Taylor Sailfish Optimizer-Based Deep Stacked Auto Encoder for Blackhole Attack Detection in Wireless Sensor Network

Abstract

Sensor nodes in Wireless sensor network (WSN) are distributed over a large area for sensing the pressure, temperature, humidity, and so on. They are at risk due to several attacks. In an attack like a black hole, the malicious node captures the whole data without any consideration of the active route, thus the source node are secured for communication. Hence, a new method name, Taylor SailFish Optimizer (TaylorSFO) is proposed to predict blackhole attacks in WSN. The training of the Deep stacked autoencoder is done through proposed Taylor-SFO, which is the integration of Taylor Series, and SailFish Optimizer (SFO). The newly developed Taylor-SFO is then applied for routing and blackhole attack detection at the WSN base station. Overall, two phases are included in the proposed model, which involves routing and blackhole attack detection at the base station. Initially, the WSN nodes are given to the routing module. Here, the routing is done based on the proposed TaylorSFO. Energy, distance as well as delay are the three fitness parameters considered for the routing. The proposed method shows the lowest delay of 21.23 ms, minimal FNR of 0.083, minimal FPR of 0.134, highest PDR of 94.87%, the highest throughput rate of 119.98 kbps, respectively.