Incentive-Based Caching and Communication in a Clustered D2D Network

Abstract

Caching at the network edge can reap significant advantages to improve service quality by reducing the transmission cost and network congestion. Edge caching with device-to-device (D2D) communication helps offload cellular traffic during a surge in network traffic. This article considers the classic clustering problem for D2D users, followed by an incentive mechanism designed for successful D2D communication. The proposed clustering scheme merges D2D users with similar interests into one cluster. Specifically, a hierarchical agglomerative clustering algorithm is applied, where clusters are formed based on the similarity in users’ social interests. The cache hit probability is then optimized for each cluster, and the performance is examined for a varying number of clusters. We then propose a monetary incentive-based mechanism based on a points system to increase user participation for successful D2D communication by keeping track of a user’s content-providing history. Using this approach, devices with a good participation rate in D2D communication are identified and rewarded for their performance. Through simulations, we show that the hit rate improves by more than 40% after ensuring incentive-based methods in the D2D network.