and the amount of offloading traffic are adjusted dynamically with the energy variation, to minimize the on-grid power consumption. Numerical results are given to demonstrate that for the daily traffic and solar energy profiles, EOTS scheme can significantly reduce the energy consumption, compared with the greedy method where users are always offloaded to the off-grid small cell with priority.

In cellular networks, the coverage of base stations (BSs) belonging to different operators often overlaps. As a result, when the traffic load is lower than the peak volume, there are opportunities to turn off a subset of BSs to save power, potentially from different operators, leaving their users to be served by other BSs. Because in this case the active BSs can be shared among different operators, it is rational to assume that operators are self-interested and hold their own private information, such as their own traffic loads. In this paper, we consider the problem of how to motivate operators to cooperate and reveal their private information such that the overall utility can be maximized, which is also called social efficiency in mechanism design. A new BS utility model that depends on the BS's energy consumption is proposed. Based on this, a game theoretic mechanism with money transfer between operators is designed, which has been proved to be incentive compatible and budget-balanced. Simulation results under various traffic load distributions show that when the operators have similar traffic load distributions, they would like to participate the cooperation voluntarily.

Motivated by the recent development of energy harvesting communications, and the trend of multimedia contents caching and push at the access edge and user terminals, this paper considers how to design an effective push mechanism of energy harvesting powered small-cell base stations (SBSs) in heterogeneous networks. The problem is formulated as a Markov decision process by optimizing the push policy based on the battery energy, user request and content popularity state to maximize the service capability of SBSs. We extensively analyze the problem and propose an effective policy iteration algorithm to find the optimal policy. According to the numerical results, we find that the optimal policy reveals a state dependent threshold based structure. Besides, more than 50% performance gain is achieved by the optimal push policy compared with the non-push policy.

The explosive growth of smart mobile user equipments (UEs) boosts the emerging of numerous mobile applications. Most of these applications require an always-online connectivity, which incurs overly-frequent Radio Resource Control (RRC) state transitions, leading to signaling storm and user access failure. To address this issue, many researches focus on avoiding frequent transitions between RRC states by maintaining UEs in the RRC connected state for longer time. However, these researches bring up substantial energy consumption. In this paper, we propose an energy-efficient system signaling control method, by which each UE adjusts its RRC release timer adaptively according to the traffic patterns of mobile applications. Numerical results show that in comparison to the conventional signaling control method, the proposed method can save 27.5% average energy consumption with well-controlled signaling load. Meanwhile, the disparity of user experience is significantly lower.

The baseband-up centralization architecture of radio access networks (C-RAN) has recently been proposed to support efficient cooperative communications and reduce deployment and operational costs. However, the massive fronthaul bandwidth required to aggregate baseband samples from remote radio heads (RRHs) to the central office incurs huge fronthauling cost, and existing baseband compression algorithms can hardly solve this issue. In this paper, we propose a graph-based framework to effectively reduce fronthauling cost through properly splitting and placing baseband processing functions in the network. Baseband transceiver structures are represented with directed graphs, in which nodes correspond to baseband functions, and edges to the information flows between functions. By mapping graph weighs to computational and fronthauling costs, we transform the problem of finding the optimum location to place some baseband functions into the problem of finding the optimum clustering scheme for graph nodes. We then solve this problem using a genetic algorithm with customized fitness function and mutation module. Simulation results show that proper splitting and placement schemes can significantly reduce fronthauling cost at the expense of increased computational cost. We also find that cooperative processing structures and stringent delay requirements will increase the possibility of centralized placement.

The introduction of micro grids (MGs) into the utility grid poses new challenges in the energy management design due to the intermittent characteristics of renewable energy sources and limited storage capacity. In this paper, we proposed a distributed energy management algorithm by taking into consideration the interactions and interconnections among utility companies, MGs, and customers. We model the energy management problem as a two-stage Stackel berg game, in which utility companies and MGs are game leaders, and customers are game followers. Utility companies and MGs make decisions about what price to offer their electricity to customers. Customers adjust their electricity procurement amounts based on the prices offered by utility companies and MGs. We prove that a Nash equilibrium exists in the proposed two-stage Stackel berg game, and the optimum solutions obtained by the distributed energy management algorithm is exactly the Nash equilibrium. We have analyzed and verified the relationships among utility functions, electricity prices, electricity demands, electricity procurement amounts, and pollutant parameters through computer simulations. We have also compared the performance of the proposed distributed algorithm with the centralized algorithm under different simulation conditions.

Dimensionality loss is defined as the channel estimation overhead, which results in a loss of time-frequency resources in pilot-assisted wireless systems. In this paper, the scaling result of dimensionality loss, i.e., the scaling factor, in frequency-division-duplex (FDD) massive multiple-input-multiple-output(MIMO) downlinks is derived. The scaling factor determines the amount of channel estimation overhead, and thus is vital to understand the downlink throughput in FDD massive MIMO systems. Moreover, the transmit diversity of the downlink channel is also derived. In the simulations, we adopt a geometry-based stochastic channel model to validate our analysis. The impact of several assumptions made in our analysis is also investigated.

