Multi-access point coordination (MAPC) is a key technology for enhancing throughput in next-generation Wi-Fi within dense overlapping basic service sets. However, existing MAPC protocols rely on static, protocol-defined rules, which limits their ability to adapt to dynamic network conditions such as varying interference levels and topologies. To address this limitation, we propose a novel Agentic AI Wi-Fi framework where each access point, modeled as an autonomous large language model agent, collaboratively reasons about the network state and negotiates adaptive coordination strategies in real time. This dynamic collaboration is achieved through a cognitive workflow that enables the agents to engage in natural language dialogue, leveraging integrated memory, reflection, and tool use to ground their decisions in past experience and environmental feedback. Comprehensive simulation results demonstrate that our agentic framework successfully learns to adapt to diverse and dynamic network environments, significantly outperforming the state-of-the-art spatial reuse baseline and validating its potential as a robust and intelligent solution for future wireless networks.

This paper presents Multimodal-Wireless, a large-scale open-source dataset for multimodal sensing and communication research. The dataset is generated through an integrated and customizable data pipeline built upon the CARLA simulator and Sionna framework, and features high-resolution communication channel state information (CSI) fully synchronized with five other sensor modalities, namely LiDAR, RGB and depth camera, inertial measurement unit (IMU) and radar, all sampled at 100 Hz. It contains approximately 160,000 frames collected across four virtual towns, sixteen communication scenarios, and three weather conditions. This paper provides a comprehensive overview of the dataset, outlining its key features, overall framework, and technical implementation details. In addition, it explores potential research applications concerning communication and collaborative perception, exemplified by beam prediction using a multimodal large language model. The dataset is open in https://le-liang.github.io/mmw/.

Judicious resource allocation can effectively enhance federated learning (FL) training performance in wireless networks by addressing both system and statistical heterogeneity. However, existing strategies typically rely on block fading assumptions, which overlook rapid channel fluctuations within each round of FL gradient uploading, leading to a degradation in FL training performance. Therefore, this paper proposes a small-scale-fading-aware resource allocation strategy using a multi-agent reinforcement learning (MARL) framework. Specifically, we establish a one-step convergence bound of the FL algorithm and formulate the resource allocation problem as a decentralized partially observable Markov decision process (Dec-POMDP), which is subsequently solved using the QMIX algorithm. In our framework, each client serves as an agent that dynamically determines spectrum and power allocations within each coherence time slot, based on local observations and a reward derived from the convergence analysis. The MARL setting reduces the dimensionality of the action space and facilitates decentralized decision-making, enhancing the scalability and practicality of the solution. Experimental results demonstrate that our QMIX-based resource allocation strategy significantly outperforms baseline methods across various degrees of statistical heterogeneity. Additionally, ablation studies validate the critical importance of incorporating small-scale fading dynamics, highlighting its role in optimizing FL performance.

Accurate channel state information (CSI) acquisition is essential for modern wireless systems, which becomes increasingly difficult under large antenna arrays, strict pilot overhead constraints, and diverse deployment environments. Existing artificial intelligence-based solutions often lack robustness and fail to generalize across scenarios. To address this limitation, this paper introduces a predictive-foundation-model-based channel estimation framework that enables accurate, low-overhead, and generalizable CSI acquisition. The proposed framework employs a predictive foundation model trained on large-scale cross-domain data to extract universal channel representations and provide predictive priors with strong cross-scenario transferability. A pilot processing network based on a vision transformer architecture is further designed to capture spatial, temporal, and frequency correlations from pilot observations. An efficient fusion mechanism integrates predictive priors with real-time measurements, enabling reliable CSI reconstruction even under sparse or noisy conditions. Extensive evaluations across diverse configurations demonstrate that the proposed estimator significantly outperforms both classical and data-driven baselines in accuracy, robustness, and generalization capability.

High-density Wi-Fi deployments often result in significant co-channel interference, which degrades overall network performance. To address this issue, coordination of multi access points (APs) has been considered to enable coordinated spatial reuse (CSR) in next generation wireless local area networks. This paper tackles the challenge of downlink spatial reuse in Wi-Fi networks, specifically in scenarios involving overlapping basic service sets, by employing hierarchical multi-agent reinforcement learning (HMARL). We decompose the CSR process into two phases, i.e., a polling phase and a decision phase, and introduce the HMARL algorithm to enable efficient CSR. To enhance training efficiency, the proposed HMARL algorithm employs a hierarchical structure, where station selection and power control are determined by a high- and low-level policy network, respectively. Simulation results demonstrate that this approach consistently outperforms baseline methods in terms of throughput, mean delay and delay jitter across various network topologies. Moreover, the algorithm exhibits robust performance when coexisting with legacy APs. Additional experiments in a representative topology further reveal that the carefully designed reward function not only maximizes the overall network throughput, but also improves fairness in transmission opportunities for APs in high-interference regions.

Semantic communication has been introduced into collaborative perception systems for autonomous driving, offering a promising approach to enhancing data transmission efficiency and robustness. Despite its potential, existing semantic communication approaches predominantly rely on analog transmission models, rendering these systems fundamentally incompatible with the digital architecture of modern vehicle-to-everything (V2X) networks and posing a significant barrier to real-world deployment. To bridge this critical gap, we propose CoDS, a novel collaborative perception framework based on digital semantic communication, designed to realize semantic-level transmission efficiency within practical digital communication systems. Specifically, we develop a semantic compression codec that extracts and compresses task-oriented semantic features while preserving downstream perception accuracy. Building on this, we propose a novel semantic analog-to-digital converter that converts these continuous semantic features into a discrete bitstream, ensuring integration with existing digital communication pipelines. Furthermore, we develop an uncertainty-aware network (UAN) that assesses the reliability of each received feature and discards those corrupted by decoding failures, thereby mitigating the cliff effect of conventional channel coding schemes under low signal-to-noise ratio (SNR) conditions. Extensive experiments demonstrate that CoDS significantly outperforms existing semantic communication and traditional digital communication schemes, achieving state-of-the-art perception performance while ensuring compatibility with practical digital V2X systems.

