Research On Node Adaptive Fault Tolerant Mechanism Based On Data Dependence

Wireless Sensor Network is a typical communication Network without infrastructure. It monitors environment real-time and collects information of the monitoring object. These data information will be processed and sent to the manage users. Wireless sensor network with low economic cost and flexible network topology, is widely used for environmental detection, military defense, intelligent household, biomedical and so on. In the application, some nodes will be placed in the place that is harsh environment, remote, and hard to reach. Typical fault-tolerant algorithm based on spatial correlation detection data including Bayesian-fault-tolerant algorithm, etc. This algorithm exist adjacent nodes communication is frequently, the problem such as energy consumption is too large. And event detection in the border, and there is a node detection information conflicting issues.In this paper, propose the traditional single threshold decision optimization mathematical model for the node decision threshold area. And using the decision threshold area, combined with the testing data of time correlation, this paper proposes a distributed fault-tolerant algorithm based on correlation data of time. This paper puts forward the fault-tolerant algorithm combined with A-SMAC protocol mechanism. Using adaptive fault-tolerant mechanism of sensor network, when the mutations in the test environment test results can be quickly sent out; When tested environment tend to be stable, according to the data of time correlation can undertake fault tolerance for the verdict, reduce unreliable data sent, thus improve the reliability of the node decision, improve the quality of the overall detection of the network.Finally using NS2simulation software simulate and analysis of adaptive fault-tolerant mechanism. Experimental results show that in the event of slow change monitoring environment, this time based on the data correlation of adaptive fault-tolerant decision mechanism, can effectively reduce the individual sensor node errors due to noise, and thus improve the accuracy of the data decision. Due to node set the decision threshold area and reduce the unnecessary data sending, compared with A-SMAC protocol, further reducing the node energy consumption, so as to improve the performance of the whole network.