An Evolutionary Game-Theoretic Approach for Cache-Enabled Cognitive D2D Networks

摘要

Device-to-device (D2D) communication is regarded as a promising technology in dealing with the dramatic congestion in network traffic. It alleviates the traffic backhaul by allowing data exchange between nearby mobile devices. However, balancing the network among cellular network users and D2D users is of great importance. Therefore, we model this problem as two evolutionary games, one among requesters and the other among helpers. Requesters have two strategies: get the content from the cellular network or from the D2D network. Helpers are users who have the content, they also have two strategies: share the content or not. To optimize the effective strategy of each user, we define a utility function for each strategy, then we adopt replicator dynamics to model the behaviors of users. Furthermore, we propose an algorithm that reflects our methodology and proves the efficiency in achieving the evolutionarily stable strategy for the two games as illustrated in simulation results.