Despite the success of large language models (LLMs) across domains, their potential for efficient channel state information (CSI) compression and feedback in frequency division duplex (FDD) massive multiple-input multiple-output (mMIMO) systems remains largely unexplored yet increasingly important. In this paper, we propose a novel LLM-based framework for CSI feedback to exploit the potential of LLMs. We first reformulate the CSI compression feedback task as a masked token prediction task that aligns more closely with the functionality of LLMs. Subsequently, we design an information-theoretic mask selection strategy based on self-information, identifying and selecting CSI elements with the highest self-information at the user equipment (UE) for feedback. This ensures that masked tokens correspond to elements with lower self-information, while visible tokens correspond to elements with higher self-information, thus maximizing the accuracy of LLM predictions. Finally, the LLM leverages its robust modeling capabilities to reconstruct complete CSI representations through contextual inference. This self-information-driven masking strategy integrates the LLM-based masked token prediction mechanism into a coherent, information-driven framework. Numerical results indicate that the proposed LLM-based CSI feedback framework significantly outperforms traditional small models in CSI reconstruction accuracy, leading to substantial improvements in communication rates in multi-user MIMO scenarios. This approach has the potential to address the limitations of CSI reconstruction accuracy that restrict multi-user communication rates. Moreover, the method deploys a lightweight network at the UE, with additional network complexity overhead only at the base station (BS). Finally, the method demonstrates strong generalization across different compression ratios and exhibits excellent transfer learning capabilities across various channel scenarios. These findings pave the way for integrating LLMs into next-generation wireless communication systems.

In recent years, Text-to-Image (T2I) models have garnered significant attention due to their remarkable advancements. However, security concerns have emerged due to their potential to generate inappropriate or Not-Safe-For-Work (NSFW) images. In this paper, inspired by the observation that texts with different semantics can lead to similar human perceptions, we propose an LLM-driven perception-guided jailbreak method, termed PGJ. It is a black-box jailbreak method that requires no specific T2I model (model-free) and generates highly natural attack prompts. Specifically, we propose identifying a safe phrase that is similar in human perception yet inconsistent in text semantics with the target unsafe word and using it as a substitution. The experiments conducted on six open-source models and commercial online services with thousands of prompts have verified the effectiveness of PGJ.

This paper investigates the downlink channel estimation problem for frequency division duplex (FDD) multi-user massive multiple-input multiple-output (MIMO) systems. We model this problem within the sparse Bayesian learning (SBL) framework, where all unknowns are treated as random variables. Due to limited scattering at the base station, the channel exhibits sparsity in the angular domain. By introducing Gaussian mixture priors to characterize the user equipment internal sparsity and partially shared sparsity, we develop a Markov chain Monte Carlo (MCMC) method to implement Bayesian inference and accurately estimate all random variables in the model, including the channel matrix. Experimental results demonstrate that the MCMC-based SBL channel estimation algorithm outperforms existing approaches by over 5 dB in multi-user scenarios while reducing pilot overhead.

This paper investigates the challenge of heterogeneous multi-agent reinforcement learning (MARL) algorithms in wireless local area networks (WLANs), where multiple stations utilize either value-based or policy-based reinforcement learning algorithms for channel access. Specifically, we propose a novel heterogeneous MARL training framework, named QPMIX, which adopts a centralized training with decentralized execution paradigm to enable heterogeneous agents to collaborate. Our method aims to maximize the network throughput and ensure fairness among stations, enhancing the overall performance of WLANs. Through the simulation results, we demonstrate that the proposed QPMIX algorithm achieves higher throughput, lower mean delay, reduced delay jitter, and decreased collision rates than conventional CSMA/CA in the saturated traffic scenario. Additionally, it can better promote cooperation between heterogeneous agents compared to independent learning.

As a promising paradigm for the sixth-generation (6G) networks, task-oriented semantic communication significantly enhances transmission efficiency. However, it faces complex security challenges, particularly the risk of eavesdropping due to the open nature of wireless channels. To address this issue, we propose a secure semantic communication framework that integrates physical-layer secure beamforming (SBF) to safeguard semantic information from eavesdropping. Specifically, we design an SBF network to generate SBF vectors that focus signal beams on legitimate users to enhance signal power while directing designed artificial noise toward potential eavesdroppers to strengthen jamming. To further improve system adaptability across varying channel conditions, we introduce attention-based channel-aware modules that dynamically optimize the encoding, decoding, and beamforming processes based on perceived channel state information. Furthermore, task-oriented artificial noise is employed to degrade the task performance of eavesdroppers more effectively. Finally, a stepwise training strategy with task-specific loss functions is employed to jointly optimize the SBF and semantic modules, maximizing the performance gap in downstream tasks between legitimate users and eavesdroppers. The simulation results demonstrate that the proposed approach effectively maintains task performance for legitimate users while significantly suppressing eavesdroppers, outperforming conventional methods.

With the development of computer vision, 3D object detection has become increasingly important in many real-world applications. Limited by the computing power of sensor-side hardware, the detection task is sometimes deployed on remote computing devices or the cloud to execute complex algorithms, which brings massive data transmission overhead. In response, this paper proposes an optical flow-driven semantic communication framework for the stereo-vision 3D object detection task. The proposed framework fully exploits the dependence of stereo-vision 3D detection on semantic information in images and prioritizes the transmission of this semantic information to reduce total transmission data sizes while ensuring the detection accuracy. Specifically, we develop an optical flow-driven module to jointly extract and recover semantics from the left and right images to reduce the loss of the left-right photometric alignment semantic information and improve the accuracy of depth inference. Then, we design a 2D semantic extraction module to identify and extract semantic meaning around the objects to enhance the transmission of semantic information in the key areas. Finally, a fusion network is used to fuse the recovered semantics, and reconstruct the stereo-vision images for 3D detection. Simulation results show that the proposed method improves the detection accuracy by nearly 70% and outperforms the traditional method, especially for the low signal-to-noise ratio regime.

In this paper, we consider a radio resource management (RRM) problem in the dynamic wireless networks, comprising multiple communication links that share the same spectrum resource. To achieve high network throughput while ensuring fairness across all links, we formulate a resilient power optimization problem with per-user minimum-rate constraints. We obtain the corresponding Lagrangian dual problem and parameterize all variables with neural networks, which can be trained in an unsupervised manner due to the provably acceptable duality gap. We develop a meta-learning approach with graph neural networks (GNNs) as parameterization that exhibits fast adaptation and scalability to varying network configurations. We formulate the objective of meta-learning by amalgamating the Lagrangian functions of different network configurations and utilize a first-order meta-learning algorithm, called Reptile, to obtain the meta-parameters. Numerical results verify that our method can efficiently improve the overall throughput and ensure the minimum rate performance. We further demonstrate that using the meta-parameters as initialization, our method can achieve fast adaptation to new wireless network configurations and reduce the number of required training data samples.

