The fast development of Artificial Intelligence (AI) agents provides a promising way for the realization of intelligent and customized wireless networks. In this paper, we propose a Wireless Multi-Agent System (WMAS), which can provide intelligent and customized services for different user equipment (UEs). Note that orchestrating multiple agents carries the risk of malfunction, and multi-agent conversations may fall into infinite loops. It is thus crucial to design a conversation topology for WMAS that enables agents to complete UE task requests with high accuracy and low conversation overhead. To address this issue, we model the multi-agent conversation topology as a directed acyclic graph and propose a reinforcement learning- based algorithm to optimize the adjacency matrix of this graph. As such, WMAS is capable of generating and self-optimizing multi-agent conversation topologies, enabling agents to effectively and collaboratively handle a variety of task requests from UEs. Simulation results across various task types demonstrate that WMAS can achieve higher task performance and lower conversation overhead compared to existing multi-agent systems. These results validate the potential of WMAS to enhance the intelligence of future wireless networks.

With the proliferation of mobile devices and IoT applications, edge caching has become vital for mitigating network congestion and enhancing user Quality of Experience (QoE). However, traditional caching policies, such as Least Frequently Used (LFU), First-In-First-Out (FIFO), and Least Recently Used (LRU), often struggle to perform effectively in highly dynamic and heterogeneous environments, particularly when content sizes vary significantly. Moreover, existing approaches, whether AI-driven or heuristic-based, typically adopt a single caching strategy, which inherently limits their flexibility and adaptability. To address these limitations, we propose ECMSA, a learning-based multi-strategy edge caching algorithm that integrates a reinforcement learning-driven proactive caching strategy with three conventional reactive caching strategies. Specifically, ECMSA operates in two stages: First, it generates four candidate cache lists—three derived from traditional caching policies (LFU, FIFO, LRU) and one produced by our self-attention-enhanced Deep Deterministic Policy Gradient (Atten-Actor DDPG)-based proactive caching strategy. Next, it employs another Atten-Actor DDPG agent to dynamically select the optimal strategy in real time, leveraging current state features. This dual-agent framework enables continuous learning and adaptation of caching decisions, effectively optimizing content placement and update policies in response to evolving user demands. Extensive experiments conducted on both synthetic and real-world datasets demonstrate that ECMSA achieves 15-17% higher cache-hit ratios and 16-22% lower latency than baseline methods under constrained cache capacities and diverse content sizes. Furthermore, ECMSA exhibits strong robustness and generalization ability, allowing it to rapidly adapt to unseen environments.

The integration of artificial intelligence (AI) and mobile networks is regarded as one of the most important scenarios for 6G. In 6G, a major objective is to realize the efficient transmission of task-relevant data. Then a key problem arises, how to design collaborative AI models for the device side and the network side, so that the transmitted data between the device and the network is efficient enough, which means the transmission overhead is low but the AI task result is accurate. In this paper, we propose the multi-link information bottleneck (ML-IB) scheme for such collaborative models design. We formulate our problem based on a novel performance metric, which can evaluate both task accuracy and transmission overhead. Then we introduce a quantizer that is adjustable in the quantization bit depth, amplitudes, and breakpoints. Given the infeasibility of calculating our proposed metric on high-dimensional data, we establish a variational upper bound for this metric. However, due to the incorporation of quantization, the closed form of the variational upper bound remains uncomputable. Hence, we employ the Log-Sum Inequality to derive an approximation and provide a theoretical guarantee. Based on this, we devise the quantized multi-link information bottleneck (QML-IB) algorithm for collaborative AI models generation. Finally, numerical experiments demonstrate the superior performance of our QML-IB algorithm compared to the state-of-the-art algorithm.