User scheduling in multiuser multiple-input-multiple-output (MU-MIMO) systems is fundamentally different with single-user systems1, in the sense that without spatial multiplexing, users in single-user systems are sharing the time-frequency degree-of-freedoms (DoFs), whereas in MU-MIMO systems, due to the fact that the number of spatial DoFs scales with the number of users (assuming sufficient base station (BS) antennas), users are not sharing the DoFs, but rather creating additional DoFs for their own use. However, instead of limited by the available DoFs, the number of simultaneous users are limited by the channel state information (CSI) acquisition overhead in pilot-assisted MU-MIMO systems. In this paper, we investigate the user scheduling scheme in pilot-assisted time-division-duplex (TDD) MU-MIMO systems. Leveraging the Lyapunov optimization techniques, we derive the throughput-optimal scheduling policy which serves as a performance bound due to its non-causality and high complexity. We then propose a heuristic scheme, which is causal and substantially decreases the complexity. Moreover, it performs fairly close to the optimum.

Multiuser Multiple input multiple output (MIMO) systems are now having more and more radio frequency (RF) chains, with larger capacity and at the same time higher energy consumption. With random data arrival, it is desired to turn off RF chains to save energy according to the traffic variations. In this paper a low-complexity traffic-aware scheme is proposed, whereby RF chains and users are selected at each frame based on the channel quality and the data queue-length. Particularly, the number of active RF chains is determined by comparing the current queue-length to the predefined thresholds, the values of which are able to control the tradeoff between the energy saving and quality of service, i.e., delay. Simulation results show that the proposed scheme saves more energy compared with conventional schemes which is designed regardless the traffic variations, and the saving gain increases when the average traffic load decreases.

Base station (BS) sleeping operation is one of the effective ways to save energy consumption of cellular networks, but it may lead to longer delay to the customers. The fundamental question then arises: How much energy can be traded off by a tolerable delay? In this paper, we characterize the fundamental tradeoffs between total energy consumption and overall delay in a BS with sleep mode operations by queueing models. Here, the BS total energy consumption includes not only the transmitting power but also basic power (for baseband processing, power amplifier, etc.) and switch-over power of the BS working mode, and the overall delay includes not only transmission delay but also queueing delay. Specifically, the BS is modeled as an M/G/1 vacation queue with setup and close-down times, where the BS enters sleep mode if no customers arrive during the close-down (hysteretic) time after the queue becomes empty. When asleep, the BS stays in sleep mode until the queue builds up to N customers during the sleep period ( N-Policy) . Several closed-form formulas are derived to demonstrate the tradeoffs between the energy consumption and the mean delay for different wake-up policies by changing the close-down time, setup time, and the parameter N. It is shown that the relationship between the energy consumption and the mean delay is linear in terms of mean close-down time, but non-linear in terms of N. The explicit relationship between total power consumption and average delay with varying service rate is also analyzed theoretically, indicating that sacrificing delay cannot always be traded off for energy saving. In other words, larger N may lead to lower energy consumption, but there exists an optimal N* that minimizes the mean delay and energy consumption at the same time. We also investigate the maximum delay (delay bound) for certain percentage of service and find that the delay bound is nearly linear in mean delay in the cases tested. Therefore, similar tradeoffs exist between energy consumption and the delay bound. In summary, the closed-form energy-delay tradeoffs cast light on designing BS sleeping and wake-up control policies that aim to save energy while maintaining acceptable quality of service.

This paper studies online power control policies for outage minimization in a fading wireless link with energy harvesting transmitter and receiver. The outage occurs when either the transmitter or the receiver does not have enough energy, or the channel is in outage, where the transmitter only has the channel distribution information. Under infinite battery capacity and without retransmission, we prove that threshold-based power control policies are optimal. We thus propose disjoint/joint threshold-based policies with and without battery state sharing between the transmitter and receiver, respectively. We also analyze the impact of practical receiver detection and processing on the outage performance. When retransmission is considered, policy with linear power levels is adopted to adapt the power thresholds per retransmission. With finite battery capacity, a three dimensional finite state Markov chain is formulated to calculate the optimal parameters and corresponding performance of proposed policies. The energy arrival correlation between the transmitter and receiver is addressed for both finite and infinite battery cases. Numerical results show the impact of battery capacity, energy arrival correlation and detection cost on the outage performance of the proposed policies, as well as the tradeoff between the outage probability and the average transmission times.

To further improve the energy efficiency of heterogeneous networks, a separation architecture called hyper-cellular network (HCN) has been proposed, which decouples the control signaling and data transmission functions. Specifically, the control coverage is guaranteed by macro base stations (MBSs), whereas small cells (SCs) are only utilized for data transmission. Under HCN, SCs can be dynamically turned off when traffic load decreases for energy saving. A fundamental problem then arises: how many SCs can be turned off as traffic varies? In this paper, we address this problem in a theoretical way, where two sleeping schemes (i.e., random and repulsive schemes) with vertical inter-layer offloading are considered. Analytical results indicate the following facts: 1) under the random scheme where SCs are turned off with certain probability, the expected ratio of sleeping SCs is inversely proportional to the traffic load of SC-layer and decreases linearly with the traffic load of MBS-layer; 2) the repulsive scheme, which only turns off the SCs close to MBSs, is less sensitive to the traffic variations; and 3) deploying denser MBSs enables turning off more SCs, which may help to improve network energy-efficiency. Numerical results show that about 50% SCs can be turned off on average under the predefined daily traffic profiles, and 10% more SCs can be further turned off with inter-layer channel borrowing.

In this paper, we study the optimal broadcast radius in cellular network from the perspective of energy efficiency. Firstly, we consider the scenario where the base stations have already been deployed. The relationship between the power-saving gain and the possible energy efficient broadcast radius is derived. Based on it, we investigate the optimal number of tiers of cells in the broadcast range. Next, we turn to the scenario with the base stations to be deployed. The power consumption per effective coverage area is defined to measure the energy efficiency of different broadcast radius, based on which we analyze the optimal broadcast radius under different channel conditions. Simulation results confirm that the optimal energy efficient broadcast radius is positively correlated with the static power of the broadcast channel, negatively correlated with the minimal required broadcast service rate in the two scenarios, and negatively correlated with the original cell radius in the first scenario.