Integrated sensing and communication (ISAC) in millimeter wave is a key enabler for next-generation networks, which leverages large bandwidth and extensive antenna arrays, benefiting both communication and sensing functionalities. The associated high costs can be mitigated by adopting a hybrid beamforming structure. However, the well-studied monostatic ISAC systems face challenges related to full-duplex operation. To address this issue, this paper focuses on a three-dimensional bistatic configuration that requires only half-duplex base stations. To intuitively evaluate the error bound of bistatic sensing using orthogonal frequency division multiplexing waveforms, we propose a positioning scheme that combines angle-of-arrival and time-of-arrival estimation, deriving the closed-form expression of the position error bound (PEB). Using this PEB, we develop two hybrid beamforming algorithms for joint waveform design, aimed at maximizing achievable spectral efficiency (SE) while ensuring a predefined PEB threshold. The first algorithm leverages a Riemannian trust-region approach, achieving superior performance in terms of SE and convergence speed compared to the conventional gradient-based methods, but with higher complexity. In contrast, the second algorithm, which employs orthogonal matching pursuit, offers a more computationally efficient solution, delivering reasonable SE while maintaining the PEB constraint. Numerical results are provided to validate the effectiveness of the proposed designs.

Device-to-device (D2D) spectrum sharing in wireless communications is a challenging non-convex combinatorial optimization problem, involving entangled link scheduling and power control in a large-scale network. The state-of-the-art methods, either from a model-based or a data-driven perspective, exhibit certain limitations such as the critical need for channel state information (CSI) and/or a large number of (solved) instances (e.g., network layouts) as training samples. To advance this line of research, we propose a novel hybrid model/data-driven spectrum sharing mechanism with graph reinforcement learning for link scheduling (GRLinQ), injecting information theoretical insights into machine learning models, in such a way that link scheduling and power control can be solved in an intelligent manner. Through an extensive set of experiments, GRLinQ demonstrates superior performance to the existing model-based and data-driven link scheduling and/or power control methods, with a relaxed requirement for CSI, a substantially reduced number of unsolved instances as training samples, a possible distributed deployment, reduced online/offline computational complexity, and more remarkably excellent scalability and generalizability over different network scenarios and system configurations.

Massive multiple-input multiple-output (MIMO) technology has significantly enhanced spectral and power efficiency in cellular communications and is expected to further evolve towards extra-large-scale MIMO. However, centralized processing for massive MIMO faces practical obstacles, including excessive computational complexity and a substantial volume of baseband data to be exchanged. To address these challenges, decentralized baseband processing has emerged as a promising solution. This approach involves partitioning the antenna array into clusters with dedicated computing hardware for parallel processing. In this paper, we investigate the gradient-based Markov chain Monte Carlo (MCMC) method—an advanced MIMO detection technique known for its near-optimal performance in centralized implementation—within the context of a decentralized baseband processing architecture. This decentralized design mitigates the computation burden at a single processing unit by utilizing computational resources in a distributed and parallel manner. Additionally, we integrate the mini-batch stochastic gradient descent method into the proposed decentralized detector, achieving remarkable performance with high efficiency. Simulation results demonstrate substantial performance gains of the proposed method over existing decentralized detectors across various scenarios. Moreover, complexity analysis reveals the advantages of the proposed decentralized strategy in terms of computation delay and interconnection bandwidth when compared to conventional centralized detectors.

Stand-alone perception systems in autonomous driving suffer from limited sensing ranges and occlusions at extended distances, potentially resulting in catastrophic outcomes. To address this issue, collaborative perception is envisioned to improve perceptual accuracy by using vehicle-to-everything (V2X) communication to enable collaboration among connected and autonomous vehicles and roadside units. However, due to limited communication resources, it is impractical for all units to transmit sensing data such as point clouds or high-definition video. As a result, it is essential to optimize the scheduling of communication links to ensure efficient spectrum utilization for the exchange of perceptual data. In this work, we propose a deep reinforcement learning-based V2X user scheduling algorithm for collaborative perception. Given the challenges in acquiring perceptual labels, we reformulate the conventional label-dependent objective into a label-free goal, based on characteristics of 3D object detection. Incorporating both channel state information (CSI) and semantic information, we develop a double deep Q-Network (DDQN)-based user scheduling framework for collaborative perception, named SchedCP. Simulation results verify the effectiveness and robustness of SchedCP compared with traditional V2X scheduling methods. Finally, we present a case study to illustrate how our proposed algorithm adaptively modifies the scheduling decisions by taking both instantaneous CSI and perceptual semantics into account.

Along with the prosperity of generative artificial intelligence (AI), its potential for solving conventional challenges in wireless communications has also surfaced. Inspired by this trend, we investigate the application of the advanced diffusion models (DMs), a representative class of generative AI models, to high dimensional wireless channel estimation. By capturing the structure of multiple-input multiple-output (MIMO) wireless channels via a deep generative prior encoded by DMs, we develop a novel posterior inference method for channel reconstruction. We further adapt the proposed method to recover channel information from low-resolution quantized measurements. Additionally, to enhance the over-the-air viability, we integrate the DM with the unsupervised Stein’s unbiased risk estimator to enable learning from noisy observations and circumvent the requirements for ground truth channel data that is hardly available in practice. Results reveal that the proposed estimator achieves high-fidelity channel recovery while reducing estimation latency by a factor of 10 compared to state-of-the-art schemes, facilitating real-time implementation. Moreover, our method outperforms existing estimators while reducing the pilot overhead by half, showcasing its scalability to ultra-massive antenna arrays.

Device-to-Device (D2D) link scheduling in wireless communications is a challenging non-convex combinatorial optimization problem. The state-of-the-art methods, either from a model-based or a data-driven perspective, exhibit certain limitations such as the critical need of Channel State Information (CSI) and a large number of instances or solved instances as training samples. To advance this line of research, we propose a novel hybrid model/data-driven approach with Graph Reinforcement Learning for Link Scheduling (GRLinQ), injecting information theoretical insights into machine leaning models. GRLinQ demonstrates superior performance to the existing model-based and data-driven link scheduling mechanisms, with a relaxed requirement of CSI, a smaller number of unsolved instances as training samples, a possible distributed deployment, and more remarkably an excellent generalization ability over different network scenarios and system configurations.

