The escalating complexity of deep neural networks introduces substantial challenges to deploying federated learning (FL) in resource-limited edge environments. To address these limitations, split federated learning (SFL) has emerged as a promising paradigm, alleviating client-side computational and communication burdens via strategic model splitting, and periodically aggregating client-side and server-side models consistent with the principles of FL. Nevertheless, existing SFL frameworks encounter significant performance degradation arising from data heterogeneity and imbalance, client heterogeneity, as well as constrained wireless resources. To overcome these issues, this paper introduces a novel data distribution deviation-aware split federated learning (DA-SFL) framework. DA-SFL dynamically adjusts aggregation weights according to the deviation of clients’ data distributions from a global distribution, effectively mitigating biases induced by data imbalance and heterogeneity. Furthermore, we theoretically establish the convergence bound of DA-SFL under a non-convex loss function setting, demonstrating that minimizing the data deviation in each training round enhances learning efficacy. Motivated by this, we formulate a mixed-integer nonlinear programming to optimize learning performance under long-term energy constraints. Leveraging the Lyapunov optimization framework, we decompose the problem into a series of tractable subproblems in each learning round, and propose efficient algorithms to find the client scheduling, adaptive cut layer selection, bandwidth allocation, and aggregation weighting policies. Extensive experimental evaluations conducted on Fashion-MNIST, CIFAR-10, and CINIC-10 datasets across diverse scenarios of data heterogeneity and imbalance demonstrate that DA-SFL significantly outperforms baselines regarding test accuracy, time and energy efficiency, while exhibiting notable robustness and scalability.

In massive MIMO systems, traditional channel estimation techniques often suffer from noise sensitivity and high computational complexity. Recently proposed deep supervised learning–based estimators have improved accuracy yet require large labeled datasets and exhibit poor generalization in dynamic channel conditions. Consequently, self-supervised methods have emerged, avoiding extensive label collection and enabling immediate online deployment. However, existing self-supervised frameworks typically rely on large networks with long run times, demanding substantial computational resources. In this work, we present a lightweight self-supervised channel estimation framework, Meta-ZSN2N. It first leverages a traditional estimator, then applies a specialized downsampling step, and finally refines the results via a lightweight two-layer neural network, resulting in a significantly simplified model and a substantially reduced runtime. To further accelerate online inference and boost generalization, we integrate a Meta-SGD module into our design. Simulation results indicate that our proposed lightweight method not only surpasses traditional estimators but also outperforms large learning networks with millions of parameters in terms of efficiency and adaptability.

In mobile wireless networks, data heterogeneity and resource constraints cause performance degradation in machine learning tasks on edge clients. To alleviate these issues, we propose a novel deviation-aware split federated learning (DevSFL) framework, which adopts an adaptive aggregation weight determination method for mitigating the effects of data heterogeneity across local datasets and improving overall learning performance. Leveraging Lyapunov optimization, we formulate a comprehensive optimization problem including client scheduling, cut layer selection, bandwidth allocation, and weight decisionmaking to enhance resource utilization and energy efficiency. To tackle this problem, we employ a sample average approximation based algorithm and a dichotomy method for optimizing cut layer selection and bandwidth allocation policies, respectively. Furthermore, a set expansion algorithm is employed to find the optimal client subset. Additionally, we introduce a deviationaware algorithm specifically designed to refine the weighting policy. Comparative analysis with benchmark schemes reveals that our proposed DevSFL framework not only achieves higher accuracy within fewer rounds but also significantly reduces the time required to reach a predefined accuracy level, thereby demonstrating the effectiveness of our proposed algorithms.

This study investigates the application of a simultaneous transmitting and reflecting reconfigurable intelligent surface (STAR-RIS)-aided medium-Earth-orbit (MEO) satellite network for providing both global positioning services and communication services in the urban canyons, where the direct satellite-user links are obstructed. Superposition coding (SC) and successive interference cancellation (SIC) techniques are utilized for the integrated navigation and communication (INAC) networks, and the composed navigation and communication signals are reflected or transmitted to ground users or indoor users located in urban canyons. To meet diverse application needs, navigation-oriented (NO)-INAC and communicationoriented (CO)-INAC have been developed, each tailored according to distinct power allocation factors. We then proposed two algorithms, namely navigation-prioritized-algorithm (NPA) and communication-prioritized-algorithm (CPA), to improve the navigation or communication performance by selecting the satellite with the optimized position dilution of precision (PDoP) or with the best channel gain.

In this article, we design an end-to-end digital semantic communication system to transmit semantic symbols that simultaneously facilitate image classification tasks and reconstruction tasks. By training a mutual information-assisted joint source-channel coding (MIJSCC) framework, the learned semantic representation can incorporate both pixel-level generative information for reconstruction and structural discriminative information for classification, which are obtained label-free via global and local mutual information estimation and maximization, as well as mean-square error (MSE) minimization. Then, the high-resolution semantic representation is quantized into finite constellation symbols to satisfy the hardware constraint on discrete control in practical radio frequency systems. Considering dynamic channel conditions in practical communication systems, we further design an adaptive MIJSCC (A-MIJSCC) framework with attention-based semantic enhancement (A-MIJSCC), which allows for the sequential activation of varying dimensions of the semantic representation according to channel signal-to-noise ratio. Compared to existing semantic communication frameworks that are dominated by end target and labels, the MIJSCC addresses the semi-supervised learning of intermediate semantics. Simulation results show that the proposed MIJSCC supports both image classification and reconstruction via task-agnostic semantic extraction, whose performance surpasses the benchmark frameworks. It is also demonstrated that the A-MIJSCC method facilitates the adaptive semantic transmission under varying channel conditions, which effectively reduces the transmission overhead while preserving task performance.

Owing to the heterogeneous computation and communication capabilities among clients, the synchronous model aggregation in wireless federated learning (FL) is susceptible to the straggler effect and exhibits low learning efficiency, while asynchronous aggregation encounters delayed gradients that lead to convergence errors and learning performance degradation. To address these obstacles, this work proposes an adaptive semi-asynchronous FL (ASAFL) approach to incorporate the strengths of synchronous and asynchronous FL while mitigating their inherent drawbacks. Specifically, the edge server dynamically adjusts the synchronous degree, i.e., the number of local gradients aggregated in each round, to strike a balance between learning latency and accuracy. Recognizing that data heterogeneity among clients may induce biased global model updating, we propose calibrating the global update by leveraging historical gradients received at the edge server from clients. Following that, we theoretically investigate the impact of synchronous degrees in different rounds on the convergence bound of ASAFL. The results imply that allocating more learning time to the later learning stages to increase the synchronous degree contributes to better learning performance. Based on this, we develop an adaptive synchronous degree control and resource allocation algorithm to enhance the learning performance of FL while adhering to the overall learning latency and wireless resources constraint. Numerical results on the MNIST and CIFAR-10 datasets demonstrate that the proposed approach is capable of attaining faster convergence speed and higher learning accuracy compared to the benchmark FL algorithms.

The emerging field of federated learning (FL) holds significant promise for advancing edge intelligence while preserving data privacy. However, as FL systems scale or become more heterogeneous, challenges such as spectrum scarcity and the straggler problem arise. To address these issues, this paper proposes SA-AirFed, a semi-asynchronous FL architecture compatible with Over-the-Air Computation (AirComp). We develop an efficient scheduling scheme that meets AirComp’s requirements and analyze the factors affecting convergence under the Lipschitz-Smooth condition. Building on insights from the convergence analysis, we design an adaptive algorithm that mitigates staleness from semi-asynchronous aggregation and noise from AirComp by dynamically adjusting aggregation weights, formulated as a convex quadratic programming problem. Experimental results on MNIST and CIFAR-10 demonstrate that SA-AirFed significantly reduces wall-clock training time while achieving greater robustness compared to baseline models.

