Cooperative device-free localization, which involves node cooperation for location-awareness services, represents pivotal opportunities for forthcoming integrated sensing and communication (ISAC) networks. In this paper, a theoretical framework is established to quantify the performance limits of device-free localization accuracy in ISAC networks. It is achieved through Fisher information analysis, accounting for signal randomness, transceiver deployment, multi-target, multipath, etc. To explore node cooperation, the theoretically optimal linear fusion, inspired by the best linear unbiased estimator, is derived. It linearly combines independent target location estimates from multi-static transceivers for localization accuracy enhancement. The optimal weights are determined by the corresponding effective Fisher information, which later constitutes the definition of theFisher Information Field(FIF), as a spatial field to quantify useful information that can be gained through location estimation. An FIF-based practical location estimator using multi-source estimates is further developed. Finally, a case study on an orthogonal frequency division multiplexing ISAC network is conducted. The target is independently localized from multi-static transceivers using joint range-angle maximum likelihood estimation; closed-form evaluations of the Fisher information are derived for performance limits study and FIF-based cooperative localization. Numerical results offer insights into the Fisher information analysis and demonstrate low complexity and high accuracy of the FIF-based estimator compared with benchmarks.

Integrated sensing and communication (ISAC) has garnered significant attention in recent years. In this paper, we delve into the topic of sensing-assisted communication within ISAC systems. More specifically, a novel sensing-assisted channel estimation scheme is proposed for bistatic orthogonal-frequency-division-multiplexing (OFDM) ISAC systems. A framework of sensing-assisted channel estimator is first developed, integrating a tailored low-complexity sensing algorithm to facilitate real-time channel estimation and decoding. To address the potential sensing errors caused by low-complexity sensing algorithms, a sensing-assisted linear minimum mean square error (LMMSE) estimation algorithm is then developed. This algorithm incorporates tolerance factors designed to account for deviations between estimated and true channel parameters, enabling the construction of robust correlation matrices for LMMSE estimation. Additionally, we establish a systematic mechanism for determining these tolerance factors. A comprehensive analysis of the normalized mean square error (NMSE) performance and computational complexity is finally conducted, providing valuable insights into the selection of the estimator’s parameters. The effectiveness of our proposed scheme is validated by extensive simulations. Compared to existing methods, our proposed scheme demonstrates superior performance, particularly in high signal-to-noise ratio (SNR) regions or with large bandwidths, while maintaining low computational complexity.

Wireless neural receivers provide a promising alternative to conventional receivers. To perform well in different channel environments, online adaption is required. However, during this process, performance remains low. Thus, an approach called federated collaborative learning with pruned-data aggregation (FedPDA) is developed to reduce online-adaptation frequency. The basic idea is that, upon online adaptation, mobile terminals further update their neural receivers collaboratively via federated learning. To reduce memory consumption, neural receivers follow a main-side network architecture where only the side network needs retraining during collaborative learning. To avoid catastrophic forgetting during continual learning, local data on terminals are pruned, with only a small percent sent to the base station. With such data, the base station also trains a neural receiver before conducting model aggregation. FedPDA is distinct with several features: 1) small memory footprint and no storage burden on terminals; 2) no catastrophic forgetting issue; 3) low communication cost. Performance results show that FedPDA reduces online adaptation by more than 90% and memory footprint by 70%. It achieves comparable performance as centralized schemes, but reducing communication cost by 78%. Compared to vanilla federated learning, FedPDA resolves the catastrophic forgetting issue without storage burden, and also reduces the communication cost by 50%.

With the development of Internet of Vehicles (IoV), vehicle-to-infrastructure (V2I) communications are becoming attractive for vehicle users (VUEs) to obtain diverse cloud service through base stations (BSs). To tackle V2I link deterioration caused by blockage and out-of-coverage cases, multihop vehicle-to-everything (V2X) routing with both vehicle-to-vehicle (V2V) and V2I links needs to be investigated. However, traditional routing reacts to statistical or real-time information, which may suffer link degradation during path switchover in fast-changing vehicular networks. Predictive routing protocols take timely actions by forecasting link connectivity, but they fail to satisfy specific QoS requirements. Low robustness to link failures is also incurred without considering imperfect prediction. To build continual paths between VUEs and BSs for QoS provision of cloud service, a robust predictive routing framework (ROPE) is proposed with three major components: 1) an early warning scheme detects V2I link deterioration in advance via predicting vehicle mobility and link signal strength to facilitate seamless path switchover; 2) a virtual routing mechanism finds top3 paths that have the highest path strength and satisfy the connectivity and hop count constraints based on the prediction results to fulfill QoS requirements of cloud service; and 3) a path verification protocol checks availability and quality of the top3 paths shortly before switchover and activates one qualified path for switchover to ensure routing robustness. Extensive experiments are conducted in a holistic simulation framework, which demonstrate the superiority of ROPE over direct V2I communications and existing predictive routing protocols under various scenarios.

Backhaul mesh networks are critical for ensuring coverage and connectivity of high-frequency 6G networks. To maintain high throughput, its architecture needs to be augmented by adding relays. However, how to place relays at appropriate sites poses two challenges: 1) there lacks a theory to capture the relationship between a certain change of network architecture and its throughput gain; 2) selecting the best sites for relays is a complicated combinatorial problem. To tackle the first challenge, this paper first establishes a clique-based bottleneck theory, through which a clique-based bottleneck structure of a given network architecture is constructed to determine the network throughput. Based on this bottleneck structure, clique gradients are then computed to quantify the impact of each clique on the overall network throughput. With the clique-based bottleneck theory, the second challenge is resolved by embedding clique gradients into a deep reinforcement learning (DRL) scheme. Specifically, the DRL actions are masked such that only the relay sites that match the highest clique gradients are selected. This DRL-based relay placement (DeepRP) scheme is evaluated via extensive simulations, and performance results show that it can boost network throughput by more than 50%, which is$\text{10.4} \!-\! \text{32.1}\% $higher than those of baseline schemes.

The interference graph, depicting the intra- and inter-cell interference channel gains, is indispensable for resource allocation in multi-cell networks. However, there lacks viable methods of interference graph estimation (IGE) for multi-cell multi-numerology (MN) networks. To fill this gap, we propose an efficient power-domain approach to IGE for the resource allocation in multi-cell MN networks. Unlike traditional reference signal-based approaches that consume frequency-time resources, our approach uses power as a new dimension for the estimation of channel gains. By carefully controlling the transmit powers of base stations, our approach is capable of estimating both intra- and inter-cell interference channel gains. As a power-domain approach, it can be seamlessly integrated with the resource allocation such that IGE and resource allocation can be conducted simultaneously using the same frequency-time resources. We derive the necessary conditions for the power-domain IGE and design a practical power control scheme. We formulate a multi-objective joint optimization problem of IGE and resource allocation, propose iterative solutions with proven convergence, and analyze the computational complexity. Our simulation results show that power-domain IGE can accurately estimate strong interference channel gains with low power overhead and is robust to carrier frequency and timing offsets.

Training data in federated learning (FL) frameworks can have label noise, since they must be stored and annotated on clients' devices. If trained over such corrupted data, the models learn the wrong knowledge of label noise, which highly degrades their performance. Although several FL schemes are designed to combat label noise, they suffer performance degradation when the clients' devices only have limited local training samples. To this end, a new scheme called federated label-noise learning (FedLNL) is developed in this paper. The key problem of FedLNL is how to estimate a noise transition matrix (NTM) accurately in the case of limited local training samples. If a gradient-based update method is used to update the local NTM on each client's device, it can generate too large gradients for the local NTM, causing a high estimation error of the local NTM. To tackle this issue, an alternating update method for the local NTM and the local classifier is designed in FedLNL, where the local NTM is updated by a Bayesian inference-based update method. Such an alternating update method makes the loss function of existing NTM-based schemes not applicable to FedLNL. To enable federated optimization of FedLNL, a new regularizer on the parameters of the classifier called local diversity product regularizer is designed for the loss function of FedLNL. The results show that FedLNL improves the test accuracy of a trained model by up to 25.98%, compared with the state-of-the-art FL schemes that tackle label-noise issues.

Integrated sensing and communications (ISAC) is considered a promising technology in the B5G/6G networks. The channel model is essential for an ISAC system to evaluate the communication and sensing performance. Most existing channel modeling studies focus on the monostatic ISAC channel. In this paper, the channel modeling framework for bistatic ISAC is considered. The proposed channel modeling scheme extends the current 3GPP channel modeling framework and ensures the compatibility with the communication channel model. To support the bistatic sensing function, several key features for sensing are added. First, more clusters with weaker power are generated and retained to characterize the potential sensing targets. Second, the target model can be either deterministic or statistical, based on different sensing scenarios. Furthermore, for the statistical case, rays are generated considering spatial coherence for different reflection models. The effectiveness of the proposed bistatic ISAC channel model is validated by both Ray-tracing simulations and experiment studies. The compatibility with the 3GPP communication channel model is also demonstrated.

Image compositing plays a vital role in photo editing. After inserting a foreground object into another background image, the composite image may look unnatural and inharmonious. When the foreground is photorealistic and the background is an artistic painting, painterly image harmonization aims to transfer the style of background painting to the foreground object, which is a challenging task due to the large domain gap between foreground and background. In this work, we employ adversarial learning to bridge the domain gap between foreground feature map and background feature map. Specifically, we design a dual-encoder generator, in which the residual encoder produces the residual features added to the foreground feature map from main encoder. Then, a pixel-wise discriminator plays against the generator, encouraging the refined foreground feature map to be indistinguishable from background feature map. Extensive experiments demonstrate that our method could achieve more harmonious and visually appealing results than previous methods.

Interactive services of mobile edge computing demand low latency in task handling, which cannot be easily satisfied due to limited per-user computing power on an edge server. Fortunately, a user can establish links with multiple base stations and the co-located edge servers in 5 G and beyond. By leveraging multi-connectivity, tasks of a computing service can be dispatched to multiple edge servers. This approach can also improve the quality of services, since tasks allocated to edge servers can be prioritized according to link quality. To exploit the benefits of multi-connectivity, a task scheduling problem is formulated to minimize latency and maximize reliability of the application, subject to task dependency and resource constraints. This problem is hard to solve due to NP-hardness and time-varying conditions. To this end, a pointer network based scheme called PtrTasking is developed to obtain a deep-learning model for the schedules and then infer new schedules on-line. To support time-varying conditions (e.g., a new application or evident change of computing load), a training strategy called PtrTrain is designed to retrain PtrTasking in a fast and efficient way. Experiments based on real-world datasets demonstrate that PtrTasking significantly improves latency and reliability, as compared to the baseline schemes.

Efficient one-hop multicasting (OHM) of high-volume sensor data plays a pivotal role in the success of cooperative autonomous driving applications. Although millimeter-Wave (mmWave) bands demonstrate huge potential for high- bandwidth OHM data transmission, the challenge lies in enabling individual vehicles to locate and communicate with suitable neighbors in a fully distributed and highly dynamic scenario. This paper introduces mmV2V, a fully distributed OHM scheme designed for vehicular networks, comprising three tightly integrated protocols. Initially, synchronized vehicles perform a probabilistic neighbor discovery procedure, wherein randomly divided transmitters (or receivers) clockwise scan (or listen to) the surroundings in synchronization with heterogeneous Tx (or Rx) beams. This approach facilitates the identification of the vast majority of neighbors within a few repeated rounds. Subsequently, vehicles engage in negotiations with their neighbors to establish an optimal communication schedule in evenly distributed slots. Finally, matched pairs of neighboring vehicles commence high data rate transmissions using refined beams. We implement a prototype testbed to validate the feasibility of the main components of mmV2V. Extensive simulations based on generated and real-world traffic traces are conducted and the results demonstrate that mmV2V consistently achieves a high completion ratio in demanding OHM tasks across various traffic conditions.

