Clustering-based device-to-device cache placement

Abstract

In this work we consider the problem of optimal cache placement in a D2D enabled cellular network. There are a number of helper devices in the area, which use their cached contents to help other users and offload traffic from the base station. The goal of cache placement is maximizing the offloaded traffic. We first formulate and optimally solve the cache placement problem as a mixed integer linear program. Then we propose a distributively implementable algorithm that clusters helpers. Helpers in each cluster collectively decide the contents to be cached, based on the content popularity. Numerical evaluations show that the proposed cache placement scheme always performs within 5% of the optimal result and it is robust to popularity profile and cache capacity.