Delta Ruled First Order Iterative Deep Neural Learning for Sybil and Wormhole Attacks Detection in Healthcare Wireless Sensor Network

Abstract

Detection and isolation of Sybil and wormhole attack nodes in healthcare WSN is a significant problem to be resolved. Few research works have been designed to identify Sybil and wormhole attack nodes in the network. However, the detection performance of Sybil and wormhole attack nodes was not effectual as the false alarm rate was higher. In order to overcome such limitations, Delta Ruled First Order Iterative Deep Learning based Intrusion Detection (DRFOIDL-ID) Technique is proposed. The DRFOIDL-ID Technique includes two main phase namely attack detection and isolation. The DRFOIDL-ID Technique constructs Delta Ruled First Order Iterative Deep Learning in attack detection phase with aim of detecting the occurrence of Sybil and wormhole attacks in healthcare WSN. After detecting the attack nodes, DRFOIDL-ID Technique carried outs isolation process with the objective of increasing the routing performance. During the isolation phase, DRFOIDL-ID Technique keep always the identified Sybil and wormhole attack nodes through transmitting the isolation messages to all sensor nodes in healthcare WSN. Hence, DRFOIDL-ID Technique improves the routing performance with lower packet loss rate. The DRFOIDL-ID Technique conducts the simulation process using factors such as attack detection rate, attack detection time, false alarm rate and packet loss rate with respect to a diverse number of sensor nodes and data packets. The simulation result proves that the DRFOIDL-ID Technique is able to improve the attack detection rate and also reduces the attack detection time as compared to state-of-the-art works.