An Extensible Multi-agent Based Traffic Simulation System

Agent technology is rapidly emerging as a powerful computing paradigm to cope with the complexity in dynamic distributed systems, such as microscopic traffic simulation systems. However, while a number of agent-based microscopic traffic simulation systems have been proposed and the multi-agent systems have been studied, to the best of our knowledge, the extensibility and interoperability have not been taken into account in this field. In this paper, we propose to integrate an extensible environmental information sensing(EIS) technology with multi-agent systems to enhance the ability of the traffic simulation systems to deal with increasing complex and large scale road network. In particular, we have developed an IEEE FIPA compliant traffic simulation system called JADE(Java based Agent Development Framework) based traffic simulation system(JTSS). The use of EIS decouples underlying infrastructure of road network and driver vehicle agents that makes it much easier to migrate between hosts of connected networks. The simulation of using multi-host deployment demonstrates that JTSS offers great flexibility in simulating large scale traffic systems.