As the complexity of deep neural networks escalates, traditional federated learning (FL) frameworks increasingly struggle since the training overhead of the full model is costly for resource-limited clients. In addition, the class imbalance among local datasets and client heterogeneity may lead to significant deterioration in learning performance. To address these challenges, we first propose a novel wireless split federated learning (SFL) framework to enhance learning efficiency and performance in resource-constrained networks, which adaptively splits the global model between the clients and server to alleviate the computation burden for clients. Then, we theoretically analyze how the client sampling and wireless network parameters impact on the convergence bound. Based on the analysis, we identify the extent of class imbalance that significantly impacts learning performance. Inspired by this, we formulate an optimization problem to strike a balance between latency and performance by jointly optimizing the client selection, model splitting, and bandwidth allocation policies. To solve this problem, we introduce a latency and class imbalance-aware double greedy algorithm to obtain client scheduling policy. Additionally, bisection-enabled optimal bandwidth allocation and model splitting algorithms are developed to adaptively determine bandwidth allocation and model splitting policies, respectively. Extensive experimental results demonstrate that our approach significantly reduces latency and enhances learning performance.

Wireless federated learning (FL) faces significant challenges due to limited wireless resources and unreliable channels. To cope with these challenges, this work proposes a gradient compensation-based FL approach (FL-GC), in which the edge server estimates the local gradients of transmission failure and unselected clients by first-order Taylor approximation based on previously received local gradients. We then theoretically analyze the convergence bound, which reveals that selecting clients with large local gradient staleness helps reduce the estimation error and improve learning performance. Based on this, we jointly optimize the client selection and resource allocation strategies to enhance the FL performance under resource-limited wireless networks. Simulation results under a typical data heterogeneity scenario demonstrate the efficacy of our proposed scheme in mitigating the adverse effects of unreliable transmission and limited resources. It improves 7.34% model accuracy compared to the considered benchmarks and is able to save 42.5% training time to achieve the target accuracy.

To reduce the synchronization latency in digital twin (DT)-enabled wireless edge networks, the DT migration provides an efficient roaming solution among edge servers by following users' trajectories. In this work, we formulate a joint DT migration, communication and computation resource management problem to minimize the data synchronization latency, where the time-varying network states and user mobility are considered. By decoupling edge servers under a deterministic migration strategy, we first derive the optimal communication and computation resource management policies at each server using convex optimization methods. For the DT migration problem between different servers, we transform it as a decentralized partially observable Markov decision process (Dec-POMDP). Then, we propose a novel agent-contribution-enabled multiagent reinforcement learning (AC-MARL) algorithm to enable distributed DT migration for users, in which the counterfactual baseline method is adopted to characterize the contribution of each agent and facilitate cooperation among agents. Simulation results show that the proposed DT migration scheme is able to reduce 30% data synchronization latency for users compared to the benchmark schemes.

This work proposes an adaptive semi-asynchronous federated learning (FL) approach, namely ASAFL, to incorporate the strengths of synchronous and asynchronous FL while mitigating their inherent drawbacks. Specifically, the edge server dynamically adjusts the synchronous degree, i.e., the number of local gradients aggregated in each round, to strike a balance between learning latency and accuracy. Recognizing that data heterogeneity among clients may induce biased global model updating, we propose calibrating the global update by leveraging historical gradients received at the edge server from clients. Following that, we experimentally revealed that allocating more learning time to the later learning stages to increase the synchronous degree contributes to better learning performance. Inspired by this, we develop an adaptive synchronous degree control and resource allocation algorithm to enhance the learning performance of FL while adhering to the overall learning latency and wireless resources constraint. Numerical results demonstrate that the proposed approach is capable of attaining faster convergence speed and higher learning accuracy compared to the benchmark FL algorithms.

Cooperative caching has emerged as an efficient way to alleviate backhaul traffic and enhance user experience by proactively prefetching popular videos at the network edge. However, it is challenging to achieve the optimal design of video caching, sharing, and delivery within storage-limited edge networks due to the growing diversity of videos, unpredictable video requirements, and dynamic user preferences. To address this challenge, this work explores cost-efficient cooperative video caching via video compression techniques while considering unknown video popularity. Firstly, we formulate the joint video caching, sharing, and delivery problem to capture a balance between user delay and system operative cost under unknown time-varying video popularity. To solve this problem, we develop a two-layer decentralized reinforcement learning algorithm, which effectively reduces the action space and tackles the coupling among video caching, sharing, and delivery decisions compared to the conventional algorithms. Specifically, the outer layer produces the optimal decisions for video caching and communication resource allocation by employing a multi-agent deep deterministic policy gradient algorithm. Meanwhile, the optimal video sharing and computation resource allocation are determined in each agent’s inner layer using the alternating optimization algorithm. Numerical results show that the proposed algorithm outperforms benchmarks in terms of the cache hit rate, delay of users and system operative cost, and effectively strikes a trade-off between system operative cost and users’ delay.

The data heterogeneity across clients and the limited communication resources, e.g., bandwidth and energy, are two of the main bottlenecks for wireless federated learning (FL). To tackle these challenges, we first devise a novel FL framework with partial model aggregation (PMA). This approach aggregates the lower layers of neural networks, responsible for feature extraction, at the parameter server while keeping the upper layers, responsible for complex pattern recognition, at clients for personalization. The proposed PMA-FL is able to address the data heterogeneity and reduce the transmitted information in wireless channels. Then, we derive a convergence bound of the framework under a non-convex loss function setting to reveal the role of unbalanced data size in the learning performance. On this basis, we maximize the scheduled data size to minimize the global loss function through jointly optimize the client selection, bandwidth allocation, computation and communication time division policies with the assistance of Lyapunov optimization. Our analysis reveals that the optimal time division is achieved when the communication and computation parts of PMA-FL have the same power. We also develop a bisection method to solve the optimal bandwidth allocation policy and use the set expansion algorithm to address the client scheduling policy. Compared with the benchmark schemes, the proposed PMA-FL improves 3.13% and 11.8% absolute accuracy on two typical datasets with heterogeneous data distribution settings, i.e., MINIST and CIFAR-10, respectively. In addition, the proposed joint dynamic client selection and resource management approach achieve slightly higher accuracy than the considered benchmarks, but they provide a satisfactory energy and time reduction: 29% energy or 20% time reduction on the MNIST; and 25% energy or 12.5% time reduction on the CIFAR-10.

In this paper, we design a resource block (RB) oriented power pool (PP) for semi-grant-free non-orthogonal multiple access (SGF-NOMA) in the presence of residual errors resulting from imperfect successive interference cancellation (SIC). In the proposed method, the BS allocates one orthogonal RB to each grant-based (GB) user, and determines the acceptable received power from grant-free (GF) users and calculates a threshold against this RB for broadcasting. Each GF user as an agent, tries to find the optimal transmit power and RB without affecting the quality-of-service (QoS) and ongoing transmission of the GB user. To this end, we formulate the transmit power and RB allocation problem as a stochastic Markov game to design the desired PPs and maximize the long-term system throughput. The problem is then solved using multi-agent (MA) deep reinforcement learning algorithms, such as double deep Q networks (DDQN) and Dueling DDQN due to their enhanced capabilities in value estimation and policy learning, with the latter performing optimally in environments characterized by extensive states and action spaces. The agents (GF users) undertake actions, specifically adjusting power levels and selecting RBs, in pursuit of maximizing cumulative rewards (throughput). Simulation results indicate computational scalability and minimal signaling overhead of the proposed algorithm with notable gains in system throughput compared to existing SGF-NOMA systems. We examine the effect of SIC error levels on sum rate and user transmit power, revealing a decrease in sum rate and an increase in user transmit power as QoS requirements and error variance escalate. We demonstrate that PPs can benefit new (untrained) users joining the network and outperform conventional SGF-NOMA without PPs in spectral efficiency.

Non-orthogonal multiple access (NOMA) is capable of serving different numbers of users in the same time-frequency resource element, and this feature can be leveraged to carry additional information. In the orthogonal frequency division multiplexing (OFDM) system, a novel enhanced NOMA scheme called NOMA with informative envelope (NOMA-IE) is proposed to explore extra flexibility from the envelope of NOMA signals. In this scheme, data bits are conveyed by the quantified signal envelope in addition to classic signal constellations. The subcarrier activation patterns of different users are jointly decided by the envelope former at the transmitter of NOMA-IE. At the receiver, successive interference cancellation (SIC) is employed, and the envelope detection coefficient is introduced to eliminate the error floor. Theoretical expressions of spectral efficiency, energy efficiency, and detection complexity are provided first. Then, considering the binary phase shift keying modulation, the block error rate and bit error rate are derived based on the two-subcarrier element. The analytical results reveal that the SIC error and the index error are the main factors degrading the error performance. The numerical results demonstrate the superiority of the NOMA-IE over the OFDM and OFDM-NOMA in terms of the error rate performance when all the schemes have the same spectral efficiency and energy efficiency.

