This work explores combining the rate-splitting multiple-access (RSMA) technique with an active reconfigurable intelligent surface (RIS) to improve the quantized multiuser multiple-input single-output network. The active RIS facilitates communication between the base station (BS) and users equipped with low-resolution quantizers, whereas RSMA improves downlink transmission efficiency. By maximizing the spectral efficiency while minimizing the power consumption at the transmitter and active RIS, we formulate an energy efficiency maximization problem by jointly designing the BS precoding matrix and active RIS reflecting matrix. The optimization problem presents nonconvexity, which makes finding the optimal solution challenging. Therefore, we reformulate the problem into a reinforcement learning-based problem that is solvable by applying deep reinforcement learning (DRL) algorithms. To ensure action accuracy, we design a constraint-matching function that integrates with the DRL algorithm, forming a DRL framework securing all problem constraints. To assess the proposed DRL algorithm, we propose an alternating-based solution that decomposes the problem into precoding matrix optimization and active reflecting matrix optimization sub-problems, which are solvable using the successive convex approximation-based method. The performance evaluations demonstrate the convergence and effectiveness of the proposed approaches in various scenarios.

Recently, there has been a growing interest in exploring new multi-antenna technologies for 6 G wireless networks, including the use of extremely largescale antenna arrays, tremendously high frequencies, and novel antenna technologies. These emerging trends introduce unique characteristics that cannot be adequately addressed by classical far-field communication techniques with planar wavefronts. As a result, there is a need to investigate near-field communication design with spherical wavefronts. In this paper, we present a comprehensive overview of near-field communications, focusing on basic concepts, applications, and future directions.

One of the essential factors for enabling sixth-generation systems is efficiently ensuring diverse quality-of-service (QoS) performance metrics to support the upcoming massive ultra-reliable low-latency communication (URLLC). This work proposes efficient transmission control in intelligent reflecting surface (IRS)-assisted nonorthogonal multiple access (NOMA) networks in the finite blocklength (FBL) regime that statistically guarantee stringent URLLC QoS requirements. Thus, we formulate a nonconvex problem that maximizes the sum effective capacity (SEC) while ensuring statistical delay QoS constraints. To make the problem more tractable, we propose a tight upper bound for the objective function based on Jensen’s inequality and employ the concept of opportunistically minimizing an expectation. Then, we decompose the problem into two subproblems: active beamforming at the base station and phase-shift optimization at the IRS. Each subproblem is convexified by employing slack variables, penalty functions, and linear approximation, and solved using successive convex approximations. The subproblems are iteratively solved until convergence using alternating optimization. The convergence to a suboptimal stationary solution and the computing complexity of the proposed algorithm are rigorously analyzed. Finally, extensive numerical evaluations confirm that the proposed control in the FBL regime significantly improves the SEC under various QoS parameters compared to existing benchmark schemes. In particular, as the number of antennas and IRS elements increases, the proposed method becomes more efficient than the semi-definite relaxation-based approach in terms of complexity and performance.

Rate-splitting multiple access (RSMA) and reconfigurable intelligent surface (RIS) techniques show promise in enhancing spectral efficiency in sixth-generation Internet of Things (IoT) networks. However, optimizing the synergy between these two methods is challenging due to the complex and dynamic environment. This study focuses on maximizing the sum-rate metric in RIS-assisted uplink multiantenna RSMA IoT networks to address this problem. We jointly optimized the base station beamforming design, power allocation, and RIS phase shifts to enhance the spectral efficiency with multiple mobile IoT devices present. The controlled parameters are continuous variables and the mathematical problem is nonconcave. Therefore, we formulated the problem as a Markov decision process and used the deep deterministic policy gradient (DDPG) to determine the optimal joint actions. We proposed a safe action shaping process for the decision-making actor network to address constraint violations. Through a rigorous performance evaluation, we demonstrated that the DDPG approach with action shaping outperforms the current DDPG algorithm regarding the maximum achievable sum rate.