Facing the explosion of mobile data traffic, cloud radio access network (C-RAN) is proposed recently to overcome the efficiency and flexibility problems with the traditional RAN architecture by centralizing baseband processing. However, there lacks a mathematical model to analyze the statistical multiplexing gain from the pooling of virtual base stations (VBSs) so that the expenditure on fronthaul networks can be justified. In this paper, we address this problem by capturing the session-level dynamics of VBS pools with a multi-dimensional Markov model. This model reflects the constraints imposed by both radio resources and computational resources. To evaluate the pooling gain, we derive a product-form solution for the stationary distribution and give a recursive method to calculate the blocking probabilities. For comparison, we also derive the limit of resource utilization ratio as the pool size approaches infinity. Numerical results show that VBS pools can obtain considerable pooling gain readily at medium size, but the convergence to large pool limit is slow because of the quickly diminishing marginal pooling gain. We also find that parameters such as traffic load and desired Quality of Service (QoS) have significant influence on the performance of VBS pools.

To further improve energy efficiency from the view of the whole network, a separation architecture has been proposed, where the control plane and data plane are separated and implemented by different base stations. Under this architecture, the data base stations (DBS) can be turned off adaptively according to the traffic load while signaling base stations (SBS) provide the guarantee of coverage. A key issue of this architecture is the design of effective BS sleeping mechanisms, which should guarantee the quality of service (QoS) and minimize network power consumption. In this paper, a probabilistic DBS sleeping mechanism is proposed and optimized under the separation architecture. Users within the sleeping DBSs are offloaded to SBSs for QoS guarantee. An optimization problem is formulated, where the sleeping probability and spectrum resource allocation are jointly optimized to minimize network power consumption. The optimal BS sleeping scheme is found to be threshold-based. When the ratio of sleeping DBSs is below a certain threshold which depends on the traffic load, the lightly-loaded DBSs should be turned off first; otherwise, only the heavily loaded DBSs go into sleep. Numerical results show nearly 30% energy can be saved under a typical daily traffic profile, and there exists a tradeoff between energy saving and network capacity.

The combination of energy harvesting and large-scale multiple antenna technologies provides a promising solution for improving the energy efficiency (EE) by exploiting renewable energy sources and reducing the transmission power per user and per antenna. However, the introduction of energy harvesting capabilities into large-scale multiple antenna systems poses many new challenges for energy-efficient system design due to the intermittent characteristics of renewable energy sources and limited battery capacity. Furthermore, the total manufacture cost and the sum power of a large number of radio frequency (RF) chains can not be ignored, and it would be impractical to use all the antennas for transmission. In this paper, we propose an energy-efficient antenna selection and power allocation algorithm to maximize the EE subject to the constraint of user's quality of service (QoS). An iterative offline optimization algorithm is proposed to solve the non-convex EE optimization problem by exploiting the properties of nonlinear fractional programming. The relationships among maximum EE, selected antenna number, battery capacity, and EE-SE tradeoff are analyzed and verified through computer simulations.

In this paper, we study how to use the solar radiation model to predict energy arrivals and to allocate energy resource at an energy harvesting base station (BS). First, some primary knowledge about solar radiation is reviewed and summarized. We present two solar energy models for cloudless days and cloudy days, respectively. Then artificial neural network (ANN) is used to predict solar energy arrivals in a short period, which has an improved performance compared with the previous linear model. In the end, the allocation of received energy is considered, and one optimal offline algorithm and four heuristics online algorithms are proposed. We evaluate the performance of the algorithms using Denver's solar radiation data in recent 27 years from National Renewable Energy Laboratory (NERL). Simulation results show our prediction and optimization algorithm achieves nearly optimal performance.

The increasing traffic demands bring heavy load to both the data and control planes of cellular networks, along with substantial energy consumption. To solve the issue, new network architecture that separates signaling and data has been proposed in literature for future green cellular networks. In this paper, we analyze the data and signaling resource configuration problem in this new network architecture. We find the optimal resource partitioning parameters to optimize the blocking performance and to minimize the overall network power consumption with a blocking probability constraint. More specifically, we adopt traffic-aware resource allocation between the data and signaling base stations (BSs) to improve network access capability while reducing the overall network power consumption. Two types of resource partitioning patterns, complete partitioning and partial partitioning, are studied. Numerical results show that great energy-saving gain can be achieved compared with the traditional fixed and traffic-proportional resource partitioning patterns. Moreover, power consumption and blocking performance tradeoffs are explored, based on which the appropriate resource partitioning pattern can be chosen according to different quality of service (QoS) requirements.

This paper addresses the outage minimization problem for a wireless link where both the transmitter and the receiver are powered by harvested energy, and the energy arrival processes of both nodes are correlated. We propose three power control policies to minimize the outage probability, including threshold-based On-Off policy, joint scheduling policy, and linear power levels policy. With infinite battery capacity, we analyze the optimality of the thresholds with different correlations between energy arrivals at the transmitter and the receiver. With finite battery capacity, we use finite state Markov chain (FSMC) to obtain the optimality of our policies and also numerically evaluate their performance. The optimal thresholds for minimum outages are derived according to the average energy arrival rate and the system parameters. The numerical results show the performance gains using different policies, as well as the tradeoff between the minimum outage probabilities and the average transmission times.