Federated learning (FL) is an efficient and privacy-preserving distributed learning paradigm that enables massive edge devices to train machine learning models collaboratively. Although various communication schemes have been proposed to expedite the FL process in resource-limited wireless networks, the unreliable nature of wireless channels was less explored. In this work, we propose a novel FL framework, namely FL with gradient recycling (FL-GR), which recycles the historical gradients of unscheduled and transmission-failure devices to improve the learning performance of FL. To reduce the hardware requirements for implementing FL-GR in the practical network, we develop a memory-friendly FL-GR that is equivalent to FL-GR but requires low memory of the edge server. We then theoretically analyze how the wireless network parameters affect the convergence bound of FL-GR, revealing that minimizing the average square of local gradients’ staleness (AS-GS) helps improve the learning performance. Based on this, we formulate a joint device scheduling, resource allocation and power control optimization problem to minimize the AS-GS for global loss minimization. To solve the problem, we first derive the optimal power control policy for devices and transform the AS-GS minimization problem into a bipartite graph matching problem. Through detailed analysis, we further transform the bipartite matching problem into an equivalent linear program which is convenient to solve. Extensive simulation results on three real-world datasets (i.e., MNIST, CIFAR-10, and CIFAR-100) verified the efficacy of the proposed methods. Compared to the FL algorithms without gradient recycling, FL-GR is able to achieve higher accuracy and fast convergence speed. In addition, the proposed device scheduling and resource allocation algorithm also outperforms the benchmarks in accuracy and convergence speed.

Existing device scheduling works in wireless federated learning (FL) mainly focused on selecting the devices with maximum gradient norm or loss function and require all devices to perform local training in each round. This may produce extra training costs and schedule devices with similar data statistics, thus degrading learning performance. To mitigate these problems, we first theoretically characterize the convergence behaviour of the considered FL system, finding that the learning performance is degraded by the difference between the aggregated gradient of scheduled devices and the full participation gradient. Inspired by this, we propose to find a subset of representative devices and the corresponding pre-device stepsizes to approximate the full participation aggregated gradient. Considering the limited wireless bandwidth, we formulate a problem to capture the trade-off between representativity and latency by optimizing device scheduling and bandwidth allocation policies. Our analysis reveals optimal bandwidth allocation is achieved when all scheduled devices have the same latency. Then, by proving the non-monotone submodularity of the problem, we develop a double greedy algorithm to solve the device scheduling policy. To avoid the local training of unscheduled devices, we utilize the historical gradient information of devices to estimate the current gradient for device scheduling design. Compared to existing scheduling algorithms, the proposed representativity-aware device scheduling algorithm improves 6.7% and 4.02% accuracies on two typical datasets under heterogeneous local data distributions, i.e., MNIST and CIFAR-10, respectively. In addition, the proposed latency- and representativity-aware scheduling algorithm saves over 16% and 12% training time for MNIST and CIFAR-10 datasets than the scheduling algorithms based on either latency and representativity individually.

Most existing wireless federated learning (FL) studies focused on homogeneous model settings where devices train identical local models. In this setting, the devices with poor communication and computation capabilities may delay the global model update and degrade the performance of FL. Moreover, in the homogenous model settings, the scale of the global model is restricted by the device with the lowest capability. To tackle these challenges, this work proposes an adaptive model pruning-based FL (AMP-FL) framework, where the edge server dynamically generates sub-models by pruning the global model for devices’ local training to adapt their heterogeneous computation capabilities and time-varying channel conditions. Since the involvement of diverse structures of devices’ sub-models in the global model updating may negatively affect the training convergence, we propose compensating for the gradients of pruned model regions by devices’ historical gradients. We then introduce an age of information (AoI) metric to characterize the staleness of local gradients and theoretically analyze the convergence behaviour of AMP-FL. The convergence bound suggests scheduling devices with large AoI of gradients and pruning the model regions with small AoI for devices to improve the learning performance. Inspired by this, we define a new objective function, i.e., the average AoI of local gradients, to transform the inexplicit global loss minimization problem into a tractable one for device scheduling, model pruning, and resource block (RB) allocation design. Through detailed analysis, we derive the optimal model pruning strategy and transform the RB allocation problem into equivalent linear programming that can be effectively solved. Experimental results demonstrate the effectiveness and superiority of the proposed approaches. The proposed AMP-FL is capable of achieving 1.9x and 1.6x speed up for FL on MNIST and CIFAR-10 datasets in comparison with the FL schemes with homogeneous model settings.

In this paper, a stochastic geometry based analytical framework is proposed for secure simultaneous transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) assisted non-orthogonal multiple access (NOMA) transmissions, where legitimate users (LUs) and eavesdroppers are randomly distributed. Both the time-switching protocol (TS) and energy splitting (ES) protocol are considered for the STAR-RIS. To characterize system performance, the channel statistics are first provided, and the Gamma approximation is adopted for general cascaded κ-μ fading. Afterward, the closed-form expressions for both the secrecy outage probability (SOP) and average secrecy capacity (ASC) are derived. To obtain further insights, the asymptotic performance for the secrecy diversity order and the secrecy slope are deduced. The theoretical results show that 1) the secrecy diversity orders of the strong LU and the weak LU depend on the path loss exponent and the distribution of the received signal-to-noise ratio, respectively; 2) the secrecy slope of the ES protocol achieves the value of one, higher than the slope of the TS protocol which is the mode operation parameter of TS. The numerical results demonstrate that: 1) there is an optimal STAR-RIS mode operation parameter to maximize the secrecy performance; 2) the STAR-RIS-NOMA significantly outperforms the STAR-RIS-orthogonal multiple access.

This paper investigates reconfigurable intelligent surface (RIS) aided multi-cell non-orthogonal multiple access (NOMA) networks with stochastic geometry methods. Under Rayleigh and Nakagami-m fading channels, we provide two types of approximate channel models to depict RIS channels, i.e., the N-fold convolution model and the curve fitting model. The analysis reveals that the N-fold convolution model is accurate and tractable when ignoring inter-cell interference, while the curve fitting model can evaluate the impact of inter-cell interference with a small error. The N-fold convolution model provides accurate diversity orders compared to other existing approaches such as the central limit model. Based on these channel models, we derive the closed-form analytical and asymptotic expressions of coverage probabilities and ergodic rates for two paired NOMA users. The analytical results demonstrate that: i) When we ignore inter-cell interference, the diversity order of the typical user is equal to the number of Rayleigh fading channels; and ii) For Nakagami-m fading channels with coefficientm, the diversity order is equal tomtimes of the channel number. Numerical results show that: i) RISs are capable of enhancing the coverage performance and ergodic rates of the proposed network; and ii) RISs provide extra flexibility for NOMA decoding orders.

For wireless federated learning (FL), this work proposes an adaptive model pruning-based FL (AMP-FL) frame-work, where the edge server dynamically generates sub-models by pruning the global model to adapt devices' heterogeneous computation capabilities and time-varying wireless channel conditions. To mitigate the negative effect of different structures of sub-models on learning convergence, this work designs a new compensating strategy for the pruned regions of sub-models via historical gradients. Since the freshness of gradients dominates the convergence speed, this work also defines an age of information (AoI) metric to characterize the staleness of the regions of the local gradients. Based on the compensating strategy, we formulate a joint device scheduling, model pruning, and resource block allocation optimization problem to minimize the average AoI for local gradients. To solve this problem, we theoretically derive an optimal model pruning scheme. After that, we transform the original problem into equivalent linear programming that can be solved with polynomial time complexity. Simulation results on the CIFAR-IO dataset show that the proposed AMP-FL outperforms the benchmark schemes with faster convergence speed and over 7% learning accuracy improvement.

An analytical framework for physical layer security in simultaneous transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) assisted non-orthogonal multiple access (NOMA) transmissions is proposed, where legitimate users and eavesdroppers are randomly deployed. To characterize system performance, the channel statistics are first provided, and the Gamma approximation is adopted for general cascaded$\kappa-\mu$fading. Afterwards, the energy splitting (ES) protocol is considered and closed-form expressions of average secrecy capacity are derived. To obtain further insights, the asymptotic secrecy slope is deduced. The theoretical results show that the secrecy slope of the ES protocol is one. The numerical results demonstrate that: 1) there is an optimal resource allocation ratio of STAR-RIS to maximize the system performance; 2) the STAR-RIS-aided NOMA significantly outperforms the STAR-RIS-aided orthogonal multiple access.

