Coded Caching: User to Access Points Distribution in Wireless Local Area Networks

Modern wireless networks face significant challenges in high-density deployments, where the existing approaches fail to fully exploit the cooperative potential of coded caching, intelligent user clustering, and graph-theoretic optimization. This paper proposes a novel coded caching framework that optimizes content delivery in a wireless local area network where users connect to cache-less access points. The system proposes a novel user-to-access-point assignment approach and leverages coded multicast opportunities across carefully paired users and access points to maximize the caching gain and improve load balancing. Our key contribution is the development of a clustering algorithm that efficiently distributes users across access points. This algorithm was inspired by the enhanced K++ means initialization technique that determines a centroid for each cluster, leveraging the benefits of coded caching to maximize caching gains. To validate its performance and complexity, we compare our clustering algorithm against a random User-to-AP assignment algorithm. In this heuristic baseline, users connect to access points randomly, ignoring load balance. Our simulations reveal the dual benefits of the clustering approach, highlighting enhanced caching performance through optimized content placement and improved network efficiency via balanced user distribution.