Infocast: A New Paradigm for Collaborative Content Distribution from Roadside Units to Vehicular Networks Using Rateless Codes

Abstract

In this paper, we address the problem of distributing a large amount of bulk data to a sparse vehicular network from roadside infostations, using efficient vehicle-to-vehicle collaboration. Due to the highly dynamic nature of the underlying vehicular network topology, we depart from architectures requiring centralized coordination, reliable MAC scheduling, or global network state knowledge, and instead adopt a distributed paradigm with simple protocols. In other words, we investigate the problem of reliable dissemination from multiple sources when each node in the network shares a limited amount of its resources for cooperating with others. By using \emph{rateless} coding at the Road Side Unit (RSU) and using vehicles as data carriers, we describe an efficient way to achieve reliable dissemination to all nodes (even disconnected clusters in the network). In the nutshell, we explore vehicles as mobile storage devices. We then develop a method to keep the density of the rateless codes packets as a function of distance from the RSU at the desired level set for the target decoding distance. We investigate various tradeoffs involving buffer size, maximum capacity, and the mobility parameter of the vehicles.