Summary

RTMonitor is a real-time data management system for traffic navigation applications. In our system, mobile vehicles initiate time-constrained navigation requests and RTMonitor calculates and communicates the best paths for the clients based on the road network and real-time traffic data. The correctness of the suggested routes highly depends on how well the system can maintain temporal consistency of the traffic data. To minimize the overheads of maintaining the real-time data, RTMonitor adopts a cooperative and distributed approach using mobile agents which can greatly reduce the amount of communications and improves the scalability of the system. To minimize the space and message overheads, we have designed a two-level traffic graph scheme to organize the real-time traffic data to support navigation requests. In the framework, the agents use an Adaptive PUSH OR PULL (APoP) scheme to maintain the temporal consistency of the traffic data. Our experiments using synthetic traffic data show that RTMonitor can provide efficient support to serve navigation requests in a timely fashion. Although several agents may be needed to serve a request, the size of each agent is very small (only a few kilobytes) and the resulting communication and processing overheads for data monitoring can be maintained within a reasonable level.