Design And Implementation Of Complex Event System In M2M Service Platform

The development of Internet of Things(IoT)shows great booming of IoT device industry and great potential of smart devices.IoT service platform becomes one of the most important problems in building large-scale IoT systems.Researchers of Machine-to-machine(M2M)service platform develop several solutions addressing these problems.Today,most M2M service platform features both device ability modeling and message dispatching with distributed message queue systems.These systems work well in dispatching messages to IoT application service but lack support on real-time processing of messages,which limits the application scenario of the platform.Therefore,complex event processing(CEP),a technology for high-speed event stream processing,has been adopted to address the limitation.In this paper,we propose a CEP system that supports semantic web technology to support event processing in M2M service platform.Benefiting from semantic representation and semantic deduction technology,CEP engine can make deduction based on user-defined ontology data or knowledge base.The introduction of semantic web technology improves both extensitibility and data interoperatibility of the M2M platform in comparation to other CEP solutions.In this paper,we first introduce research background and related researches.Then,the functional and non-functional requirements of the event processing system have been analysed based on the case of traffic information system(TIS).Then,the architecture design of the system is discussed.There we discuss the design of event processing agent(EPA)and event processing engine(EPE)which is the core component of our complex event processing system.The following chapters describes detailed desigh and implementation of our prototype system.The next chapter discusses functional verification test on our prototype system based on traffic simulation systems.And at the end of this paper,a short review to the design and implementation of the system is made with discussion on the future research topics.