Distributed and Privacy Preserving Routing of Connected Vehicles to Minimize Congestion

Abstract

With a large number of connected vehicles on the roads, there is an opportunity to leverage their connectivity to minimize congestion on roads by calculating fast routes for vehicles in a way that each vehicle contributes as little to the congestion as possible. The existing commercial and research based approaches of calculating routes for vehicles suffer from one or more of the following two limitations: 1) they are not privacy preserving in the sense that they receive destination addresses from users and may either store and use them for other commercial purposes or are at a risk of getting hacked and exposing these addresses to hackers; and 2) they require expensive infrastructure such as road side units (RSUs). To address these limitations, we propose a distributed and privacy preserving routing protocol, namely DPR, which the connected vehicles collaboratively and repeatedly execute to calculate fast routes to their destinations such that the overall congestion on the road network is significantly reduced and at the same time the privacy of the vehicles is preserved. The DPR protocol relies on direct vehicle to vehicle communication and does not need any new infrastructure such as RSUs. We have implemented and evaluated our DPR protocol through simulations on a real road network under several traffic conditions. Our results show that DPR reduces the average travel time of vehicles that travel a distance of 1000, 2500, and over 4000 meters by 15%, 32%, and 42%, respectively. This reduction in travel time is significant considering that this improvement results purely from software manipulations and without requiring any new infrastructure.