This paper studies the beamformer design for an integrated sensing and communication system where simultaneous communication for multiple downlink users and bistatic sensing for a point target are realized. Firstly, to avoid self-interference in monostatic settings, we establish the bistatic sensing model where the sensing performance is measured by the Cramér-Rao bound (CRB) for the target's coordinates and the communication performance by the signal-to-interference-plus-noise ratio (SINR). We aim to minimize the CRB while ensuring a predefined SINR for each user. Then, we derive a closed-form beamformer for the single-user case that proves near-optimal. For the multi-user case, we introduce a beamformer design algorithm based on semidefinite relaxation and a suboptimal design algorithm based on successive convex approximation with reduced complexity. Finally, numerical results are provided to validate the solutions.

Existing physical-layer authentication (PLA) schemes of tag superimposed on message signals (TSM) can achieve high authentication accuracy at the cost of increased latency and reduced communication performance. The schemes of tag superimposed on pilot signals (TSP) achieve desirable communication performance and low latency, but low randomness of the tag results in lower security. To further improve both security and communication performance, we propose a pseudo random watermark hopping-based PLA scheme in this article. The proposed scheme generates a pseudo-random sequence and designs a watermark hopping mechanism, which superimposes a carefully designed tag on the pilot or message signals accordingly. The proposed scheme enhances the security by utilizing the randomness from both tag generation and watermark hopping mechanism. Meanwhile, it decreases the authentication latency and improves the communication performance by superimposing the tag on the pilot signals without the message recovery process before authentication. The theoretical and experimental results demonstrate that the proposed scheme decreases the bit error rate (BER) and outage probability as well as increases the achievable rate of the system compared with the TSM scheme with the same key equivocation. Moreover, the security performance of our scheme is significantly improved compared with both TSM and TSP schemes.

Artificial intelligence (AI) and machine learning (ML) have shown tremendous potential in reshaping the landscape of wireless communications and are, therefore, widely expected to be an indispensable part of the next-generation wireless network. This article presents an overview of how AI/ML and wireless communications interact synergistically to improve system performance and provides useful tips and tricks on realizing such performance gains when training AI/ML models. In particular, we discuss in detail the use of AI/ML to revolutionize key physical layer and lower medium access control (MAC) layer functionalities in traditional wireless communication systems. In addition, we provide a comprehensive overview of the AI/ML-enabled semantic communication systems, including key techniques from data generation to transmission. We also investigate the role of AI/ML as an optimization tool to facilitate the design of efficient resource allocation algorithms in wireless communication networks at both bit and semantic levels. Finally, we analyze major challenges and roadblocks in applying AI/ML in practical wireless system design and share our thoughts and insights on potential solutions.

This article focuses on power allocation and sampling rate control in a spectrum-sharing vehicle-to-vehicle (V2V) status update network, where data packets are sampled in batches at the transmitter of the V2V links. In particular, we aim to minimize the overall power consumption while satisfying the age of information (AoI) requirement of all links. First, we analyze the average AoI of the resulting queueing system with periodical packet batch arrivals and geometrically distributed packet service time. Then, the primal problem is decomposed into two subproblems, i.e., sampling rate optimization and transmit power optimization. Finally, a computationally efficient algorithm with polynomial-time complexity is developed that optimally solves the two subproblems and, thus, solves the original problem with global optimality. Simulation results validate our average AoI analysis and the proposed algorithm.

Federated learning (FL) enables edge devices to collaboratively learn a model in a distributed fashion. Many existing researches have focused on improving communication efficiency of high-dimensional models and addressing bias caused by local updates. However, most FL algorithms are either based on reliable communications or assuming fixed and known unreliability characteristics. In practice, networks could suffer from dynamic channel conditions and non-deterministic disruptions, with time-varying and unknown characteristics. To this end, in this paper we propose a sparsity-enabled FL framework with both improved communication efficiency and bias reduction, termed as SAFARI. It makes use of similarity among client models to rectify and compensate for bias that results from unreliable communications. More precisely, sparse learning is implemented on local clients to mitigate communication overhead, while to cope with unreliable communications, a similarity-based compensation method is proposed to provide surrogates for missing model updates. With respect to sparse models, we analyze SAFARI under bounded dissimilarity. It is demonstrated that SAFARI under unreliable communications is guaranteed to converge at the same rate as the standard FedAvg with perfect communications. Implementations and evaluations on the CIFAR-10 dataset validate the effectiveness of SAFARI by showing that it can achieve the same convergence speed and accuracy as FedAvg with perfect communications, with up to 60% of the model weights being pruned and a high percentage of client updates missing in each round of model updates.

This paper investigates a simultaneous wireless information and power transfer-enabled hybrid-powered multicell communication network with a nonlinear energy harvesting model. In this multicell environment, information interaction and energy trading are carried out among base stations (BSs), and each BS powered by hybrid sources simultaneously provides information/energy to its user equipments (UEs) over the downlinks. A novel global utility function is proposed by comprehensively considering the incomes from information and energy transmission, energy trading, and the costs from the electricity companies. To maximize this goal, a nonconvex problem that jointly optimizing energy procurement, power allocation, and power splitting ratios is formulated to deal with the imbalance between energy supply and demand at BSs. Considering the nonconvexity of the problem and the strong coupling among the optimization variables, the formulated problem is difficult to solve directly with conventional convex optimization methods. To overcome these obstacles, a two-step solution combining alternating optimization, distributed optimization, and successive convex approximation is designed, in which the BS layer scheme and the UE layer scheme are performed alternately. Different from the existing centralized schemes, the proposed distributed scheme makes local optimal decisions independently at each BS only resorting to the information of neighbor BSs, which brings great advantages in reducing signaling and computational overheads. Moreover, the convergence and advantages of the proposed distributed scheme are verified by rigorous analyses and simulations.

In this paper, we study low-complexity linear precoding for downlink massive multiple-input multiple-output (MIMO) systems, exploiting a statistical method. In sharp contrast to traditional linear precoding algorithms, our proposed efficient randomized iterative precoding algorithm (ERIPA) not only avoids costly matrix inversion but also considers the complexity reduction of matrix multiplication involved, thus enabling more efficient linear precoding. Additionally, ERIPA is demonstrated to have both exponentially fast and global convergence, making it adaptable to various practical scenarios of massive MIMO. We also investigate the convergence phenomenon of ERIPA in relation to the selection of the sampling distribution during random iterations. After that, the concept of conditional sampling is introduced to ERIPA such that significant system potential can be beneficially exploited in terms of both precoding performance and computational complexity. Finally, simulation results regarding the downlink massive MIMO are presented to confirm the superiorities of the proposed ERIPA.

