resource allocation and scheduling

This study addresses the challenges of resource management in satellite networks, where high mobility, long propagation delays, and limited resources render traditional rigid approaches ineffective in balancing user rationality and system efficiency. From an integrated economic and systems perspective, this work presents the first comprehensive survey of incentive mechanisms grounded in game theory and auction theory, applied to critical domains such as communication resource allocation, computation offloading, privacy-preserving security, and multi-agent coordination. By explicitly incorporating users’ rational behaviors, the proposed unified framework significantly enhances the scalability, adaptability, and fairness of resource allocation. The paper further identifies learning-based mechanism design as a pivotal direction for achieving intelligent, efficient, and equitable resource management in future satellite networks.

This work addresses the challenge of coordinating multiple radio access technologies (RATs) in Beyond 3G heterogeneous wireless systems to meet the quality-of-service (QoS) requirements of multimedia traffic. The paper proposes a unified wireless resource management strategy that, for the first time, applies linear programming to jointly optimize RAT selection and radio resource allocation within a cross-RAT scheduling framework. By formulating a linear objective function, the approach dynamically assigns each user an optimal RAT type and corresponding resource amount, effectively satisfying multimedia QoS constraints while significantly improving overall system resource utilization efficiency.

To address dynamic resource allocation in wireless networks, this paper establishes a high-fidelity simulation environment featuring multi-antenna base stations and user equipment, and implements deep reinforcement learning (DRL) algorithms—including DQN and PPO—using the RLlib framework. It presents the first systematic comparative study of multiple DRL algorithms and learning rates on scheduling performance. Methodologically, the work innovatively incorporates non-stationary channel modeling and a multi-agent coordination mechanism to significantly enhance scheduling robustness. Experimental results demonstrate that the proposed DRL-based approach achieves a 37% improvement in spectral efficiency and a 29% reduction in average latency compared to conventional methods. Moreover, it maintains superior throughput and user fairness under time-varying channel conditions. These findings rigorously validate the effectiveness and practicality of DRL for complex wireless resource management.

This paper addresses the problem of minimizing average latency in RIS-assisted OFDM downlink systems under stochastic data arrivals. We formulate it as a Markov decision process with a hybrid action space and propose a multi-agent proximal policy optimization (PPO) framework. Our method features a novel dual-branch PPO-Θ/PPO-N architecture that decouples RIS phase-shift control from subcarrier allocation; introduces a delay-sensitive state representation based on buffer queue length and instantaneous arrival rate; designs a distributed subcarrier assignment mechanism to mitigate the curse of dimensionality; and incorporates transfer learning to accelerate convergence. Experimental results demonstrate that the proposed approach reduces average latency by up to 38.2% compared to baseline methods, while simultaneously improving spectral efficiency, user fairness, and robustness and adaptability under dynamic traffic conditions.

This work addresses the deterministic communication requirements of time-sensitive applications in Beyond 5G networks by proposing a predictive dynamic wireless resource scheduling mechanism. Integrating traffic prediction with dynamic scheduling, the approach intelligently reserves resources likely to be needed in the near future while satisfying current latency constraints, thereby overcoming the limitations of conventional semi-static scheduling. By proactively managing prediction uncertainty, the proposed scheme significantly enhances resource utilization efficiency under mixed traffic loads and diverse QoS requirements, while effectively guaranteeing bounded latency and service quality.

This work addresses energy efficiency and latency optimization in integrated sensing and communication (ISAC) systems under imperfect information. The authors jointly optimize time-slot allocation, beamforming adaptation, functionality selection, and user–target pairing to minimize energy consumption while prioritizing time savings, accounting for uncertainties arising from target dynamics, quantization errors, feedback delays, and hardware constraints. The problem is innovatively formulated as a semi-infinite nonconvex mixed-integer nonlinear program. By exploiting hidden convexity, the authors develop a structure-aware exact reformulation that equivalently transforms the problem into a globally solvable mixed-integer semidefinite program (MISDP). Simulations demonstrate that the proposed approach achieves up to 88% resource savings compared to baseline schemes and reveals strong coupling among the various resource management components.