We consider energy-efficient wireless resource management in cellular networks where base stations (BSs) are equipped with energy harvesting devices, using statistical information for traffic intensity and renewable energy. The problem is formulated as adapting BSs' on-off states, active resource blocks (e.g., subcarriers), and renewable energy allocation to minimize the average grid power consumption while satisfying the users' quality of service (blocking probability) requirements. It is transformed into an unconstrained optimization problem to minimize a weighted sum of grid power consumption and blocking probability. A two-stage dynamic programming algorithm is proposed to solve this problem, by which the BSs' on-off states are optimized in the first stage, and the active BSs' resource blocks are allocated iteratively in the second stage. Compared with the optimal joint BSs' on-off states and active resource blocks allocation algorithm, the proposed algorithm greatly reduces the computational complexity and can achieve the optimal performance when the traffic is uniformly distributed.

Multiuser multiple-input-multiple-output (MU-MIMO) systems are known to be hindered by dimensionality loss due to channel state information (CSI) acquisition overhead. In this paper, we investigate user-scheduling in MU-MIMO systems on account of CSI acquisition overhead, where a base station dynamically acquires user channels to avoid choking the system with CSI overhead. The genie-aided optimization problem (GAP) is first formulated to maximize the Lyapunov-drift every scheduling step, incorporating user queue information and taking channel fluctuations into consideration. The scheduling scheme based on GAP, namely the GAP-rule, is proved to be throughput-optimal but practically infeasible, and thus serves as a performance bound. In view of the implementation overhead and delay unfairness of the GAP-rule, the T-frame dynamic channel acquisition scheme and the power-law DCA scheme are further proposed to mitigate the implementation overhead and delay unfairness, respectively. Both schemes are based on the GAP-rule and proved throughput-optimal. To make the schemes practically feasible, we then propose the heuristic schemes, queue-based quantized-block-length user scheduling scheme (QQS), T-frame QQS, and power-law QQS, which are the practical versions of the aforementioned GAP-based schemes, respectively. The QQS-based schemes substantially decrease the complexity, and also perform fairly close to the optimum. Numerical results evaluate the proposed schemes under various system parameters.

In this paper, we consider the problem of energy and data control in energy cooperation networks powered by renewable energy. In such networks, nodes can provide data transmission service, and at the same time they have power lines to transfer the harvested energy to others. We develop an online algorithm called Energy and Data Aware (EDA) algorithm using Lyapunov analysis, which makes data admission control and decides energy allocation for traffic transmission and energy transfer. In our EDA algorithm, the node only needs to keep track of its own energy storage status and does not require any knowledge of the energy harvesting process. We show that the proposed algorithm achieves a utility that is within O(ε) of the optimal, for any ε > 0, while ensuring that both the network data queue length and the capacity of energy storage devices are upper bounded by bounds of size O(1/ε).

In this paper we consider the minimum base station (BS) power consumption given the sum rate requirement in large-scale multiple-input-multiple-output (MIMO) systems. A single cell with an Mtot-antenna BS and N single-antenna users is considered. The BS power consumption consists of two parts: The part accounting for the total transmit power and the part proportional to the number of active antennas. Specifically, closed-form approximations (CFAs) of the optimal transmit power and optimal number of active antennas are derived when the sum rate requirement is high. A CFA of the ergodic sum capacity upper bound for the downlink broadcast channel is also given.

The existing architecture of heterogeneous networks is not energy and spectrum efficient as many lightly loaded base stations (BS) can not be turned off for coverage guarantee. To further improve energy and spectrum efficiency, a new architecture called “hyper-cellular” network has been proposed in our previous work. Under this architecture, the function of different types of BSs may not be the same, and the mechanism of user association should consider many factors, such as user mobility, traffic load distribution, and differentiated service demands, which are usually ignored in the existing studies. In addition, the spectrum allocation strategy also has great influence on the network performance. As a starting point, we explore the user association mechanism based on the differentiated service demands of the network users, and jointly optimize it with spectrum allocation, in order to maximize the network capacity with quality of service constraints. Although closed-form expression of the optimal solution can not be derived, numerical results are obtained. Our approach is shown to improve the network capacity more than four times over the baseline strategy, where the conventional user association method is adopted and all BSs use all available spectrum.

This paper studies the resource allocation problem of a single cell powered jointly by renewable energy and power grid over a given time period (e.g. 24 hours), using statistical information of traffic intensity and harvested energy. Specifically, the problem is formulated as minimizing the average grid power input while satisfying users' quality of service (outage probability) requirements. We analyze the outage probability, and solve the grid power minimization problem indirectly by obtaining a power-outage tradeoff curve using the dynamic programming (DP) approach. Some heuristic algorithms are proposed and compared with the DP algorithm by simulations. The results show that the DP algorithm greatly reduces the grid power consumption compared with the heuristic methods, among which the joint traffic-energy-aware resource allocation performs closest to the optimal solution.

This paper explores the influence of the inhomogeneous spatial traffic distribution on the performance of energy efficiency and proposes an energy-efficient heterogeneous network planning scheme for cellular networks. We first provide a spatial modeling method of Scalable, Spatially correlated, and Log-normally distributed Traffic (SSLT). Together with the spatial modeling, the measure of inhomogeneity of spatial traffic distributions is introduced which is related to the energy efficiency performance of network planning. Then we propose a network planning scheme which deploys both micro and macro base stations (BSes) adapting to the traffic inhomogeneity. Numerical results show that when the spatial traffic distribution is more inhomogeneous, the proposed scheme improves energy efficiency by deploying more micro BSes. We also found that proportional relationships exist among three values, i.e. the ratio of the number of micro and macro BSes, energy efficiency, and the inhomogeneity.