It is prevalent for a mobile edge device to conduct local inference using a compact machine learning model, which achieves lower latency and less compromise of data privacy as compared to cloud-based inference. To work in a new environment, the compact model needs to be adapted to the target data from the environment so as to maintain a high inference accuracy. However, directly applying domain adaptation to the compact model leads to a low inference accuracy. Hence, a scheme called memory-efficient collaborative domain adaptation (MEC-DA) is developed in this paper to boost the compact model's inference accuracy on the target data while preserving data privacy. It first deploys a large model to the mobile edge devices where domain adaptation is conducted to adapt the large model to the target data. This process requires training of the large model, which causes high memory consumption. A new method called lite residual hypothesis transfer (LRHT) is thus designed to achieve memory-efficient domain adaptation. The knowledge of the large model is then transferred to the compact model via knowledge distillation. To prevent the compact model from forgetting the knowledge of the source data, a collaborative knowledge distillation (Co-KD) method is developed to unify the source data on the server and the target data on an edge device to update the compact model. MEC-DA can protect data privacy and handle user mobility properly via secure aggregation and user selection, respectively. Extensive experiments on several tasks of object recognition show that MEC-DA improves the inference accuracy by up to$12.5\%$, as compared to the state-of-the-art schemes.

The advancement of nanosatellite techniques has boosted the growth of satellite-originated data and applications. Satellite edge computing (SEC) is envisioned to provide in-orbit processing of the sensed data to save the scarce terrestrial-satellite communication resources and support mission-critical services. While most of the existing SEC studies mainly focus on general computing tasks, we present a two-tier collaborative processing framework for the important and unique hyperspectral image (HSI) processing task. Our framework carefully selects bands out of the collected HSIs and sends them back for further analysis. We first conduct a comprehensive data analysis to reveal the non-trivial relationship between the band selection and the eventual analytic performance. We then formulate the band selection problem in this collaborative setting as a utility maximization problem that jointly considers the analytic, energy, and communication factors. A novel multi-agent reinforcement learning approach, named MaHSI, is proposed to solve it in the dynamic SEC environment. Our multi-agent design judiciously embeds the complex correlations among bands as collaborations among agents and significantly reduces the exploration space. Extensive experiments on real-world HSI datasets prove that our approach not only outperforms the existing classical band selection algorithms in accuracy and inference speed but also brings the highest utility to the satellites.

A federated kernel k -means (FedKKM) algorithm is developed in this article to conduct distributed clustering with low memory consumption on user devices. In FedKKM, a federated eigenvector approximation (FEA) algorithm is designed to iteratively determine the low-dimensional approximate vectors of the transformed feature vectors, using only low-dimensional random feature vectors. To maintain high communication efficiency in each iteration of FEA, a communication-efficient Lanczos algorithm (CELA) is further designed in FEA to reduce the communication cost. Based on the low-dimensional approximate vectors, the clustering result is obtained by leveraging a distributed linear k -means algorithm. A theoretical analysis shows that: 1) FEA has a convergence rate of O(1/T) , where T is the number of iterations; 2) the scalability of FedKKM is not affected by the dataset size since the communication cost of FedKKM is independent of the number of users' data; and 3) FedKKM is a (1+ϵ) approximation algorithm. The experimental results show that FedKKM achieves the comparable clustering quality to that of a centralized kernel k -means. Compared with state-of-the-art schemes, FedKKM reduces the memory consumption on user devices by up to 94% and also reduces the communication cost by more than 40%.

A federated learning (FL) scheme (denoted as Fed-KSVM) is designed to train kernel support vector machines (SVMs) over multiple edge devices with low memory consumption. To decompose the training process of kernel SVM, each edge device first constructs high-dimensional random feature vectors of its local data, and then trains a local SVM model over the random feature vectors. To reduce the memory consumption on each edge device, the optimization problem of the local model is divided into several subproblems. Each subproblem only optimizes a subset of the model parameters over a block of random feature vectors with a low dimension. To achieve the same optimal solution to the original optimization problem, an incremental learning algorithm called block boosting is designed to solve these subproblems sequentially. After training of the local models, the central server constructs a global SVM model by averaging the model parameters of these local models. Fed-KSVM only increases the iterations of training the local SVM models to save the memory consumption, while the communication rounds between the edge devices and the central server are not affected. Theoretical analysis shows that the kernel SVM model trained by Fed-KSVM converges to the optimal model with a linear convergence rate. Because of such a fast convergence rate, Fed-KSVM reduces the communication cost during training by up to 99% compared with the centralized training method. The experimental results also show that Fed-KSVM reduces the memory consumption on the edge devices by nearly 90% while achieving the highest test accuracy, compared with the state-of-the-art schemes.

Supporting ultra-high data rates and flexible reconfigurability, Terahertz (THz) mesh networks are attractive for next-generation wireless backhaul systems that empower the integrated access and backhaul (IAB). In THz mesh backhaul networks, the efficient cross-layer routing and long-term resource allocation is yet an open problem due to dynamic traffic demands as well as possible link failures caused by the high directivity and high non-line-of-sight (NLoS) path loss of THz spectrum. In addition, unpredictable data traffic and the mixed integer programming property with the NP-hard nature further challenge the effective routing and long-term resource allocation design. In this paper, a deep reinforcement learning (DRL) based cross-layer design in THz mesh backhaul networks (DEFLECT) is proposed, by considering dynamic traffic demands and possible sudden link failures. In DEFLECT, a heuristic routing metric is first devised to facilitate resource efficiency (RE) enhancement regarding energy and sub-array usages. Furthermore, a DRL based resource allocation algorithm is developed to realize long-term RE maximization and fast recovery from broken links. Specifically in the DRL method, the exploited multi-task structure cooperatively benefits joint power and sub-array allocation. Additionally, the leveraged hierarchical architecture realizes tailored resource allocation for each base station and learned knowledge transfer for fast recovery. Simulation results show that DEFLECT routing consumes less resource, compared to the minimal hop-count metric. Moreover, unlike conventional DRL methods causing packet loss and second-level latency, DEFLECT DRL realizes the long-term RE maximization with no packet loss and millisecond-level latency, and recovers resource-efficient backhaul from broken links within 1s.

Traditional wisdom for network resource management allocates separate frequency-time resources for measurement and data transmission tasks. As a result, the two types of tasks have to compete for resources, and a heavy measurement task inevitably reduces available resources for data transmission. This prevents interference graph estimation (IGE), a heavy yet important measurement task, from being widely used in practice. To resolve this issue, we propose to use power as a new dimension for interference measurement in full-duplex millimeter-wave backhaul networks, such that data transmission and measurement can be done simultaneously using the same frequency-time resources. Our core insight is to consider the mmWave network as a linear system, where the received power of a node is a linear combination of the channel gains. By controlling the powers of transmitters, we can find unique solutions for the channel gains of interference links and use them to estimate the interference. To accomplish resource allocation and IGE simultaneously, we jointly optimize resource allocation and IGE with power control. Extensive simulations show that significant links in the interference graph can be accurately estimated with minimal extra power consumption, independent of the time and carrier frequency offsets between nodes.

Terahertz (THz) communications in 6G offer high data rates and large bandwidths, but the effective range is limited by atmospheric absorption and scattering. Beamformer forms narrow beams to improve directional gain, but beamforming alone has limited accuracy, resulting in degraded communication quality, so beam alignment is necessary. Traditional linear scanning methods for beam alignment suffer from low accuracy due to restrictions on the beam sweeping step size. This problem can be mitigated in an integrated sensing and communication (ISAC) system that has strong built-in sensing ability to localize the target. To this end, this paper builds theoretical foundation on discussion of the relationship between estimation accuracy and beam misalignment displacement, and proposes an iterative sensing-assisted beam alignment method that iteratively aligns the beam directly to the target location estimated from the sensing system for continuous refinement of beam orientation. To support the proposed method, an ISAC hybrid bandwidth-time resource allocation scheme between communication and sensing sub-systems is designed. Numerical results validate that if allocating sufficient bandwidth-time resources to the sensing sub-system, the proposed method improves the system's capacity by 18.6% and 142.5% compared with traditional beam alignment and beamforming-only approaches, respectively.

Pulse-based communication systems are essential for terahertz (THz) communications, offering exceptional performance in diverse application scenarios, such as nanonetworks, THz integrated sensing and communications, and short-range ultra-high-speed communications. However, classical pulse-based modulations, such as time spread on-off keying (TS-OOK), pulse amplitude modulation (PAM), and pulse position modulation (PPM), suffer from limited capacity, mainly due to the absence of effective high-order modulation schemes and the long duration of consecutive transmission between adjacent pulses. Although these modulations have low energy consumption, they are difficult to balance the relationship between capacity and energy consumption. Recently, the generation of THz frequency and bandwidth continuously tunable (FBCT) pulses has enabled pulse-based M-ary QAM (PMQAM) to improve capacity, but it increases energy consumption compared with the classical modulations. To address these issues, an efficient high-order modulation method called position interleaved pulse modulation (PIPM) is proposed in this paper. More specifically, PIPM increases the dimension of modulation by using multiple pulse positions and interleaves the odd and even components of FBCT orthogonal pulses at these positions. In this design, PIPM utilizes the information from the odd component, even component, and pulse position to achieve high-order modulation. Furthermore, PIPM can balance the capacity and energy consumption by flexibly adjusting the relationship between the three-dimensional information. Detailed analysis of PIPM and its performance under multiple access scenarios is performed in this paper. Numerical results show that PIMP achieves a much better balance between energy efficiency and spectrum efficiency, as compared to existing schemes.

Terahertz (THz) mesh networking is envisioned as a promising technology for 6G networks, with network capacity as one of the most critical performance metrics. To boost the network capacity of a THz mesh network, link resource planning is conducted, which poses two challenges. First, the relationship between link resources and network capacity must be captured quantitatively considering the peculiarities of THz mesh networking. Second, multi-dimensional resources including subarrays, power, and subbands need to be determined for link resource planning. To address the first challenge, a bottleneck structure is constructed by adapting the quantitative theory of bottleneck structures in the recent work [1] for THz mesh networks, such that the relationship between network capacity and a certain link resource planning result is determined. Furthermore, bottleneck gradients are computed based on the constructed bottleneck structure. Given the derived relationship and bottleneck gradients, a heuristic link resource planning algorithm is designed to allocate multi-dimensional resources, thus resolving the second challenge. Performance results show that the heuristic resource planning algorithm can boost the network capacity by 20.3% - 41.8% for various topologies.

Recently, integrated sensing and communications (ISAC) design has attracted great attention for B5G/6G networks. Existing ISAC studies are mainly focused on monostatic sensing with a full-duplex radio. However, full-duplex radio requires complicated self-interference cancellations and many current devices are half-duplex radios. Therefore, it is still interesting to investigate passive bistatic sensing with half-duplex radios. In this paper, integrating passive bistatic sensing into mmWave B5G/6G networks is investigated. The public reference signals are leveraged to extract the frequency-domain channel state information (CSI) for passive sensing. To address the problems of sampling-timing-offset and random phase error, a line-of-sight (LoS) path aided calibration mechanism is first developed. To achieve accurate localization for multiple targets, a novel super-resolution channel impulse response (CIR) based mechanism is then developed. The super-resolution CIR is achieved from CSI through a joint design of spatial-smoothing multiple signal classification (MUSIC) algorithm and template-based tap estimation. With the super-resolution CIR, multiple targets can be first distinguished in CIR taps and then localized by further estimating their angle of arrivals (AoAs). The proposed design is implemented and validated on a prototype system at 28 GHz with 500 MHz bandwidth in indoor environments. Experimental results show that our design can achieve accurate single-target and multi-target localization and tracking. The localization error is 26 cm at 80thpercentile for single-person and 29 cm on average for multi-person setups. The average error to the planned path after the moving-averaged filter is 11 cm and 19 cm for single-person and two-person tracking, respectively.

