An Agent-based Autonomic Management Model And Application For Wireless Sensor Networks

Wireless sensor networks are large scale distributed systems of small, un-tethered, low-power nodes capable of sensing, processing and communication. With the ability of sensing and processing widely distributed in space and time, wireless sensor networks can give real-time monitoring, collecting information on the objects within the coverage area, and have been widely used in environmental monitoring, precision agriculture, space exploration, telemedicine, subway monitoring etc, which is a new hot subject. However, due to large-scale of network, limited resources of a single node and the critical distribution environment, how to manage the wireless sensor networks in an effective and low cost manner is a key and difficult problem on the current research.Firstly, related technologies about wireless sensor networks management model were analyzed and discussed deeply, and then the complexity of management was pointed out. Secondly, with autonomic computing, autonomic sensor element was built on the use of the flexibility and coordinated operations of Agent. And on this base, an Agent-based self-management model for wireless sensor networks was established to reduce the management complexity. At the same time, a self-recovery method was presented for the recovery process of the model. Thirdly, PEPA was used to characterize the self-recovery process in order to facilitate reasoning and analysis for the self-management model, making sure it can do quantization analysis. Finally, an automated tool was used to analyze the PEPA model established.The experimental results show that the proposed Agent-based self-recovery model for wireless sensor networks is effective, and increasing the detection rate of detection Agent as well as the correct decision rate of decision Agent is helpful to reduce node compromise. The results of this paper provide a reference for further study of wireless sensor network management.