With increased radio frequency (RF) chains, base station (BS) with multiple antennas consumes more circuit power. Turning off RF chains will help to save energy. However, in turn, it needs more sophisticated user scheduling. Therefore, an energy-efficient scheduling scheme is proposed with which users and RF chains are jointly selected at each frame. Here, Lyapunov driftplus-penalty ratio is used to policy design. If the average data arrival rates locate in the capacity region, it is proved that the proposed policy achieves the maximum energy efficiency than any other stationary, randomized, queue-independent policies, while ensuring the stability of the system. At each frame, the selection of users and RF chains depends on the number of selected users, sum queue length of them and energy efficiency they achieve. A key observation is that the numbers of selected users and RF chains should be equal under zero-forcing beamforming. Simulation results have shown that it even achieves higher energy efficiency than the Maximum Weighted Queue scheduling scheme when average arrival rate vector is relative small.

This paper considers the power allocation of a single-link wireless communication with joint energy harvesting and grid power supply. We formulate the problem as minimizing the grid power consumption with random energy and data arrival in fading channel, and analyze the structure of the optimal power allocation policy in some special cases. For the case that all the packets are arrived before transmission, it is a dual problem of throughput maximization, and the optimal solution is found by the two-stage water filling (WF) policy, which allocates the harvested energy in the first stage, and then allocates the power grid energy in the second stage. For the random data arrival case, we first assume grid energy or harvested energy supply only, and then combine the results to obtain the optimal structure of the coexisting system. Specifically, the reverse multi-stage WF policy is proposed to achieve the optimal power allocation when the battery capacity is infinite. Finally, some heuristic online schemes are proposed, of which the performance is evaluated by numerical simulations.

Partial Spectrum Reuse (PSR) in the second tier of two-tier heterogeneous cellular networks has a potential to improve spectrum efficiency by reducing inter-cell interference, and thus energy efficiency as well by deploying less or switching off more Base Stations (BSs). In this paper, we analyze the optimal PSR factor, defined as the portion of spectrum reused by micro cells in two-tier heterogeneous networks, which is not in an explicit form generally. Then, a closed-form limit of the optimal PSR factor is derived as the ratio of the user rate requirement over the whole system spectrum bandwidth is approaching zero, based on which a threshold of the micro-BS energy cost is also derived to determine which type of BSs is preferable. Specifically, one should deploy more micro BSs or switch off more macro BSs if the micro-BS energy cost is lower than the threshold. Otherwise, the optimal choice is the opposite. This threshold with the PSR scheme is higher than that without PSR scheme, i.e., PSR can improve both spectrum efficiency and energy efficiency. Numerical results show that adopting PSR can reduce the network energy consumption by up to 50% when the transmit power of macro BSs is 10dB higher than that of micro BSs.

In this paper, the optimal BS (Base Station) density for both homogeneous and heterogeneous cellular networks to minimize network energy cost is analyzed with stochastic geometry theory. For homogeneous cellular networks, both upper and lower bounds of the optimal BS density are derived. For heterogeneous cellular networks, our analysis reveals the best type of BSs to be deployed for capacity extension, or to be switched off for energy saving. Specifically, if the ratio between the micro BS cost and the macro BS cost is lower than a threshold, which is a function of path loss and their transmit power, then the optimal strategy is to deploy micro BSs for capacity extension or to switch off macro BSs (if possible) for energy saving with higher priority. Otherwise, the optimal strategy is the opposite. The optimal combination of macro and micro BS densities can be calculated numerically through our analysis, or alternatively be conservatively approximated with a closed-form solution. Based on the parameters from EARTH, numerical results show that in the dense urban scenario, compared to the traditional macro-only homogeneous cellular network with no BS sleeping, deploying micro BSs can reduce about 40% of the total energy cost, and further reduce up to 35% with BS sleeping capability.

In this paper, a simple and elegant geometric water-filling (GWF) approach is proposed to solve the unweighted and weighted radio resource allocation problems. Unlike the conventional water-filling (CWF) algorithm, we eliminate the step to find the water level through solving a non-linear system from the Karush-Kuhn-Tucker conditions of the target problem. The proposed GWF requires less computation than the CWF algorithm, under the same memory requirement and sorted parameters. Furthermore, the proposed GWF avoids complicated derivation, such as derivative or gradient operations in conventional optimization methods, while provides insights to the problems and the exact solutions to the target problems. Most importantly, the GWF can be extended to solve a generalized form of radio resource allocation problem with more stringent constraints: (weighted) optimization problem with individual peak power constraints (GWFPP), and to include (weighted) group bounded power constraints (GWFGBP). On the other side, the CWF cannot solve these two general forms of the RRA problems, due to the difficulty to solve the non-linear system with multiple non-linear equations and inequalities in multiple dual variables. Optimality of the proposed water-filling solution is strictly proved for each of the proposed algorithms. Furthermore, numerical results show that the proposed approach is effective, efficient, easy to follow and insight-seeing.

In this paper, traffic-aware sleeping control (SC) and power matching (PM) of a single base station (BS) in cellular networks are studied. The objective is to find the sleeping control and power matching configurations that achieve the Pareto optimal tradeoff between total power consumption and average delay. Two types of sleeping control schemes are considered: The BS goes to sleep whenever there is no active user, and wakes up when N users are assembled or after a period of multiple or single vacation time. We first discuss when to incorporate sleeping control into power matching energy efficiently. The explicit relationship between total power consumption and average delay with varying service rate is analyzed theoretically, indicating that sacrificing delay cannot always be traded for energy saving, and we also provide conditions under which the energy-optimal rate exists. Moreover, the optimal pair of sleeping parameter and service rate to achieve the optimal energy-delay tradeoff, and the energy consumption lower bound are also derived. Both the analytical and simulation results show that tolerable sacrifice of delay performance can be traded for substantial amount of energy saving given that careful designs were made according to our analysis.