The utilization of wireless mesh networks has gained increasingly significance recently, with its potential applications in beyond 5G (B5G) and 6G communications, vehicle-to-vehicle (V2V) communications, and many Internet-of-Things (IoT) scenarios. Despite such growing popularity, the achievable per-node throughput has been a critical factor that hinders the large scale deployment of wireless mesh networks. The study conducted by Gupta and Kumar [1] and other related work revealed that the per-node throughput with respect to the number of mesh nodes n degrades as $\Theta (1/\sqrt {n\log n} )$. This issue has been addressed in our previous work [2] where the scalable per-node throughput can be achieved in a well-designed multi-tier mesh network with an acceptable cost of bandwidth, transmit power, and antenna number. However, the delay in such a mesh network remains unknown. In this paper, the scaling behavior of the average delay under such a multi-tier mesh network is analyzed. Based on the analytical results, potential methods are explored to reduce the average end-to-end delay while maintaining the scalable per-node throughput. Such explorations offer valuable insights and guidance for realistic deployment of a scalable wireless mesh network.

The integrated sensing and communications (ISAC) design has attracted great attention in recent years. Existing studies mainly focus on the monostatic radar sensing based on the assumption of full-duplex radio. However, since current legacy devices are all half-duplex radios, it is still important to investigate bistatic radar sensing based on half-duplex radios. In this paper, the 5G mmWave communication system with a hybrid beamforming architecture is considered, and the joint transmit and receive beamforming is designed for incorporating bistatic radar sensing function into the multi-user multi-input multi-output (MU-MIMO) communications. To overcome the critical problem of carrier-frequency-offset (CFO) and sampling-time-offset (STO) in bistatic sensing, the line-of-sight (LoS) signal is regarded as a virtual target for eliminating the CFO and STO. Therefore, the transmitter and radar receiver beamformers are jointly designed to maximize the radar signal-to-interference-plus-noise ratio (SINR), while satisfying the SINR constraints of both communication users and the virtual radar target. A hierarchical solution algorithm is developed to resolve this problem, in which a combined fully digital beamformer is first optimized for transmitter/sensing-receiver, and then the analog and digital beamformers for hybrid beamforming are derived from the fully digital beamformer. Simulation results show that compared to the separate beamforming design and the passive opportunistic sensing design, our joint design can achieve more than 5 dB and 7-22 dB radar sensing gain, respectively.

In this paper, the backhauling topology of an integrated and backhaul (IAB) network is designed to sustain bursty traffic with the highest probability. To ensure sequential DU-MT or MT-DU transmissions, the topology is characterized by a directed acyclic graph (DAG). First, the Bayes' theorem is used to transform the probability of sustaining the UE traffic into the probability of generating a DAG, and then the transformed problem is decomposed into two subproblems: 1) The link traffic load is determined under the condition that a DAG is formed; 2) Based on the link traffic load, the links that are critical for sustaining the UE traffic are identified, and then a new DAG is generated by maximizing the joint probability of links in the new DAG. Given random initial DAG, the two subproblems are iteratively solved until obtaining a final DAG. Theoretical analysis validates that the above iterative procedures converge to a single DAG, and simulations confirm that the convergence can be achieved within tens of iterations. Simulation results also show that the backhauling approach developed in this paper can support 56.41% higher traffic variations and achieve a 29.32% lower average hop-count than the existing topology generation schemes.

In recent years, full-duplex communications in a single frequency channel have become a practical technology. However, existing full-duplex medium access control (MAC) protocols can hardly take full advantages of full-duplex transmission opportunities, because they are not capable of adding one more half-duplex link on an on-going half-duplex link to form a full-duplex link. Besides, the exposed node and hidden node issues in multi-hop wireless networks limit flexible establishment of full-duplex links. Such limitations waste full-duplex transmission opportunities, and thus degrade network performance. In this paper, an efficient full-duplex link establishment protocol called full-duplex opportunity exploitation (FDOE) is developed to exploit full-duplex transmission opportunities in a wireless mesh network. FDOE leverages rateless coding and pseudo-noise (PN) sequences to establish a full-duplex link whenever a full-duplex transmission opportunity is available. Such a distinct feature leads to significant improvement of network performance. FDOE is evaluated through theoretical analysis and extensive simulations. Performance results show that FDOE exploits full-duplex transmission opportunities efficiently in wireless mesh networks, and thus significantly outperforms existing MAC protocols.

Edge computing is promising for connected vehicles. As vehicles move, their resource demands for edge servers vary. Thus, it is necessary to reserve edge servers dynamically to meet variable demands. Existing schemes of edge-server reservation usually rely on statistical information of resource demands to make reservations; they are infeasible for connected vehicles, since such schemes are not adaptive to time-varying demands. To this end, a spatio-temporal reinforcement learning scheme called DeepReserve is developed to learn variable demands and then conduct edge-server reservation. Its design is based on the deep deterministic policy gradient algorithm of deep reinforcement learning (DRL), but is featured with several enhancements. First, the fully-connected neural network in DRL is replaced by a convolutional LSTM (ConvLSTM) network to extract spatio-temporal features of resource demands, which highly improves the prediction accuracy of resource demands. Thus, the actions in DRL (i.e., reservation decisions) can adapt to future demands. Second, an action amender is designed to ensure the actions selected by the neural network follow the spatio-temporal correlation. Finally, a training method called DR-Train is designed to stabilize the training procedure for different traffic patterns. DeepReserve is evaluated through extensive experiments on real-world datasets. Results show that it outperforms state-of-the-art approaches.

Federated learning (FL) provides a promising way to train a machine learning model among mobile devices without collecting their raw data to a central node. During training, proper devices are selected to participate in the training process to avoid model unfairness. In a mobile edge network, participant selection must be considered together with three factors: non-iid datasets possessed by devices, tunable local iterations on devices, and radio resource allocation to counter the impact of time-varying channel conditions on parameter transmissions. Since datasets of devices are given, to ensure model fairness and achieve fast convergence in the FL training process, participants, local iterations, and radio resources must be scheduled jointly in each iteration of FL training. In this paper, the joint scheduling problem is analyzed and formulated. Since it is NP-hard, a heuristic scheduling method called PALORA is designed to conduct joint scheduling of participants, local iterations, and radio resources. PALORA consists of three sequentially interactive function blocks: 1) a pointer network embedded deep reinforcement learning method to select participants, 2) an estimation algorithm to determine the numbers of local iterations, and 3) a breadth-first search method to allocate radio resources to the selected participants. PALORA is evaluated via extensive simulations based on real-world datasets. Results show that it significantly outperforms benchmark approaches.

Voice interaction, as an emerging human-computer interaction method, has gained great popularity, especially on smart devices. However, due to the open nature of voice signals, voice interaction may cause privacy leakage. In this paper, we propose a novel scheme, calledSeVI, to protect voice interaction from being deliberately or unintentionally eavesdropped. SeVI actively generates jamming noise of superior characteristics, while a user is performing voice interaction with his/her device, so that attackers cannot obtain the voice contents of the user. Meanwhile, the device leverages the prior knowledge of the generated noise to adaptively cancel received noise, even when the device usage environment is changing due to movement, so that the user voice interactions are unaffected. SeVI relies on only normal microphone and speakers and can be implemented as light-weight software. We have implemented SeVI on a commercial off-the-shelf (COTS) smartphone and conducted extensive real-world experiments. The results demonstrate that SeVI can defend both online eavesdropping attacks and offline digital signal processing (DSP) analysis attacks.

A coordinated carrier sense multiple access (C-CSMA) scheme is developed to schedule packet transmissions in multihop wireless networks. It integrates distributed scheduling and centralized coordination at different time scales. At a small time scale, a distributed scheduling algorithm, which is incorporated into CSMA, runs on each node to determine key parameters for CSMA. These parameters are optimized by ensuring multiple network attributes, i.e., average link date rate, node transmission probability, and the conditional probability of a link being selected for transmission, approaching their corresponding optimal values at a large time scale, e.g., a few seconds. Such optimal values are obtained from a centralized algorithm running on a portal node. The algorithm collects topology, link, and traffic information from the network at a large time scale and determines the attributes above to fulfill the optimal tradeoff between throughput and fairness. Since the distributed algorithm is shepherded by the centralized algorithm, C-CSMA has high scalability: first, it achieves long-term optimal performance with low information collection overhead; second, it schedules packet transmissions in a distributed way and responds quickly to changes in networks. It is proved that network attributes resulting from the distributed scheduling algorithm converge to optimal values provided by the centralized algorithm. Simulation experiments demonstrate that the convergence speed is fast. Moreover, extensive simulation results show that C-CSMA achieves much higher throughput, better fairness, and lower delay than existing scheduling schemes under different carrier sensing thresholds and traffic conditions.

Carrier-free pulse-based waveforms are desired in terahertz (THz) communications since pulse-based systems have simpler transceiver architectures than carrier-based systems. For such a pulse-based THz communication system, there still lacks pulse waveforms that can support high-order modulation and flexible multiple access. In this paper, a pulse-based waveform with continuously tunable center frequencies and bandwidths is designed for carrier-free THz communications. More specifically, a Gaussian pulse is utilized as the basic pulse, and then the weighted sum of its high-order derivatives is used to generate waveforms with tunable frequencies and bandwidths according to the probability density function of Rice distribution. Such a pulse-based waveform is called frequency and bandwidth continuously tunable (FBCT) pulse. By making the derivative orders of all weighted terms odd or even, a pair of orthogonal FBCT pulses can be generated. Based on the pair of orthogonal FBCT pulses, a basic pulse-based$M$-ary quadrature amplitude modulation ($\text{P}M$QAM) scheme can be readily achieved. Moreover, with frequency and bandwidth tunability, FBCT pulse can support pulse division multiple access (PDMA) with tunable bandwidth, through which frequencies with high molecular absorption loss can be avoided. Numerical results demonstrate that FBCT pulse is significantly effective and flexible in supporting carrier-free THz communications.

Backscatter communication is visioned as one of the promising technologies for future ultra-low power Internet of Things (IoT). The orthogonal-frequency-division-multiplexing physical-layer (OFDM-PHY) WiFi backscatter communications attract great attention in recent years. However, the severe tag transmission collisions highly degrade the system performance in a backscatter network, since complicated multiple access mechanisms, e.g., carrier sense multiple access, cannot be applied on backscatter tags. To address this problem, a novel parallel decoding design called FreeCollision is developed to enable concurrent OFDM-PHY WiFi backscatter communications. Without the prior knowledge of the number of collided tags, their channel state information, and modulation types, FreeCollsion can just use the I-Q symbols to resolve the collision by the design of a series of mechanisms: collided constellation recovery, concurrent virtual channel estimation, QPSK tag detection, and parallel demodulation. Simulation results verify the effectiveness of our proposed scheme. The successful decoding rate is more than 95% for 4 collided tags with BPSK modulation. The maximum successful transmission probability can be improved from 36.7% to 88% for slotted Aloha.