In the orthogonal frequency division multiplexing with index modulation (OFDM-IM) framework, we propose an enhanced non-orthogonal multiple access (NOMA) scheme called NOMA with informative envelope (NOMA-IE) to explore the flexibility of signal envelope. In this scheme, data bits are conveyed by the quantified signal envelope in addition to classic signal constellations. Subcarrier activation patterns of different users are jointly decided by the envelope former at the transmitter. At the receiver, successive interference cancellation (SIC) is employed, and we introduce the envelope detection coefficient to eliminate the detection error floor. Considering binary phase shift keying, we derive the asymptotic bit error rate (BER). Analytical results reveal that the imperfect SIC is the main factor degrading the error performance. Numerical results demonstrate the superiority of the NOMA-IE over the OFDM and OFDM-NOMA.

The conventional model aggregation-based federated learning (FL) approaches require all local models to have the same architecture and fail to support practical scenarios with heterogeneous local models. Moreover, the frequent model exchange is costly for resource-limited wireless networks since modern deep neural networks usually have over-million parameters. To tackle these challenges, we first propose a novel knowledge-aided FL (KFL) framework, which aggregates light high-level data features, namely knowledge, in the per-round learning process. The KFL allows devices to design their machine learning models independently and reduces the communication overhead in the training process. We then experimentally show that different temporal device scheduling patterns lead to considerably different learning performance. With this insight, we formulate a stochastic optimization problem for joint device scheduling and bandwidth allocation under limited devices' energy budgets and develop an efficient online algorithm to achieve an energy-learning trade-off in the learning process. Experimental results on the CIFAR-10 dataset show that the proposed KFL can reduce over 87% communication overhead while achieving better learning performance than the baselines. In addition, the proposed device scheduling algorithm converges faster than benchmark scheduling schemes.

This work aims to address two of the main challenges for federated learning (FL), i.e., the limited communication resources and the data heterogeneity across devices. To this end, we first devise a novel FL framework with partial model aggregation (PMA), which only aggregates the lower layers of neural networks responsible for feature extraction while the upper layers corresponding to complex pattern recognition remain at devices for personalization. This design is able to address the data heterogeneity and reduce the transmitted information in wireless channels. Then, we maximize the scheduled data sample volume by joint optimizing the device scheduling, bandwidth allocation, computation and communication time division. Specifically, our analysis reveals that the optimal time division is achieved when the communication and computation parts of PMA-FL have the same power. We also develop a bisection method to solve the optimal bandwidth allocation policy and use the set expansion algorithm to address the optimal device scheduling. Experimental results on the CIFAR-10 dataset show that the proposed PMA-FL improves 11.6% accuracy compared with the state-of-art benchmarks, and the proposed joint dynamic device scheduling and resource optimization approach achieves slightly higher accuracy than the considered benchmarks but reduced 25% energy or 12.5% time budgets.

An adaptive next generation multiple access (NGMA) downlink scheme is provided, where non-orthogonal multiple access (NOMA) and space division multiple access (SDMA) users are served with the same orthogonal time and frequency resource to address the energy constraints and massive connectivity issues of Internet-of-Things networks. Based on this scheme, the long-term power-constrained sum rate maximization problem is investigated, where beamforming, power allocation, and user clustering are jointly optimized, subject to a long-term total power constraint. To solve the formulated problem, a spatial correlation-based user clustering approach is proposed and a resource allocation algorithm is designed based on the trust region policy optimization (TRPO) algorithm, which demonstrates stable convergence under large learning rates. Numerical results verify that the sum rate of the proposed NGMA scheme outperforms the conventional NOMA and SDMA schemes. Moreover, the spatial correlation-based clustering algorithm achieves an increasing sum rate gain compared to the channel correlation-based baseline algorithm as the spatial correlation in the channel model increases.

How to tackle the unreliability in wireless channels is critical for federated learning (FL). To solve this problem, we propose a novel FL framework, namely FL with gradient recycling (FL-GR), which recycles the historical gradients of unscheduled and transmission-failure devices to improve the learning performance of FL. Based on the proposed FL-GR, we theoretically analyze how the wireless network parameters affect the convergence bound of FL-GR, revealing that scheduling devices with large staleness and increasing their transmit power in each round helps improve learning performance. Simulation results on MNIST and CIFAR-10 show that FL-GR is able to achieve higher accuracy and fast convergence speed than conventional FL algorithms without gradient recycling.

Coverage and capacity are the important metrics for performance evaluation in wireless networks, while the coverage and capacity have several conflicting relationships, e.g. high transmit power contributes to large coverage but high inter-cell interference reduces the capacity performance. Therefore, in order to strike a balance between the coverage and capacity, a novel model is proposed for the coverage and capacity optimization of simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RISs) assisted networks. To solve the coverage and capacity optimization (CCO) problem, a machine learning-based multi-objective optimization algorithm, i.e., the multi-objective proximal policy optimization (MO-PPO) algorithm, is proposed. In this algorithm, a loss function-based update strategy is the core point, which is able to calculate weights for both loss functions of coverage and capacity by a min-norm solver at each update. The numerical results demonstrate that the investigated update strategy outperforms the fixed weight-based MO algorithms.

The conventional model aggregation-based federated learning (FL) approach requires all local models to have the same architecture, which fails to support practical scenarios with heterogeneous local models. Moreover, the frequent model exchange is costly for resource-limited wireless networks since modern deep neural networks usually have over a million parameters. To tackle these challenges, we first propose a novel knowledge-aided FL (KFL) framework, which aggregates light high-level data features, namely knowledge, in the per-round learning process. This framework allows devices to design their machine-learning models independently and reduces the communication overhead in the training process. We then theoretically analyze the convergence bound of the proposed framework under a non-convex loss function setting, revealing that scheduling more data volume in each round helps to improve the learning performance. In addition, large data volume should be scheduled in early rounds if the total scheduled data volume during the entire learning course is fixed. Inspired by this, we define a new objective function, i.e., the weighted scheduled data sample volume, to transform the inexplicit global loss minimization problem into a tractable one for device scheduling, bandwidth allocation, and power control. To deal with unknown time-varying wireless channels, we transform the considered problem into a deterministic problem for each round with the assistance of the Lyapunov optimization framework. Then, we derive the optimal bandwidth allocation and power control solution by convex optimization techniques. We also develop an efficient online device scheduling algorithm to achieve an energy-learning trade-off in the learning process. Experimental results on two typical datasets (i.e., MNIST and CIFAR-10) under highly heterogeneous local data distributions show that the proposed KFL is capable of reducing over 99% communication overhead while achieving better learning performance than the conventional model aggregation-based algorithms. In addition, the proposed device scheduling algorithm converges faster than the benchmark scheduling schemes.

Metaverse aims for building a fully immersive virtual shared space, where the users are able to engage in various activities. To successfully deploy the service for each user, the Metaverse service provider and network service provider generally localise the user first and then support the communication between the base station (BS) and the user. A reconfigurable intelligent surface (RIS) is capable of creating a reflected link between the BS and the user to enhance line-of-sight. Furthermore, the new key performance indicators (KPIs) in Metaverse, such as its energy-consumption-dependent total service cost and transmission latency, are often overlooked in ultra-reliable low latency communication (URLLC) designs, which have to be carefully considered in next-generation URLLC (xURLLC) regimes. In this paper, our design objective is to jointly optimise the transmit power, the RIS phase shifts, and the decoding error probability to simultaneously minimise the total service cost and transmission latency and approach the Pareto Front (PF). We conceive a twin-stage central controller, which aims for localising the users first and then supports the communication between the BS and users. In the first stage, we localise the Metaverse users, where the stochastic gradient descent (SGD) algorithm is invoked for accurate user localisation. In the second stage, a meta-learning-based position-dependent multi-objective soft actor and critic (MO-SAC) algorithm is proposed to approach the PF between the total service cost and transmission latency and to further optimise the latency-dependent reliability. Our numerical results demonstrate that 1) The proposed solution strikes a tradeoff between the total service cost and transmission latency, which provides a candidate group of optimal solutions for diverse practical scenarios. 2) The proposed meta-learning-based MO-SAC algorithm is capable of adaption to new wireless environments, compared to the benchmarkers. 3) The approximate PF depicted discovered the relationships among the KPIs for the Metaverse, which provides guidelines for its deployment.

Simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RISs) is a promising passive device that contributes to full-space coverage via transmitting and reflecting the incident signal simultaneously. As a new paradigm in wireless communications, how to analyze the coverage and capacity performance of STAR-RISs becomes essential but challenging. To solve the coverage and capacity optimization (CCO) problem in STAR-RIS-assisted networks, a multi-objective proximal policy optimization (MO-PPO) algorithm is proposed to handle long-term effects. To strike a balance between each objective, the MO-PPO algorithm provides a set of optimal solutions to approach a Pareto front (PF), where the solution on the approximate PF is regarded as an optimal result. Moreover, in order to improve the performance of the MO-PPO algorithm, two update strategies, i.e., action-value-based update strategy (AVUS) and loss function-based update strategy (LFUS), are investigated. For the AVUS, the improved point is to integrate the action values of both coverage and capacity and then update the loss function. For the LFUS, the improved point is only to assign dynamic weights for both loss functions of coverage and capacity, while the weights are calculated by a min-norm solver at every update. The numerical results demonstrated that the investigated update strategies outperform the fixed weights MO optimization algorithms in different cases, which include a different number of sample grids, the number of STAR-RISs, the number of elements in the STAR-RISs, and the size of STAR-RISs. Additionally, the STAR-RIS-assisted networks achieve better performance than conventional wireless networks without STAR-RISs. Moreover, with the same bandwidth, a millimetre wave is able to provide higher capacity than sub-6 GHz, but at a cost of smaller coverage.

The new concept of semi-integrated-sensing-and-communication (Semi-ISaC) is proposed for next-generation cellular networks. Compared to the state-of-the-art, where the total bandwidth is used for integrated sensing and communication (ISaC), the proposed Semi-ISaC framework provides more freedom as it allows that a portion of the bandwidth is exclusively used for either wireless communication or radar detection, while the rest is for ISaC transmission. To enhance the bandwidth efficiency (BE), we investigate the evolution of Semi-ISaC networks from orthogonal multiple access (OMA) to non-orthogonal multiple access (NOMA). First, we evaluate the performance of an OMA-based Semi-ISaC network. As for the communication signals, we investigate both the outage probability (OP) and the ergodic rate. As for the radar echoes, we characterize the ergodic radar estimation information rate (REIR). Then, we investigate the performance of a NOMA-based Semi-ISaC network, including the OP and the ergodic rate for communication signals and the ergodic REIR for radar echoes. The diversity gains of OP and the high signal-to-noise ratio (SNR) slopes of the ergodic REIR are also evaluated as insights. The analytical results indicate that: 1) Under a two-user NOMA-based Semi-ISaC scenario, the diversity order of the near-user is equal to the coefficient of the Nakagami-${m}$fading channels ($m$), while that of the far-user is zero; and 2) The high-SNR slope for the ergodic REIR is based on the ratio of the radar signal’s duty cycle to the pulse duration. Our simulation results show that: 1) Semi-ISaC has better channel capacity than the conventional ISaC; and 2) The NOMA-based Semi-ISaC has better channel capacity than the OMA-based Semi-ISaC.

A novel reconfigurable intelligent surface-aided multi-robot network is proposed, where multiple mobile robots are served by an access point (AP) through non-orthogonal multiple access (NOMA). The goal is to maximize the sum-rate of whole trajectories for the multi-robot system by jointly optimizing trajectories and NOMA decoding orders of robots, phase-shift coefficients of the RIS, and the power allocation of the AP, subject to predicted initial and final positions of robots and the quality of service (QoS) of each robot. To tackle this problem, an integrated machine learning (ML) scheme is proposed, which combines long short-term memory (LSTM)-autoregressive integrated moving average (ARIMA) model and dueling double deep Q-network ($\text{D}^{3}$QN) algorithm. For initial and final position prediction for robots, the LSTM-ARIMA is able to overcome the problem of gradient vanishment of non-stationary and non-linear sequences of data. For jointly determining the phase shift matrix and robots’ trajectories,$\text{D}^{3}$QN is invoked for solving the problem of action value overestimation. Based on the proposed scheme, each robot holds an optimal trajectory based on the maximum sum-rate of a whole trajectory, which reveals that robots pursue long-term benefits for whole trajectory design. Numerical results demonstrated that: 1) LSTM-ARIMA model provides high accuracy predicting model; 2) The proposed$\text{D}^{3}$QN algorithm can achieve fast average convergence; and 3) RIS-NOMA networks have superior network performance compared to RIS-aided orthogonal counterparts.

Semi-grant-free non-orthogonal multiple access (SGF-NOMA) is a potential paradigm to support massive connec-tivity for the short packets Internet of things (IoT) applications while satisfying the undistracted transmission requirements of primary IoT users. However, resource allocation in SGF-NOMA is more challenging due to the sporadic traffic of grant-free (GF) users and the need to satisfy the quality of service (QoS) requirements of grant-based (GB) users. The GF users access and choose resources at random, resulting in frequent power collisions and decoding failures at the base station (BS). This paper develops a general learning framework that enables GF users to learn from historical information to avoid power collisions. We utilize a hybrid multi-agent deep reinforcement learning (hMA-DRL) framework to maximize the connectivity and enhance the number of successful decoded users at the BS. The numerical results show that the proposed scheme achieves a solution near to the optimal one and increases the successful decoded users by 42.38% as compared to the benchmark scheme. The considered algorithm performs well with an increasing number of users as compared to the competitive and cooperative MA-DRL algorithms.

In this paper, we propose a novel distributed resource allocation mechanism for semi-grant-free non-orthogonal multiple access (SGF-NOMA) transmission to maximize the network throughput, where multi-agent deep reinforcement learning with prioritized experience replay (PER) is employed. We design a centralized training framework and decentralized decision making to increase the flexibility of the proposed scheme. More specifically, each grant-free user as an "agent" learns the dynamics of the environment and makes its decisions independently in a decentralized manner. No heavy information exchange is needed to find the optimal transmit power and sub-channel that maximize the throughput. Numerical results show that the proposed algorithm with PER enhances the learning efficiency compared to the algorithm with conventional replay buffer and outperforms the existing scheme with a 12% throughput increase.

This paper investigates non-orthogonal multiple access (NOMA) assisted integrated sensing and communication (ISaC) networks. Compared to the conventional ISaC networks, where the total bandwidth is used for both the radar detection and wireless communications, the proposed Semi-ISaC networks allow that a portion of bandwidth is used for ISaC and the rest of the bandwidth is only utilized for wireless communications. We first derive the analytical expressions of the outage probability for the communication signals, including the signals for the radar target and the communication transmitter. Additionally, we derive the analytical expressions of the ergodic radar estimation information rate (REIR) for the radar echoes. The simulation results show that 1) NOMA ISaC has better spectrum efficiency than the conventional ISaC; and 2) The REIR is enhanced when we enlarge the density of pulses.

Reconfigurable intelligent surface (RIS) has been regarded as a promising technology since it has ability to create the favorable channel conditions. This paper investigates the secrecy performance of RIS aided non-orthogonal multiple access (NOMA) networks, where the internal eavesdropping scenario is taken into consideration. More specifically, novel closed-form and asymptotic expressions of secrecy outage probability for the k-th user are derived. According to the analytical results, the secrecy diversity orders at users are acquired in the high signal-to-noise ratio region. Simulation results show that the applying of RIS in NOMA networks can remarkably improve the performance of secrecy outage behaviour and secrecy system throughput compared to RIS aided orthogonal multiple access networks.