By moving to millimeter wave (mmWave) frequencies, base stations (BSs) will be densely deployed to provide seamless coverage in sixth generation (6G) mobile communication systems, which, unfortunately, leads to severe cell-edge problem. In addition, with massive multiple-input-multiple-output (MIMO) antenna arrays employed at BSs, the beamspace channel is sparse for each user, and thus there is no need to serve all the users in a cell by all the beams therein jointly. Therefore, it is of paramount importance to develop a flexible clustered cell-free networking scheme that can decompose the whole network into a number of weakly interfered small subnetworks operating independently and in parallel. Given a per-user rate constraint for service quality guarantee, this paper aims to maximize the number of decomposed subnetworks so as to reduce the signaling overhead and system complexity as much as possible. By formulating it as a bipartite graph partitioning problem, a rate-constrained network decomposition (RC-NetDecomp) algorithm is proposed, which can smoothly tune the network structure from the current cellular network with simple beam allocation to a fully cooperative network by increasing the required per-user rate. Simulation results demonstrate that the proposed RC-NetDecomp algorithm outperforms existing baselines in terms of average per-user rate, fairness among users and energy efficiency.

Wireless simultaneous localization and mapping (SLAM) has attracted much attention as a promising technique to empower location based services. However, the accuracy of traditional wireless SLAM systems is limited as the wireless signals are easily disturbed by the uncontrollable radio environments. To mitigate this issue, in this paper, we propose a MetaSLAM system where multiple reconfigurable intelligent surfaces (RISs) are deployed to customize the wireless environments. To be specific, through adjusting the phase shifts of these RISs, the strength of reflected signals can be enhanced in order to resist the variance of radio environments. However, it is challenging to coordinate multiple RISs and optimize their phase shifts especially when their locations are unknown to the agent. In order to address these challenges, we formulate a MetaSLAM optimization problem, and design a two-stage optimization algorithm based on the genetic and particle filter algorithms to solve the formulated problem. Analysis of the complexity and the positioning error bound of the proposed SLAM system are provided. Simulation results show that compared with the benchmark schemes, the positioning error obtained by the MetaSLAM system is reduced by at least 31%.

Wireless network capacity can be regarded as the most important performance metric for wireless communication systems. With the fast development of wireless communication technology, future wireless systems will become more and more complicated. As a result, the channel gain matrix will become a large-dimensional random matrix, leading to an extremely high computational cost to obtain the capacity. In this paper, we propose a moment passing method (MPM) to realize the fast and accurate capacity estimation for future ultra-dense wireless systems. It can determine the capacity with quadratic complexity, which is optimal considering that the cost of a single matrix operation is not less than quadratic complexity. Moreover, it has high accuracy. The simulation results show that the estimation error of this method is below 2%. Finally, our method is highly general, as it is independent of the distributions of BSs and users, and the shape of network areas. More importantly, it can be applied not only to the conventional multi-user multiple input and multiple output (MU-MIMO) networks, but also to the capacity-centric networks designed for B5G/6G.

Base-stations (BSs) will be densely deployed to provide seamless coverage in sixth generation (6G) mobile communication systems, which, unfortunately, leads to severe cell-edge problem. A flexible clustered cell-free networking scheme to replace the cellular network is studied in this paper, which decomposes the whole network into a number of subnetworks operating independently. In order to reduce signaling overhead and system complexity as far as possible, we aim to maximize the number of decomposed subnetworks with a per-user rate constraint for service quality guarantee. In addition, subnetworks with BSs only are allowed to enable BS sleep mode operation. A rate-constrained network decomposition (RC-NetDecomp) algorithm is proposed, which can smoothly tune the network structure from the current cellular network to the fully cooperative network by varying the required per-user rate. Simulation results demonstrate that it outperforms the existing baselines in terms of both average per-user rate and fairness among users.

The sixth generation (6G) era is envisioned to be a fully intelligent and autonomous era, with physical and digital lifestyles merged together. Future wireless network architectures should provide a solid support for such new lifestyles. A key problem thus arises that what kind of network architectures are suitable for 6G. In this paper, we propose a capacity-guaranteed network (CGN) architecture, which provides high capacity for wireless devices densely distributed everywhere, and ensures a superior scalability with low signaling overhead and computation complexity simultaneously. Our theorem proves that the essence of a CGN architecture is to decompose the whole network into non-overlapping clusters with equal cluster sum capacity. Simulation results reveal that in terms of the minimum cluster sum capacity, the proposed CGN can achieve at least 30% performance gain compared with existing base station clustering (BS-clustering) architectures. In addition, our theorem is sufficiently general and can be applied for networks with different distributions of BSs and users.