Carrier-free pulse-based waveforms are desired in terahertz (THz) communications, since pulse-based systems have simpler transceiver architectures than carrier-based systems. Nowadays, Gaussian pulses and higher time order derivative (HTOD) Gaussian pulses are commonly used in pulse-based communications. However, existing Gaussian pulses span a large consecutive spectrum, so the communication distance is constrained by those frequencies with high molecular absorption loss. HTOD Gaussian pulses can avoid the frequency band of molecular absorption peak by tuning its center frequency, but the center frequency cannot be adjusted continuously. To resolve these issues, a pulse-based waveform with continuously tunable center frequencies and bandwidths is designed for carrier-free THz communications. More specifically, a Gaussian pulse is utilized as the basic pulse, and then the weighted sum of its high-order derivatives is used to generate waveforms with tunable frequencies and bandwidths according to the probability density function of Rice distribution. In this paper, such a pulse-based waveform is called frequency and bandwidth continuously tunable (FBCT) pulse. Moreover, with frequency and bandwidth tunability, FBCT pulse can support pulse division multiple access (PDMA) with tunable bandwidth, through which frequencies with high molecular absorption loss can be avoided. The basic mechanisms of multiple access based on FBCT pulse are analyzed. Numerical results demonstrate that FBCT pulse is significantly effective and flexible in supporting carrier-free THz communications.

Intelligent vehicles require both communications and radar sensing. Recently, the development of joint communication and automotive radar sensing has attracted great attention. In this paper, the reference signal (RS) in the orthogonal frequency division multiplexing (OFDM) waveform is leveraged for radar sensing in 5G vehicle-to-everything (V2X) communications. Unlike the existing studies that use the whole subcarriers as pilot subcarriers for radar sensing, the scattered RS pattern adopted in the 5G system is considered in our design. More specifically, the scattered RS placement and power allocation between RS and data are optimized to minimize the total transmission power for energy-efficient joint signal transmission, while satisfying both the communication and radar sensing requirements. The optimization problem is a mixed-integer non-linear programming (MINLP) problem. The optimal solution is solved by enumerating all possible RS placements with the optimal allocated power by solving a transformed convex optimization problem. Such an approach suffers a high computing complexity when there are a large number of possible RS placements. Therefore, an alternative heuristic approach is further developed to resolve the problem via successive convex approximation (SCA) method. Simulation results show that the heuristic method can well approach the optimal solution. Compared with existing studies, our proposed scheme can effectively improve energy efficiency.

One-hop multicasting (OHM) of high-volume sensor data is essential for cooperative autonomous driving applications. While millimeter-Wave (mmWave) bands can be utilized for high-bandwidth OHM data transmission, it is very challenging for individual vehicles to find and communicate with a proper neighbor in a fully distributed and highly dynamic scenario. In this paper, we propose a fully distributed OHM scheme in vehicular networks, called mmV2V, which consists of three highly integrated protocols. Specifically, synchronized vehicles first conduct a probabilistic neighbor discovery procedure, in which randomly divided transmitters (or receivers) clockwise scan (or listen to) the surroundings in pace with heterogeneous Tx (or Rx) beams. In this way, the vast majority of neighbors can be identified in a few repeated rounds. Furthermore, vehicles negotiate with each of their neighbors about the optimal communication schedule in evenly distributed slots. Finally, each agreed pair of neighboring vehicles start high data rate transmissions with refined beams. We conduct extensive simulations and the results demonstrate that mmV2V can achieve a high completion ratio in rigid OHM tasks under various traffic conditions.

It remains a challenging issue to localize blocked RFID objects. Existing solutions rely on non-line-of-sight (NLOS) signals reflected from environments, which are not reliable and cannot be accurately measured. To eliminate such limitations, this paper develops a new approach (called MetaSight) that localizes blocked objects by modulating NLOS signals via metasurfaces. More specifically, a programmable metasurface is designed such that each reflecting element can dynamically change frequencies and phases of reflected signals. With proper setting of frequencies and phases within a certain range of reflecting elements (i.e., the reflecting window) by a reflection beamforming algorithm, a unique NLOS signal path (called metasurface path or MS path) is established between an object and its reader. By shifting the reflecting window in time domain, multiple MS paths and the corresponding channel coefficients can be obtained sequentially. To make the sequential process fast, both the preamble and the payload of a frame are utilized for channel estimation. Based on the estimated channel coefficients of multiple MS paths, the 3D direction of the object viewed from the metasurface is derived through a 3D direction estimation algorithm. By consolidating the 3D directions from two separate metasurfaces, the object's 3D location is finally determined. MetaSight is implemented as a prototype system. Extensive experiments show that MetaSight can localize a blocked object with an average accuracy of 15 cm.

Localizing RFID-tagged objects by a mobile robot plays an important role in many Internet of Things (IoT) applications. Existing RFID localization systems are infeasible, since they either demand bulky RFID infrastructures or cannot achieve sufficient localization accuracy. In this article, a portable localization (POLO) system is developed for a mobile robot to locate RFID-tagged objects. POLO consists of an RFID reader, a tag array, and a lightweight receiver. The reader is used for interrogating the RFID tag on an object. The tag array is designed to reflect the RFID signal from an object into multipath signals. The receiver captures such signals and estimates their multipath channel coefficients by a tag-array-assisted channel estimation (TCE) mechanism. Such channel coefficients are further exploited to determine the object’s direction by a spatial smoothing direction estimation (SSDE) algorithm. To resist the impact of multipath reflections from surroundings, more spatial information is exploited by placing the tag elements densely and collecting the channel coefficients during the robot’s movement. Based on the object’s direction, POLO guides the robot to approach the object. When the object is in proximity, its 3-D location is finally determined by a near-range positioning (NRP) algorithm. Moreover, POLO is designed to be compatible with commercial RFID systems. POLO is prototyped and evaluated via extensive experiments. Results show that the average angular error is within 1 degree when the object is in the far range (2–6 m), and the average location error is within 6 cm while the object is in the near range (~1 m).

Considering network slicing in a cellular network, one of the most intriguing tasks is slice enforcement over air interfaces across multiple cells. The challenges lie in several aspects. First, resources allocated to different slices must achieve soft isolation at the link level. Second, users’ diverse QoS requirements must be satisfied even when communication links experience fading and interference. Third, long-term slicing policies must be conformed, no matter how unbalanced they are. To address these challenges, link-level slice enforcement is first formulated as a resource allocation problem that minimizes radio resource consumption while ensuring link-level soft slice isolation, guaranteeing users’ diverse QoS requirements, and conforming to slicing policies. Next, this problem is tackled via a deep reinforcement learning (DRL) based approach, through which LinkSlice is designed as an iterative two-stage algorithm. The first stage determines transmission rates for each link based on DRL. It is embedded with a graph neural network (GNN) to characterize link interference. Based on the transmission rates from the first stage, the second stage allocates resources to each slice. Performance results show that LinkSlice converges quickly to a near-optimal solution. It gracefully tackles the three challenges of link-level slice enforcement while further improving throughput by 18.5%.

In smartphone data analysis, both energy consumption modeling and user behavior mining have been explored extensively, but the relationship between energy consumption and user behavior has been rarely studied. Such a relationship is explored over large-scale users in this article. Based on energy consumption data, where each users’ feature vector is represented by energy breakdown on hardware components of different apps, User Behavior Models (UBM) are established to capture user behavior patterns (i.e., app preference, usage time). The challenge lies in the high diversity of user behaviors (i.e., massive apps and usage ways), which leads to high dimension and dispersion of data. To overcome the challenge, three mechanisms are designed. First, to reduce the dimension, apps are ranked with the top ones identified as typical apps to represent all. Second, the dispersion is reduced by scaling each users’ feature vector with typical apps to unit ℓ 1 norm. The scaled vector becomes Usage Pattern, while the ℓ 1 norm of vector before scaling is treated as Usage Intensity. Third, the usage pattern is analyzed with a two-layer clustering approach to further reduce data dispersion. In the upper layer, each typical app is studied across its users with respect to hardware components to identify Typical Hardware Usage Patterns (THUP). In the lower layer, users are studied with respect to these THUPs to identify Typical App Usage Patterns (TAUP). The analytical results of these two layers are consolidated into Usage Pattern Models (UPM), and UBMs are finally established by a union of UPMs and Usage Intensity Distributions (UID). By carrying out experiments on energy consumption data from 18,308 distinct users over 10 days, 33 UBMs are extracted from training data. With the test data, it is proven that these UBMs cover 94% user behaviors and achieve up to 20% improvement in accuracy of energy representation, as compared with the baseline method, PCA. Besides, potential applications and implications of these UBMs are illustrated for smartphone manufacturers, app developers, network providers, and so on.

To meet the tremendous demand of Internet of Things (IoT) applications, the 3rd generation partnership project (3GPP) has specified the narrowband IoT (NB-IoT) standard. However, collisions in the random access (RA) channel of NB-IoT can be severe due to the mismatch between frequent random access attempts by a huge number of devices and the limited radio resources. In this paper, a new random access mechanism called time-alignment-value based random access (TARA) is designed to improve the efficiency of random access of NB-IoT. The key mechanism of TARA is to conduct quick retries upon failure by exploiting the difference of time-alignment (TA) values in collided preambles. To analyze TARA rigorously, a theoretical model of TARA is derived, and its validity is verified via simulations. Further simulations are carried out to evaluate TARA with respect to different system parameters. Comparisons with existing schemes are also conducted. Both analytical and simulation results show that, under various system parameters, TARA achieves 30 percent higher success probability of random access, 75 percent higher throughput, and 40 percent lower access delay than the original slotted Aloha mechanism of NB-IoT. It also highly outperforms other random access schemes.

In a 5G integrated backhaul and access (IAB) network, an IAB-donor and multiple IAB-nodes form a multi-hop wireless backhaul network. When a terminal is connected to an IAB-node, the traffic flows generated by the terminal can be backhauled to the IAB-donor. Due to the mobility of terminals, access links are error-prone and cause difficulty in ensuring end-to-end quality of service (QoS) of traffic flows. To improve access link reliability, multi-connectivity is considered in IAB, i.e., a terminal is connected to multiple IAB-nodes. However, to ensure end-to-end QoS, multi-connectivity must be considered together with multi-hop backhauling. Thus, an effective-capacity based resource allocation (eReal) scheme is developed to establish end-to-end multi-connectivity. The objective of the scheme is to guarantee end-to-end QoS and ensure outage probability is below the given threshold. Since traffic flows are classified into the guaranteed bit rate (GBR) and the non-GBR types, the scheme is formulated as two joint route selection and resource allocation problems to provide differentiated services with minimum resource consumption. Since both problems are NP-hard, they are solved using column generation. Performance results show that eReal serves various traffic flows with over 95% QoS guarantees and also significantly outperforms existing schemes.

Coded caching has become a hot research topic in recent years. However, existing studies rarely consider the delay constrained transmissions over wireless fading channels. For strict delay constrained content delivery over wireless fading channels, power optimization is usually required to achieve energy-efficient transmissions. Moreover, in coded caching, since the data to different users is transmitted via many coded multicast transmissions, the scheduling of these transmissions over fading channels is also critical for minimizing energy consumption. To this end, a joint scheduling and power optimization problem is formulated to minimize the expected energy consumption over wireless fading channels under strict delay constraints. Causal channel state information (CSI) is considered in this paper for practical cases, which nonetheless makes the optimal solution to this problem hard to achieve due to the uncertain future channel states. Therefore, a heuristic approach is developed to solve this problem. The number of slots for each coded multicast transmission is first determined by obtaining the optimal result of a nonlinear integer programming (NLIP) problem based on statistical channel estimations. Next, a closed-form inverse-waterfilling algorithm is carried out to allocate power for the coded transmissions under delay constraints, and then the one that saves the maximum energy consumption is scheduled for transmission. The heuristic approach is proved to achieve a result that is upper-bounded by a parameter times the optimal result of the original problem under non-causal CSI. Simulation results further show that compared to the benchmark with only slot allocation for each coded transmission, our approach with joint scheduling and power optimization can significantly save energy consumption while ensuring strict delay constraints.