In multi-access edge computing (MEC), most existing task software caching works focus on statically caching data at the network edge, which may hardly preserve high reusability due to the time-varying user requests in practice. To this end, this work considers dynamic task software caching at the MEC server to assist users’ task execution. Specifically, we formulate a joint task software caching update (TSCU) and computation offloading (COMO) problem to minimize users’ energy consumption while guaranteeing delay constraints, where the limited cache size and computation capability of the MEC server, as well as the time-varying task demand of users are investigated. This problem is proved to be non-deterministic polynomial-time hard, so we transform it into two sub-problems according to their temporal correlations, i.e., the real-time COMO problem and the Markov decision process-based TSCU problem. We first model the COMO problem as a multi-user game and propose a decentralized algorithm to address its Nash equilibrium solution. We then propose a double deep Q-network (DDQN)-based method to solve the TSCU policy. To reduce the computation complexity and convergence time, we provide a new design for the deep neural network (DNN) in DDQN, named state coding and action aggregation (SCAA). In SCAA-DNN, we introduce a dropout mechanism in the input layer to code users’ activity states. Additionally, at the output layer, we devise a two-layer architecture to dynamically aggregate caching actions, which is able to solve the huge state-action space problem. Simulation results show that the proposed solution outperforms existing schemes, saving over 12% energy, and converges with fewer training episodes.

The simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) is capable of providing full-space coverage of smart radio environments. This work investigates STAR-RIS aided downlink non-orthogonal multiple access (NOMA) multi-cell networks, where the energy of incident signals at STAR-RISs is split into two portions for transmitting and reflecting. We first propose a fitting method to model the distribution of composite small-scale fading power as the tractable Gamma distribution. Then, a unified analytical framework based on stochastic geometry is provided to capture the random locations of RIS-RISs, base stations (BSs), and user equipments (UEs). Based on this framework, we derive the coverage probability and ergodic rate of both the typical UE and the connected UE. In particular, we obtain closed-form expressions of the coverage probability in interference-limited scenarios. We also deduce theoretical expressions in conventional RIS aided networks for comparison. The analytical results show that optimal energy splitting coefficients of STAR-RISs exist to simultaneously maximize the system coverage and ergodic rate. The numerical results demonstrate that: 1) STAR-RISs are able to meet different demands of UEs located on different sides; 2) STAR-RISs with appropriate energy splitting coefficients outperform conventional RISs in the coverage and the rate performance.

With the rapid growth of data traffic in cellular networks, edge caching has become an emerging technology for traffic offloading. We investigate the caching placement and content delivery in cache-enabling cellular networks. To cope with the time-varying content popularity and user location in practical scenarios, we formulate a long-term joint dynamic optimization problem of caching placement and user association for minimizing the content delivery delay which considers both content transmission delay and content update delay. To solve this challenging problem, we decompose the optimization problem into two sub-problems, the user association sub-problem in a short time scale and the caching placement in a long time scale. Specifically, we propose a low complexity user association algorithm for a given caching placement in the short time scale. Then we develop a deep deterministic policy gradient based caching placement algorithm which involves the short time-scale user association decisions in the long time scale. Finally, we propose a joint user association and caching placement algorithm to obtain a sub-optimal solution for the proposed problem. We illustrate the convergence and performance of the proposed algorithm by simulation results. Simulation results show that compared with the benchmark algorithms, the proposed algorithm reduces the long-term content delivery delay in dynamic networks effectively.

With the exponential growth of data traffic in wireless networks, edge caching has been regarded as a promising solution to offload data traffic and alleviate backhaul congestion, where the contents can be cached by an unmanned aerial vehicle (UAV) and user terminal (UT) with local data storage. In this article, a cooperative caching architecture of UAV and UTs with scalable video coding (SVC) is proposed, which provides the high transmission rate content delivery and personalized video viewing qualities in hotspot areas. In the proposed cache-enabling UAV-D2D networks, we formulate a joint optimization problem of UT caching placement, UAV trajectory, and UAV caching placement to maximize the cache utility. To solve this challenging mixed integer nonlinear programming problem, the optimization problem is decomposed into three sub-problems. Specifically, we obtain UT caching placement by a many-to-many swap matching algorithm, then obtain the UAV trajectory and UAV caching placement by approximate convex optimization and dynamic programming, respectively. Finally, we propose a low complexity iterative algorithm for the formulated optimization problem to improve the system capacity, fully utilize the cache space resource, and provide diverse delivery qualities for video traffic. Simulation results reveal that: i) the proposed cooperative caching architecture of UAV and UTs obtains larger cache utility than the cache-enabling UAV networks with same data storage capacity and radio resource; ii) compared with the benchmark algorithms, the proposed algorithm improves cache utility and reduces backhaul offloading ratio effectively.

By activating blocked users and altering successive interference cancellation (SIC) sequences, reconfigurable intelligent surfaces (RISs) become promising for enhancing non-orthogonal multiple access (NOMA) systems. To evaluate the benefits between RISs and NOMA, a downlink RIS-aided multi-cell-NOMA network is investigated via stochastic geometry. We first introduce the unique path loss model for RIS reflecting channels. Then, we evaluate the angle distributions based on a Poisson cluster process (PCP) model, which theoretically demonstrates that the angles of incidence and reflection are uniformly distributed. Additionally, we derive closed-form analytical and asymptotic expressions for coverage probabilities of the paired NOMA users. Lastly, we derive the analytical expressions of the ergodic rate for both of the paired NOMA users and calculate the asymptotic expressions for the typical user. The analytical results indicate that 1) the achievable rates reach an upper limit when the length of RIS increases; 2) exploiting RISs can enhance the path loss intercept to improve the performance without influencing the bandwidth. The simulation results show that 1) RIS-aided networks have superior performance than the networks without RISs; and 2) the SIC order in NOMA systems can be altered since RISs are able to change the channel quality of NOMA users.

In this paper, we propose a deep state-action-reward-state-action (SARSA)$\lambda $learning approach for optimising the uplink resource allocation in non-orthogonal multiple access (NOMA) aided ultra-reliable low-latency communication (URLLC). To reduce the mean decoding error probability in time-varying network environments, this work designs a reliable learning algorithm for providing a long-term resource allocation, where the reward feedback is based on the instantaneous network performance. With the aid of the proposed algorithm, this paper addresses three main challenges of the reliable resource sharing in NOMA-URLLC networks: 1) user clustering; 2) Instantaneous feedback system; and 3) Optimal resource allocation. All of these designs interact with the considered communication environment. Lastly, we compare the performance of the proposed algorithm with conventional Q-learning and SARSA Q-learning algorithms. The simulation outcomes show that: 1) Compared with the traditional Q learning algorithms, the proposed solution is able to converge within 200 episodes for providing as low as$10^{-2}$long-term mean error; 2) NOMA assisted URLLC outperforms traditional OMA systems in terms of decoding error probabilities; and 3) The proposed feedback system is efficient for the long-term learning process.

Compared with the conventional reconfigurable intelligent surfaces (RIS), simultaneous transmitting and reflecting intelligent omini-surfaces (STAR-IOSs) are able to achieve 360° coverage “smart radio environments”. By splitting the energy or altering the active number of STAR-IOS elements, STAR-IOSs provide high flexibility of successive interference cancellation (SIC) orders for non-orthogonal multiple access (NOMA) systems. Based on the aforementioned advantages, this paper investigates a STAR-IOS-aided downlink NOMA network with randomly deployed users. We first propose three tractable channel models for different application scenarios, namely the central limit model, the curve fitting model, and the M-fold convolution model. More specifically, the central limit model fits the scenarios with large-size STAR-IOSs while the curve fitting model is extended to evaluate multi-cell networks. However, these two models cannot obtain accurate diversity orders. Hence, we figure out the M-fold convolution model to derive accurate diversity orders. We consider three protocols for STAR-IOSs, namely, the energy splitting (ES) protocol, the time switching (TS) protocol, and the mode switching (MS) protocol. Based on the ES protocol, we derive closed-form analytical expressions of outage probabilities for the paired NOMA users by the central limit model and the curve fitting model. Based on three STAR-IOS protocols, we derive the diversity gains of NOMA users by the M-fold convolution model. The analytical results reveal that the diversity gain of NOMA users is equal to the active number of STAR-IOS elements. Numerical results indicate that 1) in high signal-to-noise ratio regions, the central limit model performs as an upper bound of the simulation results, while a lower bound is obtained by the curve fitting model; 2) the TS protocol has the best performance but requesting more time blocks than other protocols; 3) the ES protocol outperforms the MS protocol as the ES protocol has higher diversity gains.

