Capacity scaling of wireless device-to-device caching networks under the physical model

Abstract

We study the capacity scaling law of a device-to-device (D2D) caching network where n nodes are placed on a regular grid of area n. Each node caches some (coded) bits from a content library and requests a file from the library independently according to the Zipf popularity distribution. We propose a cache-induced hierarchical cooperation scheme which achieves the optimal capacity scaling law under a commonly used 'physical model'. When the path loss exponent α < 3, the capacity scaling law can be significantly better than the throughput scaling laws achieved by the existing state-of-the-art schemes. To the best of our knowledge, this is the first work that completely characterizes the capacity scaling law for wireless caching networks under the physical model.