In many Internet-of-Things (IoT) applications, various RFID-tagged objects need to be localized by mobile robots. Existing RFID localization systems are infeasible, since they either demand bulky RFID infrastructures or cannot achieve sufficient localization accuracy. In this paper, a portable localization (POLO) system is developed for a mobile robot to locate RFID-tagged objects. Besides a single RFID reader on board, POLO is distinguished with a tag array and a lightweight receiver. The tag array is designed to reflect the RFID signal from an object into multi-path signals. The receiver captures such signals and estimates their multi-path channel coefficients by a tag-array-assisted channel estimation (TCE) mechanism. Such channel coefficients are further exploited to determine the object's direction by a spatial smoothing direction estimation (SSDE) algorithm. Based on the object's direction, POLO guides the robot to approach the object. When the object is in proximity, its 2D location is finally determined by a near-range positioning (NRP) algorithm. POLO is prototyped and evaluated via extensive experiments. Results show that the average angular error is within 1.6 degrees when the object is in the far-range (2~6 m), and the average location error is within 5 cm while the object is in the near-range (~1 m).

Edge computing is promising to provide computational resources for connected vehicles. Resource demands for edge servers vary due to vehicle mobility. It is then challenging to reserve edge servers to meet variable demands. Existing schemes rely on statistical information of resource demands to determine edge server reservation. They are infeasible in practice, since the reservation based on statistics cannot adapt to time-varying demands. In this paper, a spatio-temporal reinforcement learning scheme called DeepReserve is developed to learn variable demands and then reserve edge servers accordingly. DeepReserve is adapted from the deep deterministic policy gradient algorithm with two major enhancements. First, by observing that the spatio-temporal correlation in vehicle traffic leads to the same property in resource demands of CVs, a convolutional LSTM network is employed to encode resource demands observed by edge servers for inference of future demands. Second, an action amender is designed to make sure an action does not violate spatio-temporal correlation. We also design a new training method, i.e., DR-Train, to stabilize the training procedure. DeepReserve is evaluated via experiments based on real-world datasets. Results show it achieves better performance than state-of-the-art approaches that require accurate demand information.

Coded caching can achieve significant caching gain by leveraging coded multicast for content delivery, which attracts great attentions in recent years. However, the coded multicast delivery design for decentralized coded caching with heterogeneous cache sizes and link rates is still an open problem. To this end, joint-cache-channel (JCC) coding is exploited in this paper. More specifically, a novel coded multicast delivery scheme that explores JCC coding is developed for the practical case with a finite number of packets per file. Instead of directly constructing the conflict graph at each packet level in existing studies, a packet-merging mechanism and a color-merging mechanism are developed to achieve the conflict graph in packet cluster level for coded multicast delivery so as to harvest the performance gain of JCC coding. The information-theoretic analysis of the transmission time via JCC coding is further conducted in the asymptotic regime when the number of packets per file approaches infinity, which provides the theoretical coded caching gain by JCC coding. Simulation results further show that compared to the existing method, the transmission time can be effectively reduced by our proposed scheme.

High-accuracy localization technology has gained increasing attention in gesture and motion control and many diverse applications. Due to multi-path fading and blockage effects in indoor propagation, 0.1m-level precise localization is still challenging. Promising for 6G wireless communications, the Terahertz (THz) spectrum provides multi-GHz ultra-broad bandwidth. Applying the THz spectrum to indoor localization, the channel state information (CSI) of THz signals, including angle of arrival (AoA), received power, and delay, has unprecedented resolution that can be explored for positioning. In this paper, a Structured Intra-Attention Bidirectional Recurrent (SIABR) deep learning method is proposed to solve the CSI-based three-dimensional (3D) THz indoor localization problem with significantly improved accuracy. As a two-level structure, the features of individual multi-path rays are first analyzed in the recurrent neural network with the attention mechanism at the lower level. Furthermore, the upper-level residual network (ResNet) of the constructed SIABR network extracts hidden information to output the geometric coordinates. Simulation results demonstrate that the 3D localization accuracy in the metric of mean distance error is within 0.25m. The developed SIABR network has very fast convergence and is robust against THz indoor line-of-sight blockage, multi-path fading, channel sparsity and CSI estimation error.

How to achieve efficient multiple access for Terahertz (THz) networks is still an open problem. The key obstacle is the extremely narrow angular coverage by a highly directional THz beam. Thus, a hybrid THzPrism beamforming (HTB) architecture is designed in this paper to greatly enlarge angular coverage by spreading frequency subcarriers to different directions. Based on the HTB architecture, a spatial-spread orthogonal frequency division multiple access (SS-OFDMA) scheme is developed. To serve users dispersed in a large angular range, a user grouping mechanism is first designed for SS-OFDMA to utilize SDMA and suppress the inter-group interference. To improve the spectrum and energy efficiency for sporadic users within a group, a non-uniform beam spreading mechanism is then developed for SS-OFDMA by joint design of digital and analog beamforming. Finally, a resource allocation algorithm is designed to minimize the transmit power of the base station by optimizing the subarray allocation among groups as well as the subcarrier and power allocation for each user within a group. Compared with the existing schemes, SS-OFDMA increases the achievable data rate by up to 124%, reduces the power consumption by up to 71%, and increases the average number of concurrently served users by up to 147%.

Container-based operation system (OS) level virtualization has been adopted by many edge-computing platforms. However, for an edge server, inter-container communications, and container management consume significant CPU resources. Given an application composed of interdependent tasks, the number of such operations is closely related to the dependency between the scheduled tasks. Thus, to improve the execution efficiency of an application in an edge server, task scheduling and task containerizing need to be considered together. To this end, a joint task scheduling and containerizing (JTSC) scheme is developed in this article. Experiments are first carried out to quantify the resource utilization of container operations. System models are then built to capture the features of task execution in containers in an edge server with multiple processors. With these models, joint task scheduling and containerizing is conducted as follows. First, tasks are scheduled without considering containerization, which results in initial schedules. Second, based on system models and guidelines gained from the initial schedules, several containerization algorithms are designed to map tasks to containers. Third, task execution durations are updated by adding the time for inter-container communications, and then the task schedules are updated accordingly. The JTSC scheme is evaluated through extensive simulations. The results show that it reduces inefficient container operations and enhances the execution efficiency of applications by 60 percent.

In recent years, crowdsensing has received extensive attention both in academia and industry. However, most of the prior works are based on centralized framework, where the users upload the sensor data to a platform alone. In this paper, we propose a decentralized data aggregation algorithm for crowdsensing, in which, participants negotiate the average of their sensor data in a pure distributed manner while enjoying differential privacy guarantee. Taking participants' privacy into consideration, we first redesign a distributed averaging algorithm by artificially adding random noise to accommodate the mobile crowdsensing scenario, where the sensor data may be sensible. We prove that, though random noises are involved, our algorithm almost surely yields an unbiased estimate of the exact average. We further theoretically formulate the trade-offs between differential privacy and the accuracy of the algorithm. Finally, we give the optimal choice for the additive noise with respect to the variance minimization of the estimate. The extensive simulations and realword test demonstrate the effectiveness of our algorithm.

The IEEE 802.11ad based joint communication and radar sensing has attracted great attentions for vehicles in recent years. The existing studies all assume full duplex communications between the transmitter and radar receiver based on perfect self-interference cancellation. However, the self-interference may not be fully cancelled due to the limitation of self-interference cancellation capability in practical cases, which will significantly degrade the sensing capability of the radar function, especially for the detection range. In this paper, the imperfect self-interference cancellation is considered and a novel joint communication and automotive long range radar sensing design is proposed based on OFDM frame structure in 802.11ad standard. The received signal model in the frequency domain synchronized with the self-interference is derived, in which the target reflection signal suffers inter-carrier-interference (ICI) and inter-symbol-interference (ISI). However, we show that the ISI can be leveraged for enhancing radar parameter estimation. Based on the received signal model, a novel pilot signal design is first developed to combat the self-interference for accurate velocity and coarse range estimation. Then, a few self-interference-free OFDM symbols at the end of the data frame are utilized to achieve accurate range estimation. Simulation results show that the decimeter-per-second level velocity estimation and centimeter level range estimation can be achieved for up to 200-meter radar sensing.

High-accuracy localization technology has gained increasing attention in gesture and motion control and many diverse applications. Due to the shadowing, multi-path fading, blockage effects in indoor propagation, 0.1m-level precise localization is still challenging. Promising for 6G wireless communications, the Terahertz (THz) spectrum provides ultra-broad bandwidth for indoor applications. Applying to indoor localization, the channel state information (CSI) of THz wireless signals, including angle of arrival (AoA), received power, and delay, has unprecedented resolution that can be explored for positioning. In this paper, a Structured Bidirectional Long Short-term Memory (SBi-LSTM) recurrent neural network (RNN) architecture is proposed to solve the CSI-based three-dimensional (3D) THz indoor localization problem with significantly improved accuracy. As a two-level structure, the features of individual multi-path ray are first analyzed in the Bi-LSTM network at the base level. Furthermore, the upper level residual network (ResNet) of the constructed SBi-LSTM network extracts for the geometric coordinates. Simulation results validate the convergence of our SBi-LSTM method and the robustness against indoor non-line-of-sight (NLoS) blockage. Specifically, the localization accuracy in the metric of mean distance error is within 0.27m under the NLoS environment, which demonstrates 60% enhancement over the state-of-the-art techniques.

Voice interaction, as an emerging human-computer interaction method, has gained great popularity, especially on smart devices. However, due to the open nature of voice signals, voice interaction may cause privacy leakage. In this paper, we propose a novel scheme, called SeVI, to protect voice interaction from being deliberately or unintentionally eavesdropped. SeVI actively generates jamming noise of superior characteristics, while a user is performing voice interaction with his/her device, so that attackers cannot obtain the voice contents of the user. Mean-while, the device leverages the prior knowledge of the generated noise to adaptively cancel received noise, even when the device usage environment is changing due to movement, so that the user voice interactions are unaffected. SeVI relies on only normal microphone and speakers and can be implemented as light-weight software. We have implemented SeVI on a commercial off-the- shelf (COTS) smartphone and conducted extensive real-world experiments. The results demonstrate that SeVI can defend both online eavesdropping attacks and offline digital signal processing (DSP) analysis attacks.

Mobile crowdsening (MCS) recently has been regarded as a newly-emerged sensing paradigm, which consists of a centralized cloud-based platform, some data demanders and some workers. However, since the platform needs to process tons of sensory data which is requested by the demanders and submitted by a large number of workers, this traditional MCS system may cause a heavy latency and serious network congestion, and thus can not be employed to some real-time and large-scale applications. Therefore, a new MCS system has been proposed recently by combining the edge computing, where some edge nodes (e.g., the base stations, laptops, smartphones) are added between the platform and workers to offload some operations from the platform to the network edges. Similar to the traditional MCS system, since participating in such edge-assisted MCS system is costly, it is important to attract more participation. However, this is more difficult since the platform and workers do not have direct communications with each other, which makes the edge nodes arbitrarily manipulate the information they transfer. Therefore, unlike the prior arts, we propose a novel incentive mechanism, namely, CHASTE, for such edge-assisted MCS system consisting of multiple demanders, some edge nodes and a crowd of workers where all of them behave strategically to maximize their own utility. Specifically, CHASTE is able to stimulate the participation, and satisfies the three-party truthfulness, three-party individual rationality, budget balance, as well as high social welfare. The desirable properties of CHASTE are validated through both theoretical analysis and extensive simulations.