In this paper, we propose a deep state-action-reward-state-action (SARSA)$A$learning approach for optimising the uplink resource allocation in non-orthogonal multiple access (NOMA) aided ultra-reliable low-latency communication (URLLC). To reduce the mean decoding error probability in time-varying network environments, this work designs a reliable learning algorithm for providing a long-term resource allocation, where the reward feedback is based on the instantaneous network performance. With the aid of the proposed algorithm, this paper addresses three main challenges of the reliable resource sharing in NOMA-URLLC networks: 1) Dynamic user clustering; 2) Instantaneous feedback system; and 3) Optimal resource allocation. All of these designs interact with the considered communication environment. The simulation outcomes show that: 1) Compared with the traditional Q learning algorithm, the proposed solution converges faster and obtains better performance; 2) NOMA assisted URLLC outperforms traditional OMA systems in terms of decoding error probabilities; and 3) The dynamic feedback system is efficient for the long-term learning process.

To achieve 360ºcoverage, we investigate a simulta-neous transmitting and reflecting reconfigurable intelligent surfaces (STAR-RIS) aided downlink non-orthogonal multiple access (NOMA) network with randomly deployed users. For different scenarios, we first derive two STAR- RIS-aided channel models, namely the central limit model and the curve fitting model. More specifically, the central limit model fits the scenarios with numerous RIS elements while the curve fitting model can be extended to multi-cell scenarios. The analytical results reveal that 1) the central limit model has closed-form expressions calculated as the error functions, and 2) the curve fitting model can be closely modeled as a Gamma distribution. We then derive the closed-form outage probability expressions for the NOMA users. Numerical results indicate that 1) the two channel models match the simulation results well in low signal-to-noise-ratio (SNR) regions and perform as boundaries in high SNR regions, 2) the central limit model performs as an upper bound of the simulation results, while a lower bound can be obtained by the curve fitting model, and 3) the both users in the NOMA pair have no error floor.

Grant-free non-orthogonal multiple access (GF-NOMA) is a potential multiple access framework for internet-of-things (IoT) networks to enhance connectivity. However, the resource allocation problem in GF-NOMA is challenging and the effectiveness of such a solution is limited due to the absence of closed-loop power control. In this paper, we design a prototype of layer-based transmit power pool by utilizing multi-agent reinforcement learning to provide open-loop power control and offload the computing tasks at the base station (BS) side. IoT users in each layer decide their own transmit power level from this layer-based power pool, instead of transmitting on the allocated sub-channel with allocated transmit power level. The proposed algorithm does not require any information exchange between IoT users and does not rely on any assistance from the BS. Numerical results confirm that the double deep Q network based GF-NOMA algorithm achieves high accuracy and finds out an accurate transmit power level for each layer. Moreover, the proposed GF-NOMA system outperforms the traditional GF with orthogonal multiple access techniques in terms of throughput.

Reconfigurable intelligent surfaces (RISs) become promising for enhancing non-orthogonal multiple access (NOMA) systems, i.e., enhancing the channel quality and altering the SIC orders. Invoked by stochastic geometry methods, we investigate the downlink coverage performance of RIS-aided multi-cell NOMA networks. We first derive the RIS-aided channel model, concluding the direct and reflecting links. The analytical results demonstrate that the RIS-aided channel model can be closely modeled as a Gamma distribution. Additionally, interference from other cells is analyzed. Lastly, we derive closed-form coverage probability expressions for the paired NOMA users. Numerical results indicate that 1) although the interference from other cells is enhanced via the RISs, the performance of the RIS-aided user still enhances since the channel quality is strengthened more obviously; and 2) the SIC order can be altered by employing the RISs since the RISs improve the channel quality of the aided user.

Non-orthogonal multiple access (NOMA) exploits the potential of the power domain to enhance the connectivity for the Internet of Things (IoT). Due to time-varying communication channels, dynamic user clustering is a promising method to increase the throughput of NOMA-IoT networks. This article develops an intelligent resource allocation scheme for uplink NOMA-IoT communications. To maximise the average performance of sum rates, this work designs an efficient optimization approach based on two reinforcement learning algorithms, namely deep reinforcement learning (DRL) and SARSA-learning. For light traffic, SARSA-learning is used to explore the safest resource allocation policy with low cost. For heavy traffic, DRL is used to handle traffic-introduced huge variables. With the aid of the considered approach, this work addresses two main problems of fair resource allocation in NOMA techniques: 1) allocating users dynamically and 2) balancing resource blocks and network traffic. We analytically demonstrate that the rate of convergence is inversely proportional to network sizes. Numerical results show that: 1) Compared with the optimal benchmark scheme, the proposed DRL and SARSA-learning algorithms have lower complexity with acceptable accuracy and 2) NOMA-enabled IoT networks outperform the conventional orthogonal multiple access based IoT networks in terms of system throughput.

Grant-free non-orthogonal multiple access (GF-NOMA) is a potential multiple access framework for short-packet internet-of-things (IoT) networks to enhance connectivity. However, the resource allocation problem in GF-NOMA is challenging due to the absence of closed-loop power control. We design a prototype of transmit power pool (PP) to provide open-loop power control. IoT users acquire their transmit power in advance from this prototype PP solely according to their communication distances. Firstly, a multi-agent deep Q-network (DQN) aided GF-NOMA algorithm is proposed to determine the optimal transmit power levels for the prototype PP. More specifically, each IoT user acts as an agent and learns a policy by interacting with the wireless environment that guides them to select optimal actions. Secondly, to prevent the Q-learning model overestimation problem, double DQN (DDQN) based GF-NOMA algorithm is proposed. Numerical results confirm that the DDQN based algorithm finds out the optimal transmit power levels that form the PP. Comparing with the conventional online learning approach, the proposed algorithm with the prototype PP converges faster under changing environments due to limiting the action space based on previous learning. The considered GF-NOMA system outperforms the networks with fixed transmission power, namely all the users have the same transmit power and the traditional GF with orthogonal multiple access techniques, in terms of throughput.

By activating blocked users and altering successive interference cancellation (SIC) sequences, reconfigurable intelligent surfaces (RISs) become promising for enhancing non-orthogonal multiple access (NOMA) systems. This work investigates the downlink performance of RIS-aided NOMA networks via stochastic geometry. We first introduce the unique path loss model for RIS reflecting channels. Then, we evaluate the angle distributions based on a Poisson cluster process (PCP) framework, which theoretically demonstrates that the angles of incidence and reflection are uniformly distributed. Lastly, we derive closed-form expressions for coverage probabilities of the paired NOMA users. Our results show that 1) RIS-aided NOMA networks perform better than the traditional NOMA networks; and 2) the SIC order in NOMA systems can be altered since RISs are able to change the channel gains of NOMA users.

In order to satisfy the requirement of high throughput in most UAV applications, the potential of integrating millimeter wave (mmWave) communications with UAV networks is explored in this paper. A tractable three-dimensional (3D) spatial model is proposed for evaluating the average downlink performance of UAV networks at mmWave bands, where the locations of UAVs and users are randomly distributed with the aid of a Poisson cluster process. Moreover, an actual 3D antenna model with the uniform planar array is deployed at all UAVs to examine the impact of both azimuth and elevation angles. Based on this framework and two typical user selection schemes, closed-form approximation equations of the evaluated coverage probability and area spectral efficiency (ASE) are derived. In a noise-limited scenario, an exact expression is provided, which theoretically demonstrates that a large scale of antenna elements is able to enhance the coverage performance. Regarding the altitude of UAVs, there exists at least one optimal height for maximizing the coverage probability. Numerical results verify the proposed insight that non-line-of-sight transmission caused by obstacles have negligible effects on the proposed system. Another interesting result is that the ASE can be maximized by optimizing both the targeted data rate and the density of UAVs.

Due to owning huge free bandwidth, millimeter-wave (mmWave) becomes a promising technique to provide ultra-low latencies and ultra-fast data rates for vehicular networks. In this paper, a practical spatial framework for mmWave-enabled vehicle-to- everything networks is proposed by utilizing stochastic geometry approach. More particularly, base stations and vehicles are modeled by a Poisson point process (PPP) and multiple type II Matern hard-core processes (MHCPs), respectively. To characterize the blockage process caused by vehicles, a novel expression is deduced to distinguish line-of-sight (LOS) and non- LOS transmission. This expression demonstrates that LOS links are independent of horizontal communication distances. Furthermore, several closed-form probability density functions of desired communication distances are derived for analyzing the generated path loss. Considering two practical user association schemes, tractable expressions for coverage probabilities are figured out. The result shows that mmWave outperforms sub-6 GHz and MHCP-based model has higher accuracy than the traditional framework with multi-PPPs.

