Cache-Aided Broadcast Channels with User Cooperation

Consider K-user cache-aided broadcast channels with user cooperation, where a server connects with users, each equipped with cache memory, and the users can communicate with each other through a cooperation network. We introduce a new definition of transmission delay and propose a decentralized coded caching scheme based on a novel randomized delivery strategy. By exploiting parallel transmission between the server and the users, and optimally allocating communication loads, the transmission delay was greatly reduced, creating cooperation gain and parallel gain. It was shown that if the cache size of each users is larger than a small threshold that tends to zero as K goes to infinity, the scheme approaches the information theoretic lower bound with a constant multiplicative factor.