Research Area:  Wireless Sensor Networks

Abstract:

Malicious attacks like denial-of-service massively affect the network activities of wireless sensor network. These attacks exploit network layer vulnerabilities and affect all the layers of the network. Anomaly based intrusion detection system (AIDS) are designed for monitoring such unpredictable attacks but it generates high false positive. In the proposed study we design robust and efficient AIDS which use fuzzy and neural network (NN) based tools. The proposed system can be implemented in each node as it is lightweight and does not consume much overhead. Also it can independently monitor the local nodes behaviour and identify whether a node is trust, distrust or enemy. The use of a trained NN filters the false alarms generated due to fuzzy logic applied in the first step thus enhancing the system accuracy. We evaluate the systems performance in NS2.35 and result shows a 100% true positive with 0% false positive.

Keywords:  

Author(s) Name:  Somnath Sinha,Aditi Paul

Publisher name:  Springer

DOI:  10.1007/s11277-020-07395-y

Volume Information:  volume 114, pages 835–851 (2020)

Paper Link:   https://link.springer.com/article/10.1007/s11277-020-07395-y