Greedy cooperative cache placement for mobile edge networks with user preferences prediction and adaptive clustering

Abstract

Caching the most likely to be requested content at the base stations in a cooperative manner can facilitate direct content delivery without fetching content from the remote content server in a mobile edge network. This alleviates the user-perceived latency, reduces the burden on backhaul and minimizes the duplicated content transmissions. In addition to static content popularity, it is also important to consider the users’ dynamic behaviour for real-time applications. Unevenly distributed users, heterogeneity of the user preferences, and activity level in large-scale mobile edge networks are important issues that have been addressed in this work. The user preference has been predicted with the long short-term memory (LSTM) to capture the dynamic user behaviour and then users are clustered based on the predicted user preferences. The cache placement problem is formulated as Integer linear programming to maximize the saved delay with deadline and capacity constraints. Further, we devise a greedy algorithm by proving the submodular property of the objective function with matroid constraints, which can achieve (1−1e) optimality with linear computational complexity. Extensive simulation results demonstrate that the proposed cooperative caching mechanism significantly improves the performance in terms of acceleration ratio, hit ratio, and cache utilization compared with existing mechanisms.