Co-operative Downloading in Vehicular Ad-Hoc Wireless Networks

Abstract

Increasing need for people to be "connected"; while at the same time remain as mobile as ever poses several interesting issues in wireless networks. It is conceivable in the near-future that wireless "hotspots" experience flash crowds-like traffic arrival pattern. A common phenomena in the Internet today characterized by sudden and unpredicted increase in popularity of on-line content. In this paper, we propose SPAWN, a cooperative strategy for content delivery and sharing in future vehicular networks. We study the issues involved in using such a strategy from the standpoint of vehicular ad-hoc networks. In particular, we show that not only content server but also wireless access network load reduction is critical. We propose a "communication efficient" swarming protocol which uses a gossip mechanism that leverages the inherent broadcast nature of the wireless medium, and a piece-selection strategy that takes proximity into account in decisions to exchange pieces. We show through simulation that gossip incorporates location-awareness into peer selection, while incurring low messaging overhead, and consequently enhancing the swarming protocol performance. We develop an analytical model to characterize the performance of SPAWN.