In this study, we consider an energy-efficient and low-complexity transmission control in a SWIPT-NOMA-based green cellular network (GCN) that consists of a green base station (GBS) and green users (GUEs). First, we formulate a non-convex problem that minimizes transmit power consumption while supporting minimum downlink user service rate, downlink data queue stability, and user battery availability. Then, we transform the problem into a Lyapunov-drift-penalty minimization problem, which can determine a new resource allocation scheme that balances transmit power consumption and queue stability. Second, the Lyapunov-drift-penalty problem is decomposed into subchannel assignment, power allocation, and power splitting (PS) ratio control problems. The subchannel assignment problem is solved using a matching theory-based low-complexity algorithm. The power allocation and PS ratio control problems are solved using the alternating optimization (AO) approach and bisection method. This decomposed subproblem-based control also enables distributed control between the GBS and GUEs. Third, we prove the convergence, optimality, and polynomial computation complexity of the proposed algorithm. Lastly, we demonstrate that the proposed control outperforms the benchmark controls regarding transmit power consumption and the achievable rate. Owing to the optimality and low complexity, the proposed control can be efficiently applied to large-scale and distributed GCNs in sixth-generation environments.

To meet the stringent demands of emerging Internet-of-Things (IoT) applications, such as smart home, smart city, and virtual reality in 5G/6G IoT networks, edge content caching for mobile/multiaccess edge computing (MEC) has been identified as a promising approach to improve the quality of services in terms of latency and energy consumption. However, the limitations of cache capacity make it difficult to develop an effective common caching framework that satisfies diverse user preferences. In this article, we propose a new content caching strategy that maximizes the cache hit ratio through flexible prediction in dynamically changing network and user environments. It is based on a hierarchical deep learning architecture: long short-term memory (LSTM)-based local learning and ensemble-based meta-learning. First, as a local learning model, we employ an LSTM method with seasonal-trend decomposition using loess (STL)-based preprocessing. It identifies the attributes for demand prediction on the contents in various demographic user groups. Second, as a metalearning model, we employ a regression-based ensemble learning method, which uses an online convex optimization framework and exhibits sublinear “regret” performance. It orchestrates the obtained multiple demographic user preferences into a unified caching strategy in real time. Extensive experiments were conducted on the popular MovieLens data sets. It was shown that the proposed control provides up to a 30% higher cache hit ratio than conventional representative algorithms and a near-optimal cache hit ratio within approximately 9% of the optimal caching scheme with perfect prior knowledge of content popularity. The proposed learning and caching control can be implemented as a core function of the 5G/6G standard’s network data analytic function (NWDAF) module.

Intelligent reflecting surface (IRS) is a promising 6G technology that can improve wireless communication capacity in a cost-effective and energy-efficient manner, by adjusting a large number of passive reflectors to appropriately change the signal propagation. In this study, we identified the achievable rate region of a two-hop interference channel with distributed multiple IRS relays. To do so, we formulated a non-convex problem that characterizes the rate-profile, and found its solution using successive convex approximation (SCA). We then proposed an alternating direction method of multipliers (ADMM) and alternating optimization (AO) based distributed and low-complex IRS control that maximizes the achievable sum-rate, and proved its convergence and optimality. We then compared the proposed IRS control with semi-definite relaxation (SDR)-, random phase-, deep reinforcement learning (DRL)- based IRS controls, and optimal amplify-and-forward (AF)-, interference neutralization (IN)-, and decode-and-forward (DF) based relaying schemes. We demonstrated that the proposed control with multiple IRS elements outperforms the benchmark controls in terms of the achievable rate region, achievable sum-rate, and energy efficiency under same power budget. We also confirmed that the discrete phase approximation of the proposed control provides near-optimal performance with fewer bits, and the proposed control is robust under imperfect CSI condition. The proposed controls can be efficiently applied to large-scale multi-pair multihop device-to-device and machine-type device communications in the interference-limited or low-powered dense networks of 5G and 6G environments.

Mobile-edge computing (MEC) and nonorthogonal multiple access (NOMA) have been regarded as promising technologies for beyond fifth-generation (B5G) and sixth-generation (6G) networks. This study aims to reduce the computational overhead (weighted sum of consumed energy and latency) in a NOMA-assisted MEC network by jointly optimizing the computation offloading policy and channel resource allocation under dynamic network environments with time-varying channels. To this end, we propose a deep reinforcement learning algorithm named ACDQN that utilizes the advantages of both actor-critic and deep Q-network methods and provides low complexity. The proposed algorithm considers partial computation offloading, where users can split computation tasks so that some are performed on the local terminal while some are offloaded to the MEC server. It also considers a hybrid multiple access scheme that combines the advantages of NOMA and orthogonal multiple access to serve diverse user requirements. Through extensive simulations, it is shown that the proposed algorithm stably converges to its optimal value, provides approximately 10%, 27%, and 69% lower computational overhead than the prevalent schemes, such as full offloading with NOMA, random offloading with NOMA, and fully local execution, and achieves near-optimal performance.