Reconfigurable intelligent surfaces (RISs) are regarded as one of the most promising techniques in the sixth-generation (6G) mobile communication networks. With the feature of smartly tuning the electromagnetic environment, RISs provide a possibility for ubiquitous and high-precision localization in 6G. However, proper system models for large indoor RIS-assisted networks and high-precision localization algorithms are still missing. In this paper, we propose a RIS-assisted downlink millimeter-wave (mmWave) indoor localization framework based on segment-by-segment far-field assumption. In addition, a brand new coarse-to-fine localization algorithm with low-complexity grid design is provided. Numerical results show that millimeter-level localization precision is achieved under the RIS-assisted indoor scenarios, which reveals that RIS can provide a solid support for accurate localization in the 6G era.

Capacity is one of the most important performance metrics for wireless communication networks. It describes the maximum rate at which the information can be transmitted of a wireless communication system. To support the growing demand for wireless traffic, wireless networks are becoming more dense and complicated, leading to a higher difficulty to derive the capacity. Unfortunately, most existing methods for the capacity calculation take a polynomial time complexity. This will become unaffordable for future ultra-dense networks, where both the number of base stations (BSs) and the number of users are extremely large. In this paper, we propose a fast algorithm TOSE to estimate the capacity for ultra-dense wireless networks. Based on the spiked model of random matrix theory (RMT), our algorithm can avoid the exact eigenvalue derivations of large dimensional matrices, which are complicated and inevitable in conventional capacity calculation methods. Instead, fast eigenvalue estimations can be realized based on the spike approximations in our TOSE algorithm. Our simulation results show that TOSE is an accurate and fast capacity approximation algorithm. Its estimation error is below 5%, and it runs in linear time, which is much lower than the polynomial time complexity of existing methods. In addition, TOSE has superior generality, since it is independent of the distributions of BSs and users, and the shape of network areas.

Recently, the intelligent omni-surface (IOS) has been proposed as a novel instance of metasurface to achieve full-dimensional communications by jointly engineering its reflective and refractive properties. However, optimal beamforming scheme for the IOS is hard to obtain due to the difficulty in acquiring perfect channel state information (CSI). To address this issue, in this paper, we consider an IOS aided system where the beamforming scheme is designed via beam training with codebooks at the base station (BS), the IOS, and users. Given that the refractive/reflective signals are closely related to both incident signals from the BS and phase shifts of IOS elements, the codebooks at the BS and the IOS are designed jointly. Based on the joint BS-IOS codebook, a multi-lobe beam training mechanism is proposed to perform beam training for multiple users simultaneously, thereby reducing the training overhead. Simulation results indicate that our proposed scheme achieves a higher sum rate than the state-of-the-art beam training schemes and performs close to the perfect CSI case.

The accelerated convergence of digital and real-world lifestyles has imposed unprecedented demands on today’s wireless network architectures, as it is highly desirable for such architectures to support wireless devices everywhere with high capacity and minimal signaling overhead. Conventional architectures, such as cellular architectures, are not able to satisfy these requirements simultaneously, and are thus no longer suitable for the future era. In this paper, we propose a capacity-centric (C2) architecture for future wireless networking. It is designed based on the principles of maximizing the number of non-overlapping clusters with the average cluster capacity guaranteed to be higher than a certain threshold, and thus provides a flexible way to balance the capacity requirement against the signaling overhead. Our analytical results reveal that C2 has superior generality, wherein both the cellular and the fully coordinated architectures can be viewed as its extreme cases. Simulation results show that the average capacity of C2 is at least three times higher compared to that of the cellular architecture. More importantly, different from the widely adopted conventional wisdom that base-station distributions dominate architecture designs, we find that the C2 architecture is not over-reliant on base-station distributions, and instead the user-side information plays a vital role and cannot be ignored.

