RL-based Competition in Reconfigurable Wireless Environments

This project addresses the intricate dynamics surrounding the coordination and competition among mobile network operators within wireless transmission environments, which are increasingly influenced by the deployment of reconfigurable intelligent surfaces (RIS). These surfaces hold the promise of revolutionizing wireless communication by dynamically modifying the propagation environment, thereby enhancing signal quality and spectral efficiency. However, as RIS technology advances, it introduces new challenges, particularly in multi-operator scenarios where multiple entities share the same frequency spectrum. The deployment of RIS has the potential to significantly impact the wireless transmission channels of coexisting operators, thereby influencing the quality of service (QoS) experienced by their respective customer bases. One of the key complexities arises from the fact that the effects of reconfigurable intelligent surfaces vary across different frequency bands. This variability can lead to conflicts and coexistence issues among operators, as a configuration optimized for one operator`s network may inadvertently disrupt the operations of others. Such conflicts create a competitive landscape where operators must balance their individual objectives with the need to ensure fair and efficient spectrum utilization. In response to these challenges, our research focuses on developing methodologies for coordinated or cooperative RIS configuration through the establishment of dedicated RIS brokers. These brokers serve as intermediaries, facilitating communication and negotiation among operators to reconcile conflicting interests and optimize overall network performance. Central to our approach is the development of an auction-based RIS configuration framework, designed to balance the competing objectives of sustainability and QoS optimization in wireless networks. Through the application of distributed reinforcement learning (RL) techniques, this framework learns and adapts optimal bidding strategies over time, fostering implicit coordination among operators while maximizing social welfare for all stakeholders. By minimizing the energy consumption of mobile networks while simultaneously meeting user requirements, our goal is to unlock the full potential of RIS technology in enhancing wireless communication systems, ensuring equitable access and improved service quality for users across diverse network environments