Joint Phase-Shift and Power Allocation Optimization in RIS-Enhanced Wireless Networks: An Intelligent Framework

The advancement of next-generation wireless networks demands innovative solutions to address escalating data requirements while ensuring equitable resource distribution and quality-of-service guarantees. Reconfigurable Intelligent Surfaces (RIS) have emerged as a transformative technology for reshaping propagation environments, yet existing approaches remain limited in holistically optimizing phase shifts and power allocation, particularly in multi-user scenarios requiring fairness-aware resource management. This paper bridges this critical gap by proposing an intelligent framework that harmonizes model based optimization with deep reinforcement learning. The primary objective is to maximize system throughput while ensuring fair resource distribution across users, addressing the inherent non-convexity of joint RIS configuration and power allocation. The methodology integrates alternating optimization decoupling phase shift adjustments, and power allocation into parallelizable subproblems with deep Q learning (DQL) to adaptively balance exploration and exploitation in dynamic environments. Key innovations include a fairness incorporated objective function, scalable batch-based resource allocation, and DQL driven policy refinement. Evaluations demonstrate that the framework achieves substantial improvements in spectral efficiency, energy efficiency, and latency compared to conventional methods, alongside robust scalability in dense deployments. Notably, it ensures equitable user service quality without compromising overall system performance. These findings underscore the viability of combining modelbased optimization with machine learning for RIS-aided networks, offering a pathway to intelligently controlled wireless environments. The studys implications extend to practical deployments in 6G networks, where adaptive resource management and environmental reconfigurability will be pivotal. By addressing both technical and fairness challenges, this work advances the realization of sustainable, high-performance wireless systems tailored for heterogeneous user demands.