Efficient Data Propagation in Traffic-Monitoring Vehicular Networks

Road congestion and traffic-related pollution have a large negative social and economic impact on several economies worldwide. We believe that investment in the monitoring, distribution, and processing of traffic information should enable better strategic planning and encourage better use of public transport, both of which would help cut pollution and congestion. This paper investigates the problem of efficiently collecting and disseminating traffic information in an urban setting. We formulate the traffic data acquisition problem and explore solutions in the mobile sensor network domain while considering realistic application requirements. By leveraging existing infrastructure such as traveling vehicles in the city, we propose traffic data dissemination schemes that operate on both the routing and the application layer; our schemes are frugal in the use of the wireless medium, rendering our system interoperable with the proliferation of competing applications. We introduce the following two routing algorithms for vehicular networks that aim at minimizing communication and, at the same time, adhering to a delay threshold set by the application: 1) delay-bounded greedy forwarding and 2) delay-bounded minimum-cost forwarding. We propose a framework that jointly optimizes the two key processes associated with monitoring traffic, i.e., data acquisition and data delivery, and provide a thorough experimental evaluation based on realistic vehicular traces on a real city map.