Collaborative Dynamic Watchdog Mechanism for Detection and Isolation of Malicious Nodes through Clustering in Wireless Sensor Networks

Abstract

One of the most important characteristics of Wireless sensor networks (WSN) is the cooperation among the nodes that are present over the network. Since WSN consists of both static and dynamic nodes, the cooperation has to be managed in an effective manner. But this cooperation activity consumes more cost in terms of resources. The cooperation among nodes is a cost consuming activity as some of the nodes in the working environment can be malicious and selfish which will affect network's overall performance. The malicious nodes are nodes which affect the other nodes in the network whereas selfish nodes are nodes that will not work in a cooperative manner in the network. There is a well-known mechanism called watchdog which is used for identifying these malicious and selfish nodes. In the existing system, the watchdogs have been implemented in a local manner. The local watchdog will take charge of its own node. Implementation of local watchdogs leads to lack of precision when detecting selfish nodes. To overcome this, a Collaborative watchdog mechanism through Clustering (CCOCOWOA) has been proposed for identifying selfish nodes and for recovering the network. This mechanism will work based on the information that will be gathered when a contact occurs with selfish nodes and it will be quickly propagated. It is also efficient by energy and communication cost compared to COCOWA based approach for detecting selfish nodes.