In this paper, we address the problem of providing video-on-demand (VoD) services to numerous clients energy-efficiently. To reduce energy consumption, multiple requests for the same video are batched and served by one single multicast stream. However, this brings additional delay to most clients. Utilizing client pre-caching is an efficient way to eliminate the delay: while the server is batching multiple requests, the clients can play the locally cached prefix of the requested video. The multicast session containing the later part of a video can be delayed till the prefix is played out. Our evaluation demonstrates that with a carefully designed pre-caching scheme, even a small cache (with the size of a video) can reduce 50% energy consumption. Moreover, we determine the optimal client cache allocation scheme to maximize the utilization of client cache and further minimize the energy consumption.

To fully understand the capacity of clustered network-MIMO systems and analyze the system performance (throughput, energy-efficiency or quality of service), one must have an analytical expression of the system capacity or capacity bounds. In this paper, the impact of cluster size on the downlink network-MIMO system capacity is analyzed considering inter-cluster interference (ICLI) based on the 1-dimensional Wyner model. For the nonfading channels, the lower and upper bounds of the per-cell capacity with ICLI are derived. The per-cell capacity with ICLI demonstrates a linear growth versus the cluster size in the interference-limited regime due to ICLI. The lower and upper bounds are generalized to a 2-dimensional cellular system. The ICLI turns out to have a great impact on the system capacity when the cluster size is small. Introducing Rayleigh fading channels, assuming the number of users in each cell is sufficiently large, a lower bound of the per-cell ergodic capacity with ICLI is derived.

Traffic-aware resource allocation and base station (BS) sleep control are key methods for energy saving in cellular networks. In this paper, first, we consider the control problem of how to adapt transmit power according to flow-level traffic variations, which leverages the tradeoff between energy consumption and delay performance. Based on different time scales of traffic variations, two power adaptation strategies are investigated: load-aware and queue-aware. The two strategies adapt transmit power according to flow arrival rate and instantaneous number of flows, respectively. Optimal solutions are given for both strategies. Since the optimal solution of the queue-aware strategy has no explicit form, tight bounds are derived as an approximation. Simulation results show that the two strategies perform closely in terms of energy consumption and average delay, while the queue-aware strategy is better in the tail distribution of delay and is more robust to system parameter variations. Secondly, for the load-aware strategy, with more practical concerns like the total BS energy consumption and BS sleep control taken into account, the relationship between energy consumption and delay is explored and energy-optimal rate can be obtained under certain conditions. Two threshold-based BS sleep strategies are investigated where the optimal threshold and rate are derived respectively.

In this paper, we adopt stochastic geometry theory to analyze the optimal macro/micro BS (base station) density for energy-efficient heterogeneous cellular networks with QoS constraints. We first derive the upper and lower bounds of the optimal BS density for homogeneous scenarios and, based on these, we analyze the optimal BS density for heterogeneous networks. The optimal macro/micro BS density can be calculated numerically through our analysis, and the closed-form approximation is also derived. Our results reveal the best type of BSs to be deployed for capacity extension, or to be switched off for energy saving. Specifically, if the ratio between the micro BS cost and the macro BS cost is lower than a threshold, which is a function of path loss and their transmit power, the micro BSs are preferred, i.e., deploy more micro BSs for capacity extension or switch off certain macro BSs for energy saving. Otherwise, the optimal choice is the opposite. Our work provides guidance for energy efficient cellular network planning and dynamic operation control.

Due to the limited diffracting ability of millimeter wave (mm-wave) signals, solving the blockage problem in 60GHz WLANs is one of major challenges. Since multiple access points (AP) are typically employed in the indoor environment to ensure the coverage and to meet the increasing demand for extremely high data rate applications, in this paper, we propose a multi-AP architecture with which a MAC layer device called an Access Controller (AC) is employed to enable each station to associate and cooperate with multiple APs. In this way, multi-AP diversity can be exploited to solve the blockage problem. Since APs may not fully obtain the state of channels, we formulate the AP selection problem as a Partially Observed Markov Decision Process (POMDP) and obtain an optimal policy. Moreover, a threshold-based policy with reduced complexity is developed with which the decision for AP selection depends only on the number of consecutive transmission failures. The optimal threshold value for AP selection is also derived. Simulation results show that the threshold-based policy achieves almost the same performance as the optimal policy derived from POMDP.

In traditional cellular networks, the network planning scheme is imperative for satisfying the coverage and traffic requirement. The rapidly growing number of users in today's cellular networks demands more base stations (BSs) to accommodate the increasing traffic load. The dense deployment results in severe energy consumption, which is often overlooked by existing network planning schemes. In this paper, an energy-aware network planning scheme is proposed to reduce the energy consumption of BSs by leveraging the coverage extension functionality of the inter-cell cooperation. The network planning problem is formulated as a mixed integer programming problem, which is solved with the Lagrangian relaxation method. Numerical results show that the energy efficiency can be enhanced by as much as 20% compared with the non-cooperative scheme without violating the QoS requirement. It is also shown that the proposed scheme is robust to the energy consumption structure of the BSs, and thus can be used with various types of BSs.

