Towards reliable intelligent transportation systems for e-government

The Intelligent Transportation Systems (ITS) is a state-of-the-art system, which maximizes mobility, safety, and usefulness through combining existing transport systems with information, communication, computer, and control technologies. Many governments have been building the ITS based on their own National ITS Architectures. The core functions of the ITS are collection, management, and provision of real time transport information, and it can be deployed based on the Common Object Request Broker Architecture (CORBA) of the Object Management Group (OMG) efficiently because it consists of many interconnected heterogeneous systems. Fault Tolerant CORBA (FT-CORBA) supports real time requirement of transport information stably through redundancy by replication of server objects. However, object replication, management, and related protocols of FT-CORBA require extra system CPU and memory resources, and can degrade the system performance both locally and as a whole. This paper proposes an improved architecture to reduce usage of CPU and memory and enhance reliability and ultimately predictability of FT-CORBA based ITS by generating and managing object replicas adaptively during system operation with an agent. The proposed architecture is expected to be applicable to other FT-CORBA based systems.