Accurately detecting symbols transmitted over multiple-input multiple-output (MIMO) wireless channels is crucial in realizing the benefits of MIMO techniques. However, optimal MIMO detection is associated with a complexity that grows exponentially with the MIMO dimensions and quickly becomes impractical. Recently, stochastic sampling-based Bayesian inference techniques, such as Markov chain Monte Carlo (MCMC), have been combined with the gradient descent (GD) method to provide a promising framework for MIMO detection. In this work, we propose to efficiently approach optimal detection by exploring the discrete search space via MCMC random walk accelerated by Nesterov’s gradient method. Nesterov’s GD guides MCMC to make efficient searches without the computationally expensive matrix inversion and line search. Our proposed method operates using multiple GDs per random walk, achieving sufficient descent towards important regions of the search space before adding random perturbations, guaranteeing high sampling efficiency. To provide augmented exploration, extra samples are derived through the trajectory of Nesterov’s GD by simple operations, effectively supplementing the sample list for statistical inference and boosting the overall MIMO detection performance. Furthermore, we design an early stopping tactic to terminate unnecessary further searches, remarkably reducing the complexity. Simulation results and complexity analysis reveal that the proposed method achieves exceptional performance in both uncoded and coded MIMO systems, adapts to realistic channel models, and scales well to large MIMO dimensions.

Optimal detection of symbols transmitted over multiple-input multiple-output (MIMO) wireless channels is known to entail exponentially increasing complexity with MIMO dimensions, making it impractical for large-scale systems. Recently, Markov chain Monte Carlo (MCMC) has been combined with the gradient descent (GD) method to create a promising machine learning solution to this issue. This paper proposes a novel algorithm for approaching optimal detection via MCMC random walk accelerated by Nesterov's gradient method, efficiently exploring the discrete search space for MIMO detection. Our proposed method utilizes multiple GDs per random walk and guarantees high sampling efficiency while mitigating the complexity associated with matrix inversions. Simulation results and complexity analysis reveal that the proposed method achieves near-optimal performance and scales effectively to large MIMO dimensions.

In this article, we investigate sequential power allocation over fast varying channels for mission-critical applications, aiming to minimize the expected sum power while guaranteeing the transmission success probability. In particular, a reinforcement learning framework is constructed with appropriate reward design so that the optimal policy maximizes the Lagrangian of the primal problem, where the maximizer of the Lagrangian is shown to have several good properties. For the model-based case, a fast converging algorithm is proposed to find the optimal Lagrange multiplier and thus the corresponding optimal policy. For the model-free case, we develop a three-stage strategy, composed in order of online sampling, offline learning, and online operation, where a backward$Q$-learning with full exploitation of sampled channel realizations is designed to accelerate the learning process. According to our simulation, the proposed reinforcement learning framework can solve the primal optimization problem from the dual perspective. Moreover, the model-free strategy achieves a performance close to that of the optimal model-based algorithm.

As one technique for autonomous driving, vehicular networks can achieve high efficiency with vehicle-and-infrastructure cooperation, bringing high safety and many value-added services. To achieve higher communication efficiency, much effort has been done to cope with the resource allocation issues for vehicular networks. Nevertheless, due to the strong nonconvexity and nonlinearity, the classical joint resource allocation problem in vehicular networks is typically NP-hard. The multiagent reinforcement learning (MARL) has emerged as a promising solution to tackle this challenge but its stability and scalability are not satisfactory when the amount of vehicles gets increased. In this article, we mainly investigate the issue of joint spectrum and power allocation in vehicular communication networks, and carefully consider the interactions between the vehicles and environment by incorporating the cooperative stochastic game theory with MARL, named complete-game MARL (CG-MARL), to achieve a better convergence and stability with the theoretical computational complexity$\mathcal {O}(n^{N})$with$n$denoting the dimension of action space and$N$denoting the number of V2X Vehicular. Furthermore, the mean-field game (MFG) theory is employed to further enhance the MARL for decreasing the horrible computing resource consumption caused by the CG-MARL to$\mathcal {O}(n^{2})$while maintaining an approximate performance. The simulation results demonstrate that the proposed mean-field-aided MARL (MF-MARL) for vehicular network resource allocation can achieve 95% near-optimal performance with much lower complexity, which indicates its significant potentials in the scenarios with massive and dense vehicles.

This work considers a wireless powered communication network (WPCN), in which wireless nodes store the energy from an energy access point in their batteries for subsequent data transmission. An online energy consumption optimization strategy is proposed for adaptively determining the beamforming vector, data routing, network operation mode and transmitted power based only on the current state of WPCN. In most existing results, the energy/data transmission of WPCNs is based on the ideal battery models and the energy storage efficiencies therein are always assumed to be non-zero constants. Since the energy storage efficiency of batteries may be affected by the ambient environment or aging in real-time, this work considers a WPCN with a time-varying energy storage efficiencies sequence and correspondingly develops an improved Lyapunov optimization strategy to offset the impact of the time-varying energy storage efficiencies. More importantly, a distributed strategy is proposed to optimize the cooperation of wireless nodes over unrestricted numbers of hops, and thus the energy access point does not require channel state information of all data links and the data backlog queues of all nodes during the solving process. Accordingly, the computational burden at the energy access point is greatly reduced due to the use of this distributed strategy. Under this strategy, the time-averaged expected energy consumption of WPCN can be within a bounded gap of the minimum energy required to maintain stability of the network. Finally, the theoretical analysis is further corroborated by simulation results.

In this paper, we investigate the resource allocation problem in a dynamic vehicular environment, where multiple vehicle-to-vehicle links attempt to reuse the spectrum of vehicle-to-infrastructure links. It is modeled as a deep reinforcement learning problem that is subject to proximal policy optimization. Training a well-performing policy usually requires a massive amount of interactions with the environment for a long time and thus is typically performed on a simulator. However, an agent well trained in a simulated environment may still fail when deployed in a live network, due to inevitable difference between the two environments, termed reality gap. We make preliminary efforts to address this issue by leveraging meta reinforcement learning that allows the learning agent to quickly adapt to a new environment with minimal interactions after being trained across a variety of similar tasks. We demonstrate that only a few episodes are required for the meta trained policy to adapt to a new environment and the proposed method is shown to achieve near-optimal performance and exhibit rapid convergence.

