Research On Network Topology Repair Mechanism Of Wireless Sensor And Actuator

Wireless Sensor and Actuator Network (WSANs) is formed through add actuators into wireless sensor networks (WSNs). Sensor node is stillness normally, however actuator node is movable and with more power, more computing and storage capability, more communication capability. Relative to WSNs could only monitor the physical environment, WSANs could interact with the environment besides monitor the physical environment, so as to affect the environment. The topology control of WSANs management and control the topology which composed of sensor nodes and actuator nodes, it is the basic of up layer protocols and application. Topology control includes two stages, topology establishment and topology maintenance. The function of topology maintenance is management and maintenance the change of topology, to keep superior performance, provides a redundancy network with high reliability of communication quality and data transmission, and routing path as shorter as possible.This paper studied the topology maintenance problem of actuator topology which consists of actuators in WSANs, and proposed a novel topology maintenance scheme focus on effectively repair when actuators failed. This scheme includes two basic algorithms, detection cut vertex base on loop algorithm (DCVL) and vertex-cut based topology repair algorithm (CVTR). DCVL and CVTR are running at actuators distributed, and DCVL runs after network has been built to check which actuators are cut vertices and which actuators are normal nodes. DCVL is based on local 2-hop neighbor nodes information, it estimates the node is cut vertex or normal node by means of analyze the relationship of these neighbor nodes. CVTR runs after actuator has been failed to repair the network and maintenance the topology. CVTR uses the results of DCVL, it chooses a agent node form its 1-hop neighbor nodes for every cut vertex. If the failure node is cut vertex, its agent node will relocate to its location, and replace it. The movement of agent node could deal as the agent node fails, its neighbor nodes will trigger the CVTR algorithm automatically. If the failure node is normal node, its neighbor nodes will move towards it until they meets the stop condition. The performance of the scheme is validated through simulation experiments, and the results show that DCVL could check every actuator is cut vertex or normal node accurately, except the individual loop. CVTR could timely and efficiently repair the network and maintenance the topology, and with excellent performance in total distance nodes moved and number of nodes moved.