Due to the low traffic congestion, high fuel efficiency, and comfortable travel experience, vehicular platoon systems (VPSs) become one of the most promising applications in millimeter wave (mmWave) vehicular networks. In this paper, an effective spatial framework for mmWave vehicle-to-everything (V2X) networks with VPSs is proposed by utilizing stochastic geometry approaches. Base stations (BSs) are modeled by a Poisson point process and vehicles are distributed according to multiple type II Matérn hard-core processes. To characterize the blockage process caused by vehicles, a closed-form expression is deduced to distinguish line-of-sight (LOS) and non-LOS transmission. This expression demonstrates that LOS links are independent of horizontal communication distances. Several closed-form probability density functions of the communication distance between a reference platoon and its serving transmitter (other platoons or BSs) are derived for analyzing the generated path loss. After designing three practical user association techniques, tractable expressions for coverage probabilities are figured out. Our work theoretically shows that the maximum density of VPSs exists and large antenna scales benefit the networks' coverage performance. The numerical results illustrate that platoons outperform individual vehicles in terms of road spectral efficiency and the considered system is LOS interference-limited.

For unmanned aerial vehicle (UAV) aided millimeter wave (mmWave) networks, we propose a unified three-dimensional (3D) spatial framework in this paper to model a general case that uncovered users send messages to base stations via UAVs. More specifically, the locations of transceivers in downlink and uplink are modeled through the Poisson point processes and Poisson cluster processes (PCPs), respectively. For PCPs, Matern cluster and Thomas cluster processes, are analyzed. Furthermore, both 3D blockage processes and 3D antenna patterns are introduced for appraising the effect of altitudes. Based on this unified framework, several closed-form expressions for the coverage probability in the uplink and downlink, are derived. By investigating the entire communication process, which includes the two aforementioned phases and the cooperative transmission between them, tractable expressions of system coverage probabilities are derived. Next, three practical applications in UAV networks are provided as case studies of the proposed framework. The results reveal that the impact of thermal noise and non-line-of-sight mmWave transmissions is negligible. In the considered networks, mmWave outperforms sub-6 GHz in terms of the data rate, due to the sharp direction beamforming and large transmit bandwidth. Additionally, there exists an optimal altitude of UAVs, which maximizes the system coverage probability.

We introduce clustered millimeter-wave (mmWave) networks with invoking non-orthogonal multiple access (NOMA) techniques, where the NOMA users are modeled as Poisson cluster processes and each cluster contains a base station (BS) located at the center. To provide realistic directional beamforming, an actual antenna array pattern is deployed at all BSs. We propose three distance-dependent user selection strategies to appraise the path loss impact on the performance of our considered networks. With the aid of such strategies, we derive tractable analytical expressions for the coverage probability and system throughput. Specifically, closed-form expressions are deduced under a sparse network assumption to improve the calculation efficiency. It theoretically demonstrates that the large antenna scale benefits the near user, while such influence for the far user is fluctuant due to the randomness of the beamforming. Moreover, the numerical results illustrate that: 1) the proposed system outperforms traditional orthogonal multiple access techniques and the commonly considered NOMA-mmWave scenarios with the random beamforming; 2) the coverage probability has a negative correlation with the variance of intra-cluster receivers; 3) 73 GHz is the best carrier frequency for the near user, and 28 GHz is the best choice for the far user; and 4) an optimal number of the antenna elements exists for maximizing the system throughput.

To evaluate the system-level performance of clustered unmanned aerial vehicle (UAV) networks, we apply stochastic geometry in order to establish a tractable analytical framework. The users' locations are modeled as Poisson cluster processes (PCPs) and the UAVs are assumed to hover above cluster centers at a fixed altitude. In order to enhance the generality of the analysis, two typical patterns of PCPs, namely Thomas cluster process and Matern cluster process, are analyzed. Based on this spatial framework, a unified expression for the coverage probability is derived, for both millimeter wave (mmWave) and sub-6 GHz scenarios. Theoretical and numerical results demonstrate that mmWave outperforms sub-6 GHz, and an optimal altitude of UAVs exists, which maximizes the coverage probability. This result indicates that in most cases UAVs perform better than traditional terrestrial base stations, due to their mobility.

In this paper, we introduce a clustered millimeter wave network with non-orthogonal multiple access (NOMA), where the base station (BS) is located at the center of each cluster and all users follow a Poisson Cluster Process. To provide a realistic directional beamforming, an actual antenna pattern is deployed at all BSs. We provide a nearest-random scheme, in which near user is the closest node to the corresponding BS and far user is selected at random, to appraise the coverage performance and universal throughput of our system. Novel closed- form expressions are derived under a loose network assumption. Moreover, we present several Monte Carlo simulations and numerical results, which show that: 1) NOMA outperforms orthogonal multiple access regarding the system rate; 2) the coverage probability is proportional to the number of possible NOMA users and a negative relationship with the variance of intra-cluster receivers; and 3) an optimal number of the antenna elements is existed for maximizing the system throughput.

In this paper, we consider a novel cache-enabled heterogeneous network (HetNet), where macro base stations (BSs) with traditional sub-6 GHz are overlaid by dense millimeter wave (mmWave) pico BSs. These two-tier BSs, which are modeled as two independent homogeneous Poisson Point Processes, cache multimedia contents following the popularity rank. High-capacity backhauls are utilized between macro BSs and the core server. A maximum received power strategy is introduced for deducing novel algorithms of the success probability and area spectral efficiency (ASE). Moreover, Monte Carlo simulations are presented to verify the analytical conclusions and numerical results demonstrate that: 1) the proposed HetNet is an interference limited system and it outperforms the traditional HetNets; 2) there exists an optimal pre-decided rate threshold that contributes to the maximum ASE; and 3) 73 GHz is the best mmWave carrier frequency regarding ASE due to the large antenna scale.

In this paper, we consider a novel cache-enabled heterogeneous network (HetNet), where macro base stations (BSs) with traditional sub-6 GHz are overlaid by dense millimeter wave (mmWave) pico BSs. These two-tier BSs, which are modeled as two independent homogeneous Poisson point processes, cache multimedia contents following the popularity rank. High-capacity backhauls are utilized between macro BSs and the core server. In contrast to the simplified flat-top antenna pattern analyzed in previous articles, we employ an actual antenna model with the uniform linear array at all mmWave BSs. To evaluate the performance of our system, we introduce two distinctive user association strategies: 1) maximum received power (Max-RP) scheme; and 2) maximum rate (Max-Rate) scheme. With the aid of these two schemes, we deduce new theoretical equations for success probabilities and area spectral efficiencies. Considering a special case with practical path loss laws, several closed-form expressions for coverage probabilities are derived to gain several insights. Monte Carlo simulations are presented to verify the analytical conclusions. We show that: 1) the proposed HetNet is an interference-limited system and it outperforms the traditional HetNets in terms of the success probability; 2) there exists an optimal pre-decided rate threshold that contributes to the maximum ASE; and 3) Max-Rate achieves higher success probability and ASE than Max-RP but it needs the extra information of the interference effect.

This paper investigates the performance of millimeter wave (mmWave) communications in clustered device-to-device (D2D) networks. The locations of D2D transceivers are modeled as a Poisson Cluster Process. In each cluster, devices are equipped with multiple antennas, and the active D2D transmitter (D2D-Tx) utilizes mmWave to serve one of the proximate D2D receivers. Specifically, we introduce three user association strategies: 1) uniformly distributed D2D-Tx model; 2) nearest D2D-Tx model; and 3) closest line-of-site (LOS) D2D-Tx model. To characterize the performance of the considered scenarios, we derive new analytical expressions for the coverage probability and area spectral efficiency (ASE). Additionally, in order to efficiently illustrating the general trends of our system, a closed-form lower bound for the special case interfered by intra-cluster LOS links is derived. We provide Monte Carlo simulations to corroborate the theoretical results and show that: 1) the coverage probability is mainly affected by the intra-cluster interference with LOS links; 2) there exists an optimum number of simultaneously active D2D-Txs in each cluster for maximizing ASE; and 3) the closest LOS model outperforms the other two scenarios but at the cost of extra system overhead.