Exploiting the computing capability of mobile devices with specialized engines (e.g., Neural Engine in iPhone), an attractive paradigm of federated learning that combines the mobile crowdsensing (MCS) has been deeply investigated recently (e.g., Google AI and Nvidia), where the training task is offloaded to the mobile crowd. However, this new paradigm still has numerous problems. Since executing the training task is costly for individual workers, the first problem is how to attract more participants. Following the incentive requirement, the second is how to preserve the workers' bid privacy since the reported costs are usually sensitive. Finally, the third problem is to guarantee the privacy protection on locally training models in the federated learning which involve the private information of local data. In this paper, we propose an incentivized federated learning with differential privacy in MCS system, namely, SHIELD, to solve the three significant problems. In fact, SHIELD satisfies the truthfulness and individual rationality while preserving the differential privacy of workers' bids and locally training models. Furthermore, for accuracy, the excess empirical risk of 1 SHIELD is proved to be upper bounded by O((ln(Knmin))1/2/Knmin+ ln(Knmin)/K2nmin2), where a special case for totally distributed scenario leads to a much sharper bound O( log(n)/n2) than the latest result O(ln(mnmin)/m2nmin2). Finally, comparing with the state-of-art apmin proaches, SHIELD illustrates superior performance by numerous experiments in classification and regression tasks.

Integrated access and backhaul (IAB) networking is envisioned as a key technology to support more flexible and dense deployment of base stations (BSs). However, existing directed acyclic graph (DAG) based IAB networking highly limits the flexibility and efficiency of link scheduling, due to the fixed parent-to-child relation between two adjacent IAB nodes. In this paper, a mesh-architecture based approach is developed to improve the efficiency of IAB networking. The key idea of the mesh architecture is to make the relationship between two adjacent IAB nodes configurable. Based on the mesh architecture, the two-stage scheduling scheme of IAB networks is revised. The typical scenarios where the mesh-architecture based IAB networking outperforms the DAG-based one are analyzed. Simulation results show that the mesh-architecture based IAB networking can effectively improve the throughput by 6.70%40.56% and substantially reduce the delay under various traffic loads, as compared to the DAG-based IAB networking.

Existing physical layer security schemes are inapplicable to legacy devices operating on long coherence-time channels, since they demand changes in the physical layer. To this end, a new physical-layer security scheme is developed to secure legacy clients. In this scheme, secret keys are generated by a client and transmitted to the access point (AP). To protect these keys, a separated device called secrecy protector (SP) transmits jamming signals to prevent eavesdroppers from overhearing the keys. The SP is equipped with multiple antennas, each of which transmits an independent jamming stream with a pseudo-preamble. Since the SP can share jamming signals with the AP secretly, the AP can use a certain analog network coding scheme to remove jamming signals and decode the keys. In contrast, eavesdroppers have no knowledge of jamming signals to decode the keys. However, the long coherence-time channels are vulnerable to eavesdroppers, as they can guess the channel coefficients in a brute force way and then remove the jamming signals. Thus, frequency diversity is exploited to enhance the security of the system. Moreover, the design of jamming mechanisms resists collusion among eavesdroppers. Therefore, secret keys can be secretly shared between the AP and a client. The developed scheme is implemented on a software-defined radio platform. Performance results demonstrate that it can effectively deliver secure communications without any changes to the physical layer of legacy clients.

IEEE 802.11ax has adopted orthogonal frequency division multiple access (OFDMA) to support multi-user (MU) transmissions. There exist two uplink MU OFDMA access methods in IEEE 802.11ax. The first one is uplink OFDMA random access (UORA) in which stations randomly select resource units (RUs) to send physical protocol data units (PPDUs), so its access efficiency is low. The second one is uplink OFDMA nonrandom access (UONRA) in which the access point (AP) schedules MU transmissions based on buffer status reports (BSR) from stations. However, stations usually rely on UORA to send BSRs, so the low efficiency of UORA can be a bottleneck of UONRA. Thus, UORA needs to be renovated to improve the performance of itself and UONRA. To this end, a multi-dimensional busy-tone arbitration (MBTA) mechanism is developed in this paper to reduce collisions among stations contending the same RU. Since there lacks an algorithm in IEEE 802.11ax to support coexistence of UORA and UONRA, a dynamic access-method selection (DAMS) algorithm is designed for the AP and stations to choose an optimal access method. Both MBTA and DAMS are analyzed rigorously and are further validated via simulations. The analytical and simulation results show that: 1) The MBTA dramatically improves the access efficiency of UORA; 2) DAMS always achieves a higher throughput than both UORA and UONRA.

The joint considering of caching and coded multicasting can significantly improve caching gain, which has become a promising approach to address the explosive growth of wireless traffic demand. In this paper, the design of caching placement and coded multicasting for wireless backhaul networks is explored under heterogeneous file preferences, where the file preference of each small base station (BS) is assumed to be aware at the macro BS. To address the heterogeneous file preferences, a group-based caching and cooperative coded multicasting scheme is developed in this paper. By utilizing the spectral clustering method, the small BSs are first clustered into different groups with similar file preferences. Moreover, a cooperative caching strategy with symmetric file division is designed for each group, where the suboptimal caching proportion of the most popular files is achieved by an approximation analysis. By utilizing the group-based caching structure, an efficient greedy-based two-level cooperative coded multicasting algorithm is then developed. The proposed algorithm not only utilizes the coded multicasting opportunities within each group, but also explores the cooperative coding opportunities among groups. The effectiveness of our proposed scheme is verified by simulation results, which show that our proposed scheme can significantly reduce the traffic load compared to the existing schemes.

In this paper, a multi-user concurrent beam training scheme, named AnalogMUSIC, is designed for mmWave systems with a single radio frequency (RF) chain. In AnalogMUSIC, a one-RF chain based multiple signal classification (MUSIC) algorithm is developed for accurate direction estimation, where the training beams are switched in a time-division manner. To pursue a short training time, the training beams are switched at rough directions predefined by a coarse- resolution codebook. We prove that the traditional MUSIC algorithm can be achieved in the frequency domain. For orthogonal frequency division multiple access (OFDMA) systems, the multi-user training signals can be decoupled in the frequency domain. Thus, AnalogMUSIC achieves the concurrent beam training for multiple users by allocating different subcarriers to different users. To further adapt the number of switched training beams to various channel conditions, two coarse-resolution codebooks and a beam mapping mechanism are designed. Extensive simulations show that AnalogMUSIC can effectively improve beam training accuracy and substantially reduce delay overhead, compared with the existing schemes.

RFID-based passive positioning is essential for many Internet-of-Things applications. However, existing RFID-based passive positioning systems either have low precision or require multiple antenna arrays, which are bulky and expensive. In this paper, a 3D passive positioning scheme is developed based on RFID tag arrays. One of the tags can harvest the energy of the surrounding electromagnetic waves to power the embedded orientation sensors (an accelerometer and a magnetometer) and then send the orientation information of an object to an RFID reader. When an object is equipped with such RFID tag arrays, its direction can be estimated by exploring the phases of the signals reflected by the tags and the orientation of the object. Furthermore, based on the triangulation principle, the 3D location of an object can be determined by using multiple RFID tags on the object and two antennas at the RFID reader. Thus, compared with existing schemes, the required number of RFID antennas for 3D passive positioning is greatly reduced. The proposed scheme is implemented based on a COTS RFID platform. The experimental results show that the 90-percentile accuracy of the proposed system is within 9 centimeters.

To meet the tremendous demand of Internet of Things (IoT) applications, the 3rd Generation Partnership Project (3GPP) has specified the Narrowband IoT (NB-IoT) standard. However, collisions in the random access (RA) channel of NB-IoT can be severe due to mismatch between frequent random access attempts by a huge number of devices and the extremely limited radio resources. In this paper, a new random access mechanism called time-alignment-value based random access (TARA) is designed to increase the efficiency of the random access channel. The key idea of TARA is to conduct quick retries by an RA-failure device based on the time-alignment (TA) values of colliding UEs in a random access response (RAR) message. Simulation results show that TARA performs much better than the original slotted Aloha mechanism, i.e., higher success probability, higher throughput, and lower access delay. Comparison with other existing schemes further demonstrates high performance and advantages of TARA.

Visible light communication by light-emitting diodes (LEDs) has become one of the promising technologies to boost the capacity of mobile networks, due to the low cost and high energy efficiency of LED lamps and the vast unregulated visible light bandwidth. However, VLC is only suitable for downlink data transmissions from LED illumination infrastructures to mobile users. Moreover, due to the line-of-sight (LOS) transmission feature of visible light, the VLC downlink can be easily interrupted by blockage of obstacles or rotation of receivers, which leads to unreliable coverage of mobile users. To overcome the above two drawbacks of VLC, an integrated VLC and WiFi system is designed in this paper. In our proposed system, a 2.5 sublayer called link convergence (LC) layer between the IP layer and the data link layer is designed to integrate WiFi radio and VLC radio. The following key mechanisms are designed in the LC layer to support efficient communications of the integrated system: (a) VLC radio access; (b) VLC ARP table; (c) VLC selective ARQ; (d) VLC link maintenance; (e) handover between WiFi and VLC. Since the integration is above date link layer, no MAC driver or protocol modifications are needed for WiFi in our system. A prototype of our designed integrated system is implemented. Experiment results validate the feasibility and effectiveness of our proposed system.

Coded caching has emerged as a promising component of solutions to the exponential growth in network traffic. Previous approaches to network coding are all based on (simple) XOR coding, which is appropriate when links have the same rate. However, in typical wireless networks, different users experience different link rates. Thus, XOR coding is sub-optimal and cannot achieve full broadcast gain. In this paper, we consider the coded caching design for wireless networks with unequal link rates. More specifically, the backhaul networks of LTE-A or 5G system are considered, in which the link rates between the macroBS and microBSs are different. We leverage a new network coding scheme nested coded modulation (NCM) in the delivery phase and develop a novel file partition scheme for the placement phase based on unequal cache size allocation. This scheme adapts to unequal link rates for increasing broadcast gains. The achievable transmission time and the information-theoretic lower bound are derived; we show that transmission time of the NCM-based coded caching can achieve a constant gap to the lower bound. Moreover, the NCM-based coded caching can achieve significant performance improvement over the XOR-based coded caching; the example with normalized link rates from 1 to 6 achieves up to 250% throughput improvement. Numerical results also show that the NCM-based coded caching can well utilize the unequal link rates, which cannot be achieved by the XOR-based coded caching.

In this paper, a centralized coded caching scheme is developed for wireless backhaul networks with heterogeneous channel conditions, i.e., the wireless channels between microBSs and a macroBS have different link rates and packet loss rates. To address these heterogeneous channel conditions, a joint random linear network coding and nested coded modulation (RLNCNCM) encoding scheme is designed for the delivery phase. Based on RLNC-NCM encoding, broadcast opportunities can be fully utilized for each coded subfile transmission. To support RLNC-NCM transmission, unequal file partition and cache size allocation are conducted in the placement phase. It is shown that a microBS with weaker channel condition needs to be allocated with more cache size. The theoretical achievable transmission time of the RLNC-NCM coded caching scheme is also derived. Moreover, numerical results show that the coded caching scheme developed in this paper significantly reduces the transmission time as compared to the existing schemes.

