Fault node detection using matrix calculus algorithm and reward andpunishment method in sensor networks

Abstract

Wireless Sensor Networks (WSNs) are composed of massive, small and low-cost sensor nodes deployed in a monitoring region, forming a multi-hop self-organized network system through wireless communication. The target is to cooperatively sense, collect and process the information about the objects in the coverage area, and then, sending them to the observer for processing and analyzing. It is a system with multi-functional and low energy consumption. Data delivery in sensor nodes is intrinsically faulty and unpredictable. Failures in remote sensor systems can happen for different reasons. This research work mainly focuses on the faulty node detection in WSN in an efficient manner. This research work mainly focuses on the multiple failure nodes and faulty nodes detection effectively and efficiently in WSN. Consequently, the lack of issues regarding the failure of sensor node, energy efficiency, reliability and faulty node affect the QoS by network partitioning and connectivity loss. Also, the data loss will be more and QoS will not be obtained. The faulty sensor nodes degrade the performance of the complete sensor network and hence, it is necessary to detect the faulty sensor nodes as well as it is essential to find the failure nodes in WSNs to attain better Quality of Service (QoS). In the first stage of the research work, failure node detection has been focused by novel matrix calculus (MCS) method which has been proposed to detect multiple failure nodes from the faulty path with minimum amount of detection time, in which row and column wise Round Trip Paths (RTPs) will be created. The MCS algorithm generates RTPs based on the sequence of row and column. By considering all the sensor nodes as good nodes, the Round Trip Delay (RTD) is calculated in all the generated RTPs. newline