Existing user scheduling approaches in multi-user MIMO systems predominantly rely on greedy algorithms, which struggle to achieve global resource optimization and consequently suffer from inadequate interference suppression and limited quality of service. This work proposes a global user scheduling framework based on approximate solutions to non-convex optimization problems. By jointly optimizing the subset of concurrently scheduled users in each time slot, the framework overcomes the limitations of conventional greedy strategies while accommodating diverse objective functions and resource occupancy constraints. The method is applicable to both millimeter-wave and sub-6 GHz cell-free massive MIMO scenarios and demonstrates significantly superior performance compared to existing algorithms, closely approaching the optimal results attainable by exhaustive search.

This work addresses the stringent requirements of 6G for energy efficiency, spectral efficiency, and support of diverse services such as enhanced mobile broadband (eMBB) and ultra-reliable low-latency communication (URLLC) by proposing a cross-layer optimization framework that integrates reconfigurable intelligent surfaces (RIS) with Open Radio Access Network (Open RAN) to enable dynamic, real-time intelligent RAN control. The core contributions include the first system architecture unifying RIS and Open RAN, the PandORA framework for automated deployment of deep reinforcement learning (DRL) applications, novel online reinforcement learning algorithms for network slicing, scheduling, and link adaptation, and intelligent spectrum sharing across terrestrial and non-terrestrial networks. Validated through Colosseum emulation, digital twin models, and a real-world 5G testbed, the proposed approach demonstrates significant improvements in energy efficiency, spectral efficiency, and quality of service, alongside a comprehensive system-level evaluation of RIS-assisted channels integrated into Open RAN.

This work addresses the joint optimization of user selection and resource allocation in uplink scheduling for MU-MIMO-OFDMA wireless local area networks. It proposes a novel scheduling approach based on a multi-agent framework, which uniquely integrates a pre-trained small-scale language model (xLM) into wireless resource management. By enabling collaborative, data-driven decision-making among multiple agents, the method facilitates autonomous scheduling without explicit rule-based programming. The effectiveness of the proposed scheme is validated on the AI-enhanced WiSER platform under diverse configurations of station (STA) counts and antenna setups. Experimental results demonstrate that the approach consistently outperforms existing benchmark schemes, achieving substantial gains in uplink throughput and establishing a new paradigm for intelligent wireless access.

This work addresses the challenge that conventional configured grant (CG) scheduling struggles to meet the bounded latency requirements of deterministic communication under variable traffic conditions. To overcome this limitation, the paper proposes a novel CG scheduling mechanism that integrates traffic prediction with robust optimization, explicitly incorporating prediction uncertainty into resource pre-allocation decisions for the first time. The approach dynamically adapts to the heterogeneous latency constraints of mixed traffic types while ensuring bounded end-to-end delays. By jointly optimizing resource allocation under uncertainty, the method significantly improves resource utilization without compromising timing guarantees. Extensive evaluations demonstrate that the proposed scheme maintains superior performance even in highly dynamic and diverse traffic scenarios, thereby substantially enhancing the system’s capability to support deterministic services.

This work addresses the joint wireless resource management challenge in uplink hybrid beamforming systems, where constraints on the number of radio-frequency chains and per-user power-time allocation complicate system optimization. To tackle this, the paper proposes a low-complexity heuristic algorithm that jointly optimizes, for each time slot, analog beam selection, user scheduling, power allocation, modulation and coding scheme, and digital zero-forcing beamforming. Leveraging codebook-based analog beamforming combined with zero-forcing digital processing, the proposed method achieves near-optimal performance while reducing computational complexity by two orders of magnitude and enabling scalability to large numbers of users. Experimental results demonstrate that the online algorithm closely approaches the theoretical performance upper bound and provide insights into the practical impact of key system parameters.