Group Behavior-Based Collaborative Caching for Mobile Edge Computing

Mobile edge computing (MEC) enables computation and storage resources at mobile edge networks. Utilizing MEC storage resources to caching popular content has been recognized as an effective solution to tackle the backhaul bandwidth constraint of the network. However, most related works ignore or only focus on the spatiotemporal distribution of base stations and individual user preferences in mobile networks, which is insufficient in a dense network with users as the main body. In this paper, we proposed a Group Behavior and Popularity Prediction based Collaborative Caching (GPCC) strategy based on MEC architecture to reduce access delay. In GPCC, user access behavior is analyzed by collected a large amount of UDR (User Detail Record) and user groups are divided by community detection, the MEC logical domains are constructed based on the user groups to obtain the collaborative relationship between MECs so that the servers of the same collaborative domain serve a certain user group together. Then, the popularity of each content category in each user group is predicted by the autoregressive integrated moving average (ARIMA) model. Finally, we formulate the collaborative caching optimization model and resort to a heuristic algorithm for solving the cache content placement problem. Simulation results prove that the proposed caching strategy can achieve better performance in terms of cache hit rate and average access delay.