Recently, the intelligent omni-surface (IOS) has been proposed as a novel instance of metasurface to achieve full-dimensional communications by jointly engineering its reflective and refractive properties. However, optimal beamforming scheme for the IOS is hard to obtain due to the difficulty in acquiring perfect channel state information (CSI). To address this issue, in this paper, we consider an IOS aided system where the beamforming scheme is designed via beam training with codebooks at the base station (BS), the IOS, and users. Given that the refractive/reflective signals are closely related to both incident signals from the BS and phase shifts of IOS elements, the codebooks at the BS and the IOS are designed jointly. Based on the joint BS-IOS codebook, a multi-lobe beam training mechanism is proposed to perform beam training for multiple users simultaneously, thereby reducing the training overhead. The training/feedback overhead of the proposed beam training and the impact of the codebook size are then analyzed theoretically. Simulation results indicate that the proposed scheme achieves a higher sum rate than the state-of-the-art beam training schemes and performs close to the perfect CSI case.

The 5th generation (5G) wireless networks are conceived in the form of heterogeneous networks (HetNets), where small cells are deployed over the conventional macrocell networks to improve the spectral efficiency. In HetNets, the interference between different tiers is the main bottleneck for achieving high spectral efficiency. Many spectrum access schemes have been proposed to manage the cross-tier interference. Unfortunately, the optimal spectrum access scheme remains unknown. In this paper, we propose an F-ALOHA based cognitive spectrum access scheme for macro-femto HetNets, where the femtocells can access the idle macro-tier spectrum with a certain probability. Therefore, besides the degrees of freedom from the conventional spectrum deployment and co-tier spectrum access, the proposed scheme obtains a new degree of freedom from cross-tier spectrum access for interference management and spectral efficiency optimization. Simulation results will show that the proposed scheme outperforms existing F-ALOHA based spectrum access schemes in terms of the area spectral efficiency (ASE). More importantly, it is observed that the maximum ASE is achieved when the number of active links per unit area, which governs the interference level, reaches a certain value. The advantage of the proposed scheme comes from its ability to offload the traffic between two tiers through the cross-tier spectrum access probability, which flexibly manages the cross-tier interference.

Deploying femtocells over the conventional macrocell network is a promising way to increase network capacity, whereas the main bottleneck is the interference between the femtocell and macrocell tiers. Recent research has proposed many effective methods for cross-tier interference mitigation, but unfortunately, the optimal spectrum access scheme remains unknown. In this paper, a cognitive spectrum access scheme is proposed, where each femtocell can dynamically explore and access the idle macro-tier spectrum besides its dedicated femto-tier spectrum. The optimum probabilities for each femtocell to access the femto-tier and idle macro-tier spectrum which maximize the area spectral efficiency (ASE) are investigated. With perfect spectrum sensing, the ratio between the optimum probabilities to access the femto-tier and macro-tier spectrum is equal to the amounts of available subchannels in the femto-tier and macro-tier spectrum for most cases. With non-perfect spectrum sensing, a lower bound of the optimum probability to access the femto-tier spectrum is determined, which is a piecewise function of the femtocell intensity. Simulation results validate the accuracy of our analytical results and reveal the advantages of the proposed scheme over existing schemes.

The deployment of femtocells in a conventional cellular network is a promising way to increase network capacity, whereas the main bottleneck is the interference between and within tiers. Previous work proposed channel splitting and F-ALOHA to manage the cross-tier and co-tier interference, respectively. However, such spectrum allocation scheme is not efficient given the often scenarios where part of the macro-tier spectrum is vacant but the femto-tier spectrum is overused. In this paper, a cognitive spectrum access scheme is proposed, where femtocells can access both femto-tier and macro-tier spectrum with certain probabilities, to increase the area spectral efficiency (ASE). The closed-form expressions of the optimum spectrum access probabilities in maximizing the ASE are derived for two scenarios where macrocell base stations (MBSs) are modeled as Poisson point process (PPP) and periodic grid. Analytical results reveal that for most cases, the ratio between the optimum probabilities for femtocells to access the femto-tier and macro-tier spectrum is equal to the ratio between the number of subchannels in the femto-tier and idle macro-tier spectrum. Simulation results show that with both models, the proposed scheme outperforms previous work in terms of the ASE.