Multihop Caching-Aided Coded Multicasting for the Next Generation of Cellular Networks

TL;DR

This paper proposes a multihop coded multicasting scheme for cellular networks with femtocaching, reducing helper deployment costs and improving content delivery efficiency by leveraging user terminal storage and multihop communication.

Contribution

It introduces a multihop index coding technique that reduces femtocache requirements and enhances capacity, especially under Zipfian content popularity distributions.

Findings

Multihop coding reduces femtocache deployment costs.

Heuristic linear index codes achieve near-optimal capacity.

Multicast replaces unicast in high content reuse scenarios.

Abstract

Next generation of cellular networks deploying wireless distributed femtocaching infrastructure proposed by Golrezaei et. al. are studied. By taking advantage of multihop communications in each cell, the number of required femtocaching helpers is significantly reduced. This reduction of femtocaches is achieved by using the underutilized storage and communication capabilities in the User Terminals (UTs), which results in reducing the deployment costs of distributed femtocaches. A multihop index coding technique is proposed to code the cached contents in helpers to achieve order optimal capacity gains. This can serve as an efficient content delivery algorithm for the solution provided by Golrezaei et. al. As an example, we consider a wireless cellular system in which contents have a popularity distribution. It has been shown that if the contents follow a high content reuse popularity distribution, our approach can replace many unicast communication with multicast communication. We will prove that simple linear index codes found by heuristics based on graph coloring algorithms can achieve order optimal capacity under Zipfian content popularity distribution.