We consider multiple base station (BS) cooperative transmission in downlink cellular networks to improve the spectral efficiency and the system capacity. Grouping BSs into clusters is a practical solution to realize BSs cooperation and reduce system complexity. However, it still suffers from inter-cluster interference, especially for the cluster-edge users. In this paper, clustering and scheduling are jointly considered to deal with the problem. The clusters are formed dynamically from users' point of view to minimize the inter-cluster interference, and are allowed to be overlapped. Accordingly, coordinated precoding scheme is designed to manage the intra-cluster interference. A greedy scheduling algorithm is proposed jointly with dynamic clustering. Simulations show that the proposed joint algorithm provides impressive average throughput gain over the non-joint ones, and the user fairness is improved significantly.

The base station (BS) turning off scheme has been considered as a feasible solution to save energy of wireless networks. At the same time, it is also important to maintain quality of service (QoS) of the cell whose BS is turned off. In this paper, we consider multi-hop relay (MR) and cooperative transmission (CT) in cellular network and evaluate performances focusing on the tradeoff between QoS (i.e., user throughput and outage probability) and energy consumption when a BS is turned off. The result shows that combining cellular network with MR (MR network) is more robust in maintaining QoS level when a BS is turned off. Moreover, the MR network can also reduce overall energy consumption depending on the energy consumption level of a relay station (RS). Finally, a cooperative multi-hop relay (CMR) network which supports both intra-sector and inter-sector cooperative relaying is proposed to fill up the coverage holes.

We consider the problem of allocating transmit power in the downlink of a distributed wireless communication system. We account for the power used in both channel estimation and data transmission, with the objective of minimizing the overall transmitted power while satisfying specified Quality of Service (QoS) constraints to the mobile users. We consider both single user and multi-user power control optimization; the problem formulation for both cases lead to a nonconvex program. We proposed solution strategies for both scenarios: For the single user case, a simple intuitive solution, where power is allocated to the antennas sequentially until the QoS constraint is satisfied, is presented. For the multi-user case, we use successive convex approximation (based on the single condensation method) to find a provably convergent solution. We also demonstrate, via numerical simulation, the convergence of the proposed multi-user power allocation strategy. Our numerical results indicate that the proposed single and multi-user power allocation lead to an overall savings of up to 45% when compared to the baseline method of equal power allocation.

Adaptive modulation and coding (AMC) has been widely used to improve the spectral efficiency. In this paper, we take a different look at it from energy saving point of view. Specifically, we analyze the queuing behavior of AMC systems jointly with sleep mode where the wake-up process incurs time and energy cost. We formulate the optimization problem by jointly considering energy-efficiency, queuing delay and packet loss rate, and find the solution with cross-layer adjustment of the transmit power and the sleep threshold. Numerical results show that at low traffic range, when the power consumption of idle (no data transmission) mode is un-negligible, introducing sleep mode to the AMC system significantly improves the energy efficiency compared with non-sleep system. To achieve the energy-efficiency gain, the system tends to use higher-order modulation by increasing transmit power, which also reduces the number of dropped packets.

We consider transceiver design in uncoded multiple input multiple-output (MIMO) systems with noisy channel state estimates. Specifically, we design a transceiver that takes into account the statistics of the CSI errors to minimize the average bit error rate (BER) of the system. Our design utilizes the noisy CSI estimates and the error statistics at the transmitter to partition the spatial channels into 'almost' independent streams. We also propose a joint bit and power loading (allocation) scheme to allocate information rate and power to each spatial stream. Exact maximum likelihood (ML) decoding incurs a high complexity at the receiver; to circumvent this, stream-by-stream ML decoding is used at the receiver. We verify, via numerical results, that for a 4 × 4 system, the BER performance of the proposed joint bit and power loading transmission scheme far surpasses that of the schemes where only bit loading or only power loading is used. At a BER of 10-3, the joint bit and power loading scheme has an approximately 4 dB gain over the bit loading scheme. In contrast, the scheme that ignores CSI errors has poor BER performance.

We consider the joint optimization of relay station (RS) placement and RS sleep/active probability to enhance the energy efficiency of a one-dimensional cellular network. When the RSs are always active, conditions for optimal RS placement that minimizes transmission power are derived, based on which closed-form solution is obtained with path-loss exponent being two, and a simple numerical method for general values of path-loss exponent is proposed. When the circuit power consumption of active RSs is considered, RSs should enter sleep mode appropriately to save power. An algorithm based on projected Newton method is proposed to jointly optimize the RS placement and sleep/active probability. It is shown via numerical examples that the benefit of implementing RSs and optimizing RS placement is substantial and increases with the path-loss exponent. We also justify the interaction between RS placement and RS sleep control, which is effectively tackled by the proposed algorithm to minimize the total power consumption.

This paper focuses on quantized channel state information (CSI) feedback for downlink network MIMO systems. Specifically, we propose to quantize and feedback the CSI of a subset of BSs, namely the feedback set. Our analysis reveals the tradeoff between better interference mitigation with large feedback set and high CSI quantization precision with small feedback set. Given the number of feedback bits and instantaneous/long-term channel conditions, each user optimizes its feedback set distributively according to the expected SINR derived from our analysis. Simulation results show that the proposed feedback adaptation scheme provides substantial performance gain over non-adaptive schemes, and is able to effectively exploit the benefits of network MIMO under various feedback bit budgets.