With the large-scale deployment and commercialization of 5G, the traditional coverage enhancement technology is facing severe challenges due to high hardware cost and power consumption. At the same time, the emerging reconfigurable intelligent surface (RIS) technology, which can flexibly regulate incident waves and optimize the transmission path of wireless signals, is considered to be one of the most potential 6G key enabling technologies. This paper presents a 5G commercial network coverage enhancement prototype system based on RIS, which directly uses the base station in China Mobile 5G commercial network as transmitter, 5G network signal test terminal as receiver, and RIS code optimization algorithm is employed. The whole system forms a closed optimization loop of “obtaining data-analyzing data-feedback code” and is tested in multiple scenarios. In selected scenarios, the prototype system achieves a received power gain of about 6 dB, which is significant for the research of RIS in 5G commercial network.

Recently, task-oriented semantic communication has received increasing attention due to its potential to transform the communication landscape by going beyond the Shannon paradigm. While most existing researches focus on a single task, we propose a multi-task semantic communication system for text tasks, inspired by the impressive results of multitask learning and bidirectional encoder representations from transformers (BERT). Based on BERT, the proposed system extracts semantic information from the text at the transmitter and then transmits it to the receiver, which aims to accomplish a series of different tasks. The transmitter and receiver would be trained jointly in an end-to-end manner. Compared with the traditional communication system, the proposed model is more robust to distortion caused by physical channels and exhibits better performance, especially in the low signal-to-noise ratio regime. Moreover, we investigate the relationship between the number of tasks and the required length of transmitted symbols in a multi-task setting.

Recently, unmanned aerial vehicle (UAV) swarm communication has drawn much attention in search and rescue (SAR) missions owing to its wide wireless coverage and increasing autonomy in navigation. In this article, we consider maximizing the downlink wireless coverage of a UAV swarm in an unknown mission area by controlling the quasistationary deployments of UAVs. Particularly, the stochastic wireless link failures caused by channel fading and noise in UAV-to-UAV communication links are considered in coverage control. Specifically, due to delay sensitivity and onboard energy limitation of UAV-enabled SAR networks, we study a distributed control strategy where swarm UAVs can address the coverage problem by exchanging only the local information. In this case, the wireless coverage problem is divided into several distributed optimization subproblems. However, due to the integer variable and nonlinear constraints, each subproblem is nonconvex and mutually coupling, which makes it difficult to solve via standard convex optimization solvers. Thus, we model the UAV swarm network as an undirected random graph and then solve the optimization subproblems by formulating a UAV swarm wireless coverage game. As per the designed utility function and potential function of the formulated game, existence of the pure Nash equilibrium is discussed and a distributed algorithm is developed to achieve the best Nash equilibrium. We analyze the convergence property and computational complexity of the proposed algorithm. Meanwhile, we analyze effects of the initial learning rate and step size on algorithm performance from both theoretical and simulation results. Simulation results show that the proposed algorithm improves the coverage by around 58% when compared with initial performance.

Abstract Scene classification for remote sensing is a popular topic, and many recent convolutional neural networks (CNNs)‐based methods have shown the great model capacity and learning ability of highly discriminative features. Given a large number of training data, CNN can extract extensive features and learn to predict a remote sensing image. However, for supervised learning tasks, deep models often rely on a large number of labelled remote sensing images, which are difficult to pre‐process. Thus, training a lightweight deep learning model is essential. Easy‐classified and hard samples may also cause an imbalance of training set and lead the model to overwhelm the loss function. Accordingly, a novel Efficient Recurrent Attention Network (ERANet) for remote sensing scene classification is proposed. Different from traditional deep learning methods, Efficientnet‐B0 is introduced as a lightweight backbone for the ARCNet framework, replacing the original one. By applying the modified efficient backbone, the low Floating Point Operations (FLOPs) and parameter numbers of the proposed ERANet are maintained. The significance of focal loss is determined and applied to address the sample imbalance problem and yield a desirable performance. Extensive experiments on several challenging remote sensing scene classification data sets prove the efficiency of the proposed ERANet.

Physical layer signal processing algorithms in the wireless domain are seeing increased use of machine learning algorithms, especially Bayesian methods. This work presents the hardware architecture and implementation of a vector processor for one such application, Bayesian channel estimation (CE) (BCE). The BCE vector processor supports a generic instruction set with a supplement of specialized instructions to realize Bayesian algorithms in the signal processing context. The vector processor is designed to work as an accelerator in a system-on-chip (SoC) with an AHB/AXI bus interface or as stand-alone unit. The vector processor achieves more than$4\times $improvement in performance when compared with a traditional CE algorithm running on a commercial vector processor. To the best of authors knowledge, this is a first known hardware implementation of a variational Bayesian inference algorithm for a wireless communication application.

It has been a long-held belief that judicious resource allocation is critical to mitigating interference, improving network efficiency, and ultimately optimizing wireless communication performance. The traditional wisdom is to explicitly formulate resource allocation as an optimization problem and then exploit mathematical programming to solve the problem to a certain level of optimality. Nonetheless, as wireless networks become increasingly diverse and complex, for example, in the high-mobility vehicular networks, the current design methodologies face significant challenges and thus call for rethinking of the traditional design philosophy. Meanwhile, deep learning, with many success stories in various disciplines, represents a promising alternative due to its remarkable power to leverage data for problem solving. In this article, we discuss the key motivations and roadblocks of using deep learning for wireless resource allocation with application to vehicular networks. We review major recent studies that mobilize the deep-learning philosophy in wireless resource allocation and achieve impressive results. We first discuss deep-learning-assisted optimization for resource allocation. We then highlight the deep reinforcement learning approach to address resource allocation problems that are difficult to handle in the traditional optimization framework. We also identify some research directions that deserve further investigation.

