Cognitive Agent Based Critical Information Gathering and Dissemination in Vehicular Ad hoc Networks

Next generation vehicles will have capability of sensing, computing, and communicating capabilities. Different components in a vehicle have to constantly exchange available information with other vehicles on the road and cooperate for the purpose of ensuring safety and comfort using a Vehicular Ad hoc Network (VANET). Critical information like navigation, cooperative collision avoidance, lane-changing, speed limit, accident, obstacle or road condition warnings, etc. play a significant role for safety-related applications in VANET. Such kind of critical information gathering and dissemination is challenging, because of their delay-sensitive nature. This paper proposes an agent based model that consists of heavy-weight static cognitive (based on Belief Desire Intention : BDI) and light-weight mobile agents. Proposed model executes push (gather/store and disseminate) and pull (gather/store) operations on information gathered based on information relevance, criticalness and importance. The simulation results show that BDI based information gathering and dissemination scheme performs better than the reliable broadcast scheme in terms of bandwidth utilization, packet delivery ratio, push latency (information saturation time) and push/pull decision latency.