Optimizing MDS Coded Caching in Wireless Networks with Device-to-Device Communication

TL;DR

This paper optimizes MDS coded caching strategies in mobile wireless networks with device-to-device communication, reducing network load by leveraging mobility and coded packet spreading.

Contribution

It introduces an analytical model for device mobility, optimizes MDS codes for minimal network load, and demonstrates the benefits of maximal spreading of coded packets.

Findings

Optimized MDS codes significantly reduce network load.

Maximal spreading of coded packets is shown to be optimal.

Mobility-aware caching improves content delivery efficiency.

Abstract

We consider the caching of content in the mobile devices in a dense wireless network using maximum distance separable (MDS) codes. We focus on an area, served by a base station (BS), where mobile devices move around according to a random mobility model. Users requesting a particular file download coded packets from caching devices within a communication range, using device-to-device communication. If additional packets are required to decode the file, these are downloaded from the BS. We analyze the device mobility and derive a good approximation of the distribution of caching devices within the communication range of mobile devices at any given time. We then optimize the MDS codes to minimize the network load under a cache size constraint and show that using optimized MDS codes results in significantly lower network load compared to when caching the most popular files. We further show numerically that caching coded packets of each file on all mobile devices, i.e., maximal spreading, is optimal.