Resource allocation has a direct and profound impact on the performance of vehicle-to-everything (V2X) networks. In this paper, we develop a hybrid architecture consisting of centralized decision making and distributed resource sharing (the C-Decision scheme) to maximize the long-term sum rate of all vehicles. To reduce the network signaling overhead, each vehicle uses a deep neural network to compress its observed information that is thereafter fed back to the centralized decision making unit. The centralized decision unit employs a deep Q-network to allocate resources and then sends the decision results to all vehicles. We further adopt a quantization layer for each vehicle that learns to quantize the continuous feedback. In addition, we devise a mechanism to balance the transmission of vehicle-to-vehicle (V2V) links and vehicle-to-infrastructure (V2I) links. To further facilitate distributed spectrum sharing, we also propose a distributed decision making and spectrum sharing architecture (the D-Decision scheme) for each V2V link. Through extensive simulation results, we demonstrate that the proposed C-Decision and D-Decision schemes can both achieve near-optimal performance and are robust to feedback interval variations, input noise, and feedback noise.

In this article, we develop an end-to-end wireless communication system using deep neural networks (DNNs), where DNNs are employed to perform several key functions, including encoding, decoding, modulation, and demodulation. However, an accurate estimation of instantaneous channel transfer function, i.e., channel state information (CSI), is needed in order for the transmitter DNN to learn to optimize the receiver gain in decoding. This is very much a challenge since CSI varies with time and location in wireless communications and is hard to obtain when designing transceivers. We propose to use a conditional generative adversarial net (GAN) to represent channel effects and to bridge the transmitter DNN and the receiver DNN so that the gradient of the transmitter DNN can be back-propagated from the receiver DNN. In particular, a conditional GAN is employed to model the channel effects in a data-driven way, where the received signal corresponding to the pilot symbols is added as a part of the conditioning information of the GAN. To address the curse of dimensionality when the transmit symbol sequence is long, convolutional layers are utilized. From the simulation results, the proposed method is effective on additive white Gaussian noise (AWGN) channels, Rayleigh fading channels, and frequency-selective channels, which opens a new door for building data-driven DNNs for end-to-end communication systems.

As wireless networks evolve toward high mobility and providing better support for connected vehicles, a number of new challenges arise due to the resulting high dynamics in vehicular environments and thus motive rethinking of traditional wireless design methodologies. Future intelligent vehicles, which are at the heart of high mobility networks, are increasingly equipped with multiple advanced onboard sensors and keep generating large volumes of data. Machine learning, as an effective approach to artificial intelligence, can provide a rich set of tools to exploit such data for the benefit of the networks. In this paper, we first identify the distinctive characteristics of high mobility vehicular networks and motivate the use of machine learning to address the resulting challenges. After a brief introduction of the major concepts of machine learning, we discuss its applications to learn the dynamics of vehicular networks and make informed decisions to optimize network performance. In particular, we discuss in greater detail the application of reinforcement learning in managing network resources as an alternative to the prevalent optimization approach. Finally, some open issues worth further investigation are highlighted.

Vehicular communications face a tremendous challenge in guaranteeing low latency for safety-critical information exchange due to fast varying channels caused by high mobility. Focusing on the tail behavior of random latency experienced by packets, latency violation probability (LVP) deserves particular attention. Based on only large-scale channel information, this paper performs spectrum and power allocation to maximize the sum ergodic capacity of vehicle-to-infrastructure (V2I) links while guaranteeing the LVP for vehicle-to-vehicle (V2V) links. Using the effective capacity theory, we explicitly express the latency constraint with introduced latency exponents. Then, the resource allocation problem is decomposed into a pure power allocation subproblem and a pure spectrum allocation subproblem, both of which can be solved with global optimum in polynomial time. Simulation results show that the effective capacity model can accurately characterize the LVP. In addition, the effectiveness of the proposed algorithm is demonstrated from the perspectives of the capacity of the V2I links and the latency of the V2V links.

This paper investigates the spectrum sharing problem in vehicular networks based on multi-agent reinforcement learning, where multiple vehicle-to-vehicle (V2V) links reuse the frequency spectrum preoccupied by vehicle-to-infrastructure (V2I) links. Fast channel variations in high mobility vehicular environments preclude the possibility of collecting accurate instantaneous channel state information at the base station for centralized resource management. In response, we model the resource sharing as a multi-agent reinforcement learning problem, which is then solved using a fingerprint-based deep Q-network method that is amenable to a distributed implementation. The V2V links, each acting as an agent, collectively interact with the communication environment, receive distinctive observations yet a common reward, and learn to improve spectrum and power allocation through updating Q-networks using the gained experiences. We demonstrate that with a proper reward design and training mechanism, the multiple V2V agents successfully learn to cooperate in a distributed way to simultaneously improve the sum capacity of V2I links and payload delivery rate of V2V links.

Proximity-based communications have been considered as a promising candidate for supporting vehicular communications. However, the high mobility in vehicular communications makes it hard to obtain accurate fast varying channel information, which poses significant challenges on meeting the requirements of high reliability and low latency. Based only on slowly varying large-scale fading channel information, this paper performs a reliability and latency aware resource allocation, which maximizes the throughput of vehicular-to-network (V2N) links while satisfying reliability and latency requirements of vehicular-to-vehicular (V2V) links. First, we obtain steady-state reliability and latency expressions based on queueing analysis for each possible spectrum reusing pair of a V2N link and a V2V link. Then, an optimal power allocation algorithm is developed for each possible spectrum reusing pair. Afterward, the spectrum reusing pattern is optimized by addressing a polynomial time solvable bipartite matching problem. The simulation results demonstrate the accuracy of the proposed queueing analysis and confirm the effectiveness of the proposed resource allocation comparing with available strategies.

Vehicular communications have stringent latency requirements on safety-critical information transmission. However, lack of instantaneous channel state information due to high mobility poses a great challenge to meet these requirements and the situation gets more complicated when packet retransmission is considered. Based on only the obtainable large- scale fading channel information, this paper performs spectrum and power allocation to maximize the ergodic capacity of vehicular-to- infrastructure (V2I) links while guaranteeing the latency requirements of vehicular-to-vehicular (V2V) links. First, for each possible spectrum reusing pair of a V2I link and a V2V link, we obtain the closed- form expression of the packets' average sojourn time (the queueing time plus the service time) for the V2V link. Then, an optimal power allocation is derived for each possible spectrum reusing pair. Afterwards, we optimize the spectrum reusing pattern by addressing a polynomial time solvable bipartite matching problem. Numerical results show that the proposed queueing analysis is accurate in terms of the average packet sojourn time. Moreover, the developed resource allocation always guarantees the V2V links' requirements on latency.

