How to utilize caching to improve spectral efficiency in device-to-device wireless networks

In this paper, we study the impact of caching on a recently-proposed spectrum sharing mechanism for device-to-device networks, ITLinQ, which schedules communication based on information-theoretic optimality principles. We provide a lower bound on the spectral reuse (the fraction of users simultaneously scheduled in the ITLinQ framework) as a function of the caching redundancy by proposing a greedy algorithm for source-user association and determining its achievable reuse. In order to demonstrate that higher spectral reuse also results in more efficient communication, we numerically evaluate the rates achieved by our scheme and show that it provides substantial throughput gain over the state-of-the-art.