Abstract
In many real time applications, Wireless Sensor Networks (WSN)is one of the most promising technologies that have efficient data transmission. With the various and increasingly malicious attacks on WSNs, traditional security tools such as firewalls and anti-virus programs are not sufficient to provide free, integrated, reliable and secure networks. One of the most tested and reliable Intrusion Detection Systems (IDSs) technologies bring upgraded monitoring which enhances integrity and efficiency of the system. In this work, a Fuzzy Neural Network with Expectation Maximization (FNN-EM) classifier is proposed for perceiving malicious behavior and presenting authentication as well as data integrity. Cluster based routing is ascertained based on trust vector and residual energy of neighbor nodes in random topology to attain the above. In proposed system, the packet transmission between the source nodes to destination node is encrypted using the asymmetric key authentication scheme. Then, trust recommendation value is used to identify the malicious nodes. The normal and malicious nodes are classified using FNN-EM. For FNN, the EM algorithm is a suitable method to attain maximum likelihood evaluation for each class conditional density. The performance result demonstrates that the proposed FNN-EM perceived the network intrusion in efficient manner and its performance is superior compared to existing methods.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call