This paper investigates the resource allocation problem in device-to-device (D2D)-based vehicular communications, based on slow fading statistics of channel state information (CSI), to alleviate signaling overhead for reporting rapidly varying accurate CSI of mobile links. We consider the case when each vehicle-to-infrastructure (V2I) link shares spectrum with multiple vehicle-to-vehicle (V2V) links. Leveraging the slow fading statistical CSI of mobile links, we maximize the sum V2I capacity while guaranteeing the reliability of all V2V links. We propose a graph- based algorithm that uses graph partitioning tools to divide highly interfering V2V links into different clusters before formulating the spectrum sharing problem as a weighted 3-dimensional matching problem, which is then solved through adapting a high-performance approximation algorithm.

We investigate a general channel estimation problem in the massive multiple-input multiple-output system which employs the hybrid analog/digital precoding structure with limited radio-frequency (RF) chains. By properly designing RF combiners and performing multiple trainings, the proposed channel estimation can approach the performance of fully-digital estimations depending on the degree of channel spatial correlation and the number of RF chains. Dealing with the hybrid channel estimation, the optimal combiner is theoretically derived by relaxing the constant-magnitude constraint in a specific single-training scenario, which is then extended to the design of combiners for multiple trainings by sequential and alternating methods. Further, we develop a technique to generate the phase-only RF combiners based on the corresponding unconstrained ones to satisfy the constant-magnitude constraints. The performance of the proposed hybrid channel estimation scheme is examined by simulations under both nonparametric and spatial channel models. The simulation results demonstrate that the estimated channel state information can approach the performance of fully-digital estimations in terms of both mean square error and spectral efficiency. Moreover, a practical spatial channel covariance estimation method is proposed and its effectiveness in hybrid channel estimation is verified by simulations.

This paper investigates the resource allocation problem in device-to-device-based vehicular communications, based on slow fading statistics of channel state information (CSI), to alleviate signaling overhead for reporting rapidly varying accurate CSI of mobile links. We consider the case when each vehicle-to-infrastructure (V2I) link shares spectrum with multiple vehicle-to-vehicle (V2V) links. Leveraging the slow fading statistical CSI of mobile links, we maximize the sum V2I capacity while guaranteeing the reliability of all V2V links. We use graph partitioning tools to divide highly interfering V2V links into different clusters before formulating the spectrum sharing problem as a weighted 3-D matching problem. We propose a suite of algorithms, including a baseline graph-based resource allocation algorithm, a greedy resource allocation algorithm, and a randomized resource allocation algorithm, to address the performance-complexity tradeoffs. We further investigate resource allocation adaption in response to slow fading CSI of all vehicular links and develop a low-complexity randomized algorithm.

The widely deployed cellular network, assisted with device-to-device (D2D) communications, can provide a promising solution to support efficient and reliable vehicular communications. In this paper, we identify differentiated requirements for different types of vehicular links, i.e., high capacity for vehicle-to-infrastructure (V2I) links and ultra reliability for vehicle-to-vehicle (V2V) links, and attempt to maximize the ergodic capacity of V2I connections while ensuring reliability guarantee for each V2V link. To account for fast channel variations caused by high mobility, we propose to perform spectrum sharing and power allocation based only on slowly varying large-scale fading information of wireless channels. A novel algorithm that yields optimal resource allocation and is robust to channel variations is proposed. Their desirable performance is confirmed by computer simulation.

The widely deployed cellular network, assisted with device-to-device (D2D) communications, can provide a promising solution to support efficient and reliable vehicular communications. Fast channel variations caused by high mobility in a vehicular environment need to be properly accounted for when designing resource allocation schemes for the D2D-enabled vehicular networks. In this paper, we perform spectrum sharing and power allocation based only on slowly varying large-scale fading information of wireless channels. Pursuant to differing requirements for different types of links, i.e., high capacity for vehicle-to-infrastructure (V2I) links and ultrareliability for vehicle-to-vehicle (V2V) links, we attempt to maximize the ergodic capacity of the V2I connections while ensuring reliability guarantee for each V2V link. Sum ergodic capacity of all V2I links is first taken as the optimization objective to maximize the overall V2I link throughput. Minimum ergodic capacity maximization is then considered to provide a more uniform capacity performance across all V2I links. Novel algorithms that yield optimal resource allocation and are robust to channel variations are proposed. Their desirable performance is confirmed by computer simulation.

In the above paper[1], the text discussion of several equations were misrepresented. Below is the corrected text ofSections IIIandIV, in which the errors appear.

The relay technology is effective in extending radio coverage and improving the performance of cell edge users. In multi-antenna relay channels, good knowledge of the channel state information (CSI) at the transmitter is important to achieve multiplexing gains of the multiple-input multiple-output technique. In this paper, we study the multi-antenna relay downlink channel with limited feedback CSI from both two-hop links. Data streams from the base station (BS) are first transmitted to a relay station (RS) with singular value decomposition-based precoding and receiver pulse shaping at the BS and RS, respectively. The block diagonalization precoding is then applied at the RS to forward the received signals to the remote multi-antenna users. We derive an upper bound for the system throughput loss due to CSI quantization error, and then propose a feedback quality control strategy to maintain a bounded rate loss relative to the perfect CSI case. It reveals that the feedback size B_1 from the RS to BS needs to scale in proportion to both transmit power at the BS and RS while the feedback size B_2 from each user to the RS only needs to scale linearly with the transmit power at the RS.

This paper considers device-to-device (D2D) communications underlaying cellular networks with a multi-antenna base station (BS). The BS serves its own cellular users while letting another remote terminal directly transmit signals to its nearby receiver via a D2D link. Two transmit strategies including beamforming (BF) and interference cancellation (IC) are considered at the BS for performance evaluation in terms of achievable channel capacity. The capacity performance of two different cases with perfect and quantized channel knowledge at the transmitter is derived with closed-form expressions. Based on these results, an adaptive transmission scheme to switch between BF and IC is proposed. Numerical results verify the accuracy of the derived expressions and draw the operating regions of BF/IC strategies.

This paper studies the capacity performance of coordinated multi-point (CoMP) downlink transmissions based on limited feedback. We consider both path loss effects and channel imperfections including feedback delay and quantization error. Closed-form expressions are derived to characterize ergodic achievable rates for joint processing (JP) and coordinated beamforming (CBF) techniques, respectively. According to the derived expressions, an adaptive transmission strategy to switch between JP and CBF is proposed to maximize cell throughput. Simulation shows the CBF scheme is preferred at medium SNR with varying switch points jointly determined by the feedback size, delay, and locations of CoMP users.