Currently, two cooperative transmission strategies, Coordinated Multi-Point (CoMP) Transmission and wireless relaying, are expected to be deployed in future cellular systems to improve the performance of cell-edge users. Due to the cooperation diversity, these technologies can potentially lead to more energy saving. Therefore in this paper, we analyze their energy saving performance with an average outage constraint. The impact of the traffic intensity and BS density are also investigated. Based on the typical parameters setting, our calculation results show that traffic intensity can be divided into three classes: ``coverage-limited'' region, ``energy-efficient'' region, and ``capacity-limited'' region. The coverage-limited region prefers offline fixed algorithms, while dynamic online algorithms are more suitable for the energy-efficient region. As BS density goes higher, the energy-efficient region becomes larger. However, the traffic load region where the cooperation schemes bring benefits becomes smaller. To overcome CoMP, relay stations need to consume energy as small as possible. Even through relay cost energy low enough, the traffic intensity region where wireless relaying overcomes CoMP will become smaller as network goes denser. Our analytical results are of great help for future ``green'' network planning.

The energy consumption of information and communication technology (ICT) industry has become a serious problem, which mostly comes from the network infrastructure, rather than the mobile terminals. In this paper, we consider densely deployed cellular networks where the coverage of base stations (BSs) overlaps and the traffic intensity varies over time and space. An energy saving algorithm is proposed by dynamically adjusting the working modes (active or sleeping) of BSs according to the traffic variation with respect to certain blocking probability requirement. In addition, to prevent frequent mode switching, BSs are set to hold their current working modes for at least a given interval. Simulations demonstrate that the proposed strategy can greatly reduce energy consumption with blocking probability guarantee, and the performance is insensitive to the mode holding time within certain range.

Multiple base station (Multi-BS) cooperation has been considered as a promising mechanism to suppress cochannel interference and boost the capacity for cellular networks. However, the large feedback and signaling overhead hinder it from practice. Therefore, limited cooperation among BSs is recognized as a good tradeoff between the performance gain and the relevant cost. In this paper, the whole network is divided into small disjointing BS cooperation groups, namely, clusters. A decentralized framework is proposed to facilitate the BS cluster formation on the downlink, in order to maximize the sum-rate of the scheduled mobile stations (MSs) under the cluster size constraint. Moreover, an efficient BS negotiation algorithm is designed for cluster formation, of which the feedback overhead per MS is irrelevant to the network size, and the number of iteration rounds scales very slowly with the network size. Simulations show that our strategy leads to significant sum-rate gain over static clustering and performs almost the same as the centralized greedy approach. With its low signaling overhead and complexity, the proposed framework is well suited for implementation in large-scale cellular networks.

In this paper, a distributed power control algorithm is proposed for wireless relay networks in interference-limited environments. The objective is to minimize the total transmission power while satisfying the signal-to-interference-plus-noise ratio (SINR) requirements. Two forwarding techniques, i.e., decode- and-forward (DF) and amplify-and-forward (AF), are considered. The proposed algorithm only requires locally measured SINR on the relay nodes (RNs) and the destination nodes (DNs), based on which each cooperation unit (defined as one source node (SN) and DN pair with the RN associated to it) iteratively updates the transmission power of the SN and the RN by solving a local optimization problem. We prove that the convergence is guaranteed when the parameters adopted in the algorithm are sufficiently large, and then a parameter adjusting method is also designed. Simulation results indicate that the proposed algorithm converges fast and leads to only 7% more power consumption than the optimal power allocation in the considered scenarios. It is also shown that even in interference-limited environments, relaying can still improve system performance substantially in terms of outage and power consumption.

The combined MIMO with adaptive modulation and coding (AMC) technology can provide high spectral efficiency and link robustness. While most existing adaptive algorithms focus on physical layer, cross-layer analysis on the queuing behavior of MIMO-AMC systems is necessary, but remains open. In this paper, under the conditions of unsaturated traffic and finite-length buffer, we investigate the queuing characters of two representative categories of MIMO systems, namely the BLAST system and the space-time block coding (STBC) system. We model the service processes of both STBC and BLAST coupled with AMC, which is the most challenging part of the queuing analysis. We observe a new tradeoff between diversity and multiplexing in terms of link layer packet loss rate and queuing delay, based on which we propose a cross-layer design of diversity-multiplexing switching scheme to optimize the QoS of the MIMO-AMC systems.

In this paper, we propose a contention based uplink medium access control (MAC) protocol design for wireless local area networks (WLANs) with spatial division multiple access (SDMA) support. Our protocol does not require sophisticated smart antenna equipments, and it can be implemented in simple omni-directional multiple-antenna WLANs. Different from the super-frame based approaches, the proposed one is a pure contention based MAC protocol and can be easily implemented into standard 802.11 systems with slight modifications. By jointly considering the the physical and the MAC layer situations, dynamic system parameter adjustment is designed to enhance throughput and protocol efficiency. In addition, our protocol provides interface for user scheduling, which makes it more extensible. Simulation results show that our scheme can achieve high network throughput, and discussions regarding different system factors are also included.

The combined OFDM/SDMA approach has raised lots of research interests recently as it appears to be quite suitable for future broadband wireless transmission. In order to build practical OFDM/SDMA systems, we need the performance evaluation of such systems under real-world conditions. The performance degradation exists of an OFDM/SDMA system due to carrier frequency offsets (CFO), because single-user plain OFDM is highly sensitive to CFO. However, the characteristic of the degradation is not clear. In this paper, the impact of CFO on the performance of an uplink OFDM/SDMA system is analyzed. Different from the existing analysis, we jointly consider the impact of CFO on channel estimation and multi-user signal detection. Analytical expressions are derived for the variance of channel estimation error and multi-user signal detection error in the presence of CFO, whose accuracies are validated by simulation results. Observations related to channel estimation accuracy and the interaction of the CFO effects from different users are also given.