Collaborative Downloading in VANET using Network Coding

TL;DR

This paper explores how network coding can improve collaborative data downloading in vehicular networks by reducing download time and eliminating the need for uplink communication.

Contribution

It analytically demonstrates the benefits of network coding in vehicular collaborative downloading, especially in sparse roadside infrastructure scenarios.

Findings

Network coding slightly reduces download time.

Eliminates need for uplink communication from vehicles.

Provides probability distribution and expected value of download time.

Abstract

Data downloading on the fly is the base of commercial data services in vehicular networks, such as office-onwheels and entertainment-on-wheels. Due to the sparse spacial distribution of roadside Base Stations (BS) along the road, downloading through Roadside-to-Vehicle (R2V) connections is intermittent.When multiple vehicles with geographical proximity have common interest in certain objects to download, they can collaborate to reduce significantly their overall download time. In this paper, we investigate application of Network Coding (NC) in collaborative downloading (CD). We focus on the R2V part of CD, and analytically derive probability distribution and expected value of amount of time needed to deliver all information to the vehicles with and without NC. Our results show that using NC slightly improves the downloading time in addition to removing any need for having any sort of uplink communications from vehicles to the infrastructure.