Charging electric vehicles (EVs) at home is attractive to EV users. However, when the penetration level of EVs becomes high, a distribution grid suffers from problems such as under-voltage and transformer overloading. EV users also experience a fairness problem, i.e., the limited capacity is unfairly shared among EVs. To solve these problems, a physical fair-queueing framework is established for EV charging. In this framework, a distribution sub-grid is first mapped to a multi-server queueing system, and then a fluid-model based queueing scheme called physical multi-server generalized processor sharing (pMGPS) is designed. pMGPS ensures perfect fairness but cannot be used practically due to its nature of fluid model. To this end, a packetized scheme called physical start-time fair queueing (pSTFQ) is developed to schedule tasks of EV charging. The fairness performance of the pSTFQ scheduling scheme is characterized by the ratio of energy difference between pSTFQ and pMGPS. This critical performance metric is studied through theoretical analysis and is also evaluated via simulations. Performance results show that the pSTFQ scheduling scheme achieves an energy difference ratio of less than 4 percent in various scenarios without causing under-voltage and transformer overloading problems.

Recently, device-to-device (D2D) communication and full duplex communication are both considered key technologies for 5G networks. In this paper, two novel cooperative modes are developed for full duplex D2D communication underlaying cellular networks: the network MU-MIMO based mode (N-mode) and the sequential forwarding mode (S-mode). In the N-mode, two D2D users work as network MIMO to forward the data to cellular users and thus leverage the channel diversity. In the S-mode, the spatial distribution of D2D users and cellular users is explored to improve the transmission rate. Based on these modes, two D2D users share the downlink resources with two nearby cellular users simultaneously to achieve both proximity gain and reuse gain. Moreover, these modes are well suited for edge cellular users. To optimize the performance of the two modes, optimal power allocation is conducted by considering the influence of residual self-interference at full duplex radios and the requirements of the minimum transmission rate for both cellular users and D2D users. Simulation results show that, compared with the dedicated mode, the sum rate of cooperative modes in the perfect self-interference cancellation case can achieve 32% overall improvement and about 70% improvement in the scenario with non-edge D2D users and edge cellular users. With residual self-interference at a full duplex radio, the overall performance improvement is reduced to 15%. However, the cooperative modes can still achieve 40% improvement in the scenario with non-edge D2D users and edge cellular users.

Due to the scarcity of the licensed spectrum, Licensed-Assisted Access Long-Term Evolution (LAA- LTE) network can be deployed in unlicensed spectrum, which is currently occupied by different Wi-Fi systems. It is a very challenging problem to ensure fair coexistence between LAA-LTE and Wi-Fi networks, in terms of spectrum sharing and traffic management. To solve this problem, a Fair Downlink Traffic Management (FDTM) scheme is proposed in this paper for hybrid LAA-LTE/Wi-Fi networks. By using the genetic algorithm, FDTM tunes the minimum Contention Window (CW min ) values and assigns feasible weights for the LAA eNBs with different traffic loads, thus to achieve (1) fair spectrum sharing with the existing Wi-Fi networks in unlicensed bands, and (2) fair service differentiation for downlink LAA-LTE traffic. Numerical results show our FDTM scheme can guarantee the throughput of WiFi networks in shared unlicensed spectrum, while supporting proportional fairness for the LAA eNBs with different weights and CW min values.

Analog network coding (ANC) is effective in improving spectrum efficiency. To coordinate ANC among multiple nodes without relying on complicated scheduling algorithm and network optimization, a new random access MAC protocol, called ANC-ERA, is developed to dynamically form ANC-cooperation groups in an ad hoc network. ANC-ERA includes several key mechanisms to maintain high performance in medium access. First, network allocation vectors (NAV) of control frames are properly set to avoid over-blocking of channel access. Second, a channel occupation frame (COF) is added to protect vulnerable periods during the formation of ANC cooperation. Third, an ACK diversity mechanism is designed to reduce potentially high ACK loss probability in ANC-based wireless networks. Since forming an ANC cooperation relies on bi-directional traffic between the initiator and the cooperator, the throughput gain from ANC drops dramatically if bi-directional traffic is not available. To avoid this issue, the fourth key mechanism, called flow compensation, is designed to form different types of ANC cooperation among neighboring nodes of the initiator and the cooperator. Both theoretical analysis and simulations are conducted to evaluate ANC-ERA. Performance results show that ANC-ERA works effectively in ad hoc networks and significantly outperforms existing random access MAC protocols.

Compared to half duplex communications, full duplex communications can significantly improve link capacity. However, in a large scale wireless network such as a wireless mesh network, the capacity gain from full duplex communications has not been fully investigated. To this end, a metric of network capacity called transmission capacity is studied in this paper for a full duplex wireless network. It captures the maximum transmission throughput in a unit area, subject to a certain outage probability. The key challenge of deriving transmission capacity is to characterize the aggregate interference of the typical link in a full duplex wireless network, which is completely different from that in a half duplex wireless network. In this paper, stochastic geometry is employed to model the network topology as a Thomas cluster point process and then the aggregate interference is characterized as a shot-noise process. Based on these models, the transmission capacity is derived. Analytical results show that under the same network density the distribution of aggregate interference in a full duplex wireless network is more dispersed than that in a half duplex wireless network. Comparisons of transmission capacity between a full duplex network and a half duplex network reveal that the capacity gain from full duplex communications is limited due to severe aggregate interference. This result implies that self-interference cancellation alone cannot ensure scalable full duplex wireless networking.

The applicability of analog network coding (ANC) to a wireless network is constrained by several limitations: 1) some ANC schemes demand fine-grained frame-level synchronization, which cannot be practically achieved in a wireless network; 2) others support only a specific type of modulation or require equal frame size in concurrent transmissions. In this paper, a new ANC scheme, called restriction-free analog network coding (RANC), is developed to eliminate the above limitations. It incorporates several function blocks, including frame boundary detection, joint channel estimation, waveform recovery, circular channel estimation, and frequency offset estimation, to support random concurrent transmissions with arbitrary frame sizes in a wireless network with various linear modulation schemes. To demonstrate the distinguished features of RANC, two network applications are studied. In the first application, RANC is applied to support a new relaying scheme called multi-way relaying, which significantly improves the spectrum efficiency as compared to two-way relaying. In the second application, RANC enables random-access-based ANC in an ad hoc network where flow compensation can be gracefully exploited to further improve the throughput performance. RANC and its network applications are implemented and evaluated on universal software radio peripheral (USRP) software radio platforms. Extensive experiments confirm that all function blocks of RANC work effectively without being constrained by the above limitations. The overall performance of RANC is shown to approach the ideal case of interference-free communications. The results of experiments in a real network setup demonstrate that RANC significantly outperforms existing ANC schemes and achieves constraint-free ANC in wireless networks.

In this paper, we study energy-efficient transmit power control for multi-tier multi-antenna (MIMO) heterogeneous cellular networks (HetNets), where each tier operates in closed-access policy and base stations (BSs) in each tier are distributed as a stationary Poisson point process (PPP). Each BS serves multiple users at a fixed distance away from it. We first study noncooperative energy-efficient power control, where each tier selfishly chooses its transmit power to maximize its network energy efficiency (EE). This is modeled by a noncooperative power control game. We prove the existence and the uniqueness of the Nash equilibrium of the game. Moreover, we analyze the effects of circuit power and BS densities of the tiers on their transmit power at the Nash equilibrium. Then, we investigate cooperative energy-efficient power control, where all the tiers cooperatively choose their transmit power to optimize their network EE. This cooperative power control is formulated as a multiobjective problem. To obtain Pareto-optimal solutions of the problem, we develop an algorithm that alternately updates the transmit power of the tiers. In addition to the transmit power, the circuit power consumption of the operating BSs and their active antennas affects the network EE. Due to the circuit power, activating all the BSs and turning on all the antennas may not be optimal in maximizing the network EE. Motivated by this observation, we also develop energy-efficient BS activation control and antenna activation control schemes. Finally, we extend the analysis to the highest signal-to-interference-plus-noise ratio (SINR) association, where each user connects to the BS that offers the highest SINR among the BSs in its tier. Simulation results show that the proposed energy-efficient designs significantly improve the network EE at the cost of small spectral efficiency loss compared with the spectral-efficient designs.

Demand for Internet access from moving vehicles has been rapidly growing. Meanwhile, the overloading issue of cellular networks is escalating due to mobile data explosion. Thus, WiFi networks are considered as a promising technology to offload cellular networks. However, there pose many challenging problems in highly dynamic vehicular environments for WiFi networks. For example, connections can be easily disrupted by frequent handoffs between access points (APs). A scheme, called SWIMMING, is proposed to support seamless and efficient WiFi-based Internet access for moving vehicles. In uplink, SWIMMING operates in a “group unicast” manner. All APs are configured with the same MAC and IP addresses, so that packets sent from a client can be received by multiple APs within its transmission range. Unlike broadcast or monitor mode, group unicast exploits the diversity of multiple APs, while keeping all the advantages of unicast. To avoid possible collisions of ACKs from different APs, the conventional ACK decoding mechanism is enhanced with an ACK detection function. In downlink, a packet destined for a client is first pushed to a group of APs through multicast. This AP group is maintained dynamically to follow the moving client. The packet is then fetched by the client. With the above innovative design, SWIMMING achieves seamless roaming with reliable link, high throughput, and low packet loss. Testbed implementation and experiments are conducted to validate the effectiveness of the ACK detection function. Extensive simulations are carried out to evaluate the performance of SWIMMING. Experimental results show that SWIMMING outperforms existing schemes remarkably.

As full-duplex wireless communication evolves into a practical technique, it will be built into communication nodes in many application scenarios. However, it is difficult to do so for legacy communication nodes. Thus, full-duplex communication nodes will coexist with half-duplex communication nodes in the same application environment. In this paper, a wireless local area network with a full-duplex access point (AP) and half-duplex clients is studied, and a media access control (MAC) protocol called asymmetrical duplex (A-Duplex) is developed to support efficient coexistence between half-duplex clients and the full-duplex AP. A-Duplex explores packet-alignment-based capture effect to establish dual links between the AP and two different clients. In this way, the capability of a full-duplex AP can be utilized by half-duplex clients, which leads to much improved network throughput. Moreover, to ensure fairness of the MAC protocol, a virtual deficit round-robin algorithm is proposed for the AP to select appropriate half-duplex clients for dual-link setup. A-Duplex does not require any change in the physical layer of half-duplex clients; only an update of MAC driver is necessary. Thus, it is well suited for coexistence between half-duplex clients and a full-duplex AP. Both analysis and simulations are conducted to evaluate performance of A-Duplex. Results show that it improves the throughput by 48% and 188% and reduces the average packet delay by 26% and 22%, as compared to the IEEE 802.11 Distributed Coordination Function with and without RTS/CTS, respectively. Moreover, the throughput remains steady as the number of clients grows. A-Duplex also maintains a high level of fairness.

Full duplex wireless communications have become practical in recent years. However, there exist many communication nodes that only support half duplex communications. Thus, it is important to consider coexistence between full duplex and half duplex communications. In this paper, a wireless LAN with a full duplex AP and half duplex stations is studied, and a media access control(MAC) protocol is developed to ensure effective operation of such a wireless LAN. The MAC protocol explores capture effect to establish dual links between two different stations and the AP to improve network throughput. Furthermore, it does not need any change in the physical layer of half duplex stations; only an update of MAC driver is required. Thus, the MAC protocol is well suited for supporting legacy half duplex stations. Simulation results show that the new MAC protocol improve throughput by 39% and 180% as compared to the MAC protocol of 802.11 DCF with and without RTS/CTS, respectively. More interestingly, throughput performance of the new MAC protocol remains steady as the number of stations grows.

Maximum power point tracking (MPPT) achieves maximum power output for a photovoltaic (PV) system under various environmental conditions. It significantly improves the energy efficiency of a specific PV system. However, when an increasing number of PV systems are connected to a distribution grid, MPPT poses several risks to the grid: 1)over-voltage problem, i.e., voltage in the distribution grid exceeds its rating; and 2)reverse power-flow problem, i.e., power that flows into the grid exceeds an allowed level. To solve these problems, power point tracking of all PV systems in the same distributed grid needs to be coordinated via a communication network. Thus, coordinated power point tracking (CPPT) is studied in this paper. First, an optimization problem is formulated to determine the power points of all PV systems, subject to the constraints of voltage, reverse power flow, and fairness. Conditions that obtain the optimal solution are then derived. Second, based on these conditions, a distributed and practical CPPT scheme is developed. It coordinates power points of all PV systems via a communication network, such that: 1) voltage and reverse power flow are maintained at a normal level; and 2) each PV system receives a fair share of surplus power. Third, a wireless mesh network (WMN) is designed to support proper operation of the distributed CPPT scheme. CPPT is evaluated through simulations that consider close interactions between WMN and CPPT. Performance results show that: 1) CPPT significantly outperforms MPPT by gracefully avoiding both overvoltage and reverse power-flow problems; 2) CPPT achieves fair sharing of surplus power among all PV systems; and 3) CPPT can be reliably conducted via a WMN.

Broadcast is an important operation in wireless ad hoc networks where control information is usually propagated as broadcasts for the realization of most networking protocols. In traditional ad hoc networks, since the spectrum availability is uniform, broadcasts are delivered via a common channel which can be heard by all users in a network. However, in cognitive radio (CR) ad hoc networks, different unlicensed users may acquire different available channels depending on the locations and traffic of licensed users. This non-uniform channel availability leads to several significant differences and causes unique challenges when analyzing the performance of broadcast protocols in CR ad hoc networks. In this paper, a novel unified analytical model is proposed to address these challenges. Our proposed analytical model can be applied to any broadcast protocol with any CR network topology. We propose to decompose an intricate network into several simple networks which are tractable for analysis. We also propose systematic methodologies for such decomposition. Results from both the hardware implementation and software simulation validate the analysis well. To the best of our knowledge, this is the first analytical work on the performance analysis of broadcast protocols for multi-hop CR ad hoc networks.

A multi-antenna access point (AP) can communicate simultaneously with multiple clients, however, this multi-user MIMO (MU-MIMO) capability is underutilized in conventional 802.11 wireless LANs (WLANs). To address this problem, researchers have recently developed a CSMA/CA-based MAC protocol to support concurrent transmissions from different clients. In this paper, we propose an analytical model to characterize the saturation throughput and mean access delay of this CSMA/CA-based MAC protocol operating in an MU-MIMO WLAN. We also consider and model a distributed opportunistic transmission scheme, where clients are able to contend for the concurrent transmission opportunities only when their concurrent rates exceed a threshold. Comparisons with simulation results show that our analytical model provides a close estimation of the network performance. By means of the developed model, we evaluate the throughput and delay performance with respect to different network parameters, including the backoff window sizes, the number of AP's antennas, the network size, and the threshold of the opportunistic transmission scheme. Performance optimization over key parameters is also conducted for the transmission schemes.

In this paper, we deal with possible data transmission congestion on the sink node in wireless sensor networks (WSNs). We consider a scenario in which all the sensor nodes have a certain amount of storage space and acquire data from the surroundings at heterogeneous speed. Because receiving bandwidth of the sink node is limited, a proper bandwidth allocation mechanism should be implemented to avoid possible congestion or data loss due to the overflow of some sensor nodes. To address this problem, we firstly design a novel bandwidth allocation mechanism, SWM, that can maximize the social utility, an indicator of every sensor node's satisfaction degree and the social fairness. Furthermore, we model the allocation process under the SWM as a noncooperative game and figure out the unique Nash Equilibrium. The uniqueness of the equilibrium demonstrates that this network will actually approach to a fair and stable state.

The association policy in current IEEE 802.11 networks usually considers Received Signal Strength Indication (RSSI) to be the only metric to capture access link quality. However, when a Mesh Client (MC) in IEEE 802.11-based Wireless Mesh Network (WMN) needs to be associated with the most appropriate Mesh Access Point (MAP), the quality of both the access link and the routing path in mesh backhaul should be considered. To take into account this requirement, most existing approaches rely on MAPs to provide more information for MCs such as traffic load, routing metric, airtime cost, etc. These solutions inevitably need to modify the IEEE 802.11 standard or wireless interface drivers on both MAPs and MCs, which is not feasible or flexible in actual deployment. In this paper, a network-leading association scheme is proposed for IEEE 802.11 WMNs. It is completely operated by MAPs and does not require any modification on MCs. It also adapts to the dynamic network environment and always selects the best MAP to serve an MC. Simulation results on ns-3 platform indicate that the network-leading association scheme remarkably improves the performance of IEEE 802.11 WMNs as compared with the existing approaches.

Scheduling schemes of wireless networks determine packet transmission opportunities for each network node by considering network parameters such as link quality, transmission rate, and delay. However, security has not been taken into account in these schemes, although it plays a critical role in network performance. In this paper a new scheduling scheme is proposed to determine packet transmissions under the constraint of achieving perfect physical layer security. Incorporating this new scheduling scheme into the distributed coordination function (DCF) of IEEE 802.11 medium access control (MAC) leads to a secure MAC protocol called SecDCF. Simulations are conducted to evaluate the performance of SecDCF, and results illustrate that SecDCF significantly outperform IEEE 802.11 DCF when physical layer security is enforced.

Lunar surface networks are critical to a manned mission to the Moon, but encounter many challenges including lack of infrastructure, shortage of power, long distance communications, and quality of service (QoS) support. These challenges demand both power efficient and reliable networking protocols. In this paper a power efficient and reliable MAC protocol is proposed based on commercial IEEE 802.11 chipsets. It includes an innovative mechanism, called time controlled CSMA/CA (TC-CSMA/CA), to reduce power consumption of a communication device to the minimum. Additionally, due to time control, QoS of various applications is highly improved. Both simulations and testbed implementation are carried out to evaluate the performance of the proposed MAC protocol. Experimental results illustrate that the TC-CSMA/CA MAC is highly effective in reducing power consumption and increasing service reliability.

Random access is critical to OFDMA wireless networks. However, due to special features of an OFDMA system, how to carry out random access in an efficient manner is still an open issue for OFDMA wireless networks. In this paper, a new random access protocol is proposed for OFDMA wireless networks. It is distinguished by a novel mechanism called concurrent multi-channel carrier sense multiple access (CM-CSMA/CA). Simulation results show that the new random access protocol is highly efficient in uplink access and also significantly outperforms existing random access protocols for OFDMA wireless networks.

In this paper we exploit a setting for cognitive radio networks by utilizing cooperation from secondary users (SUs) to assist the transmissions of operators' primary users (PUs). On the other hand, SUs can share the spare spectrum of operators. Such a scenario can be viewed as a market where multiple operators trade their spare spectrum for the assistance of SUs, and multiple SUs trade the transmission energy for access opportunities from operators. We model the system using transferable payoff coalitional game theory. An outcome of a coalitional game is a specification of the coalition that forms and the joint action it takes. We show that the optimum joint action strategy can be obtained as a solution of convex optimization problem. Then, based on dual technique, we show that there is an operating point that maximizes the sum utility over the operators and SUs while providing each player a share such that no subset of operators and SUs has an incentive to break away from the brand coalition.

An infrastructure wireless mesh network (WMN) is a hierarchical network consisting of mesh clients, mesh routers and gateways. Mesh routers constitute a wireless mesh backbone, to which mesh clients are connected as a star topology, and gateways are chosen among mesh routers providing Internet access. In this paper, the throughput capacity of infrastructure WMNs is studied. For such a network with Nc randomly distributed mesh clients, Nr regularly placed mesh routers and Ng gateways, assuming that each mesh router can transmit at W bits/s, the per-client throughput capacity has been derived as a function of Nc , Nr , Ng and W . The result illustrates that, in order to achieve high capacity performance, the number of mesh routers and the number of gateways must be properly chosen. It also reveals that an infrastructure WMN can achieve the same asymptotic throughput capacity as that of a hybrid ad hoc network by choosing only a small number of mesh routers as gateways. This property makes WMNs a very promising solution for future wireless networking.

A frequency division duplex (FDD) wideband code division multiple access (CDMA) medium access control (MAC) protocol is developed for wireless wide area multimedia networks. In order to reach the maximum system capacity and guarantee the heterogeneous bit error rates (BERs) of multimedia traffic, a minimum-power allocation algorithm is first derived, where both multicode (MC) and orthogonal variable spreading factor (OVSF) transmissions are assumed. Based on the minimum-power allocation algorithm, a multimedia wideband CDMA generalized processor sharing (GPS) scheduling scheme is proposed. It provides fair queueing to multimedia traffic with different QoS constraints. It also takes into account the limited number of code channels for each user and the variable system capacity due to interference experienced by users in a CDMA network. To control the admission of real-time connections, a connection admission control (CAC) scheme is proposed, in which the effective bandwidth admission region is derived based on the minimum-power allocation algorithm. With the proposed resource management algorithms, the MAC protocol significantly increases system throughput, guarantees BER, and improves QoS metrics of multimedia traffic.

A wide-band time-division-code-division multiple-access (TD-CDMA) medium access control (MAC) protocol is introduced in this paper. A new minimum-power allocation algorithm is developed to minimize the interference experienced by a code channel such that heterogeneous bit-error rate (BER) requirements of multimedia traffic are satisfied. Further, from analysis of the maximum capacity of a time slot, it is concluded that both rate and BER scheduling are necessary to reach a maximum capacity. Based on the new minimum-power allocation algorithm as well as on rate and BER scheduling concepts, a new scheduling scheme is proposed to serve packets with heterogeneous BER and quality of service (QoS) requirements in different time slots. To further enhance the performance of the MAC protocol, an effective connection admission control (CAC) algorithm is developed based on the new minimum-power allocation algorithm. Simulation results show that the new wide-band TD-CDMA MAC protocol satisfies the QoS requirements of multimedia traffic and achieves high overall system throughput.

A medium access control (MAC) protocol is developed for wireless multimedia networks based on frequency division duplex (FDD) wideband code division multiple access (CDMA). In this protocol, the received power levels of simultaneously transmitting users are controlled by a minimum-power allocation algorithm such that the heterogeneous bit error rates (BERs) of multimedia traffic are guaranteed. With minimum-power allocation, a multimedia wideband CDMA generalized processor sharing (GPS) scheduling scheme is proposed. It provides fair queueing to multimedia traffic with different QoS constraints. It also takes into account the limited number of code channels for each user and the variable system capacity due to interference experienced by users in a CDMA network. The admission of real-time connections is determined by a new effective bandwidth connection admission control (CAC) algorithm in which the minimum-power allocation is also considered. Simulation results show that the new MAC protocol guarantees QoS requirements of both real-time and nonreal-time traffic in an FDD wideband CDMA network.