Particle swarm optimization based artificial neural network (PSO-ANN) model for effective k-barrier count intrusion detection system in WSN

Abstract

Wireless Sensor Networks (WSN) offers an extensive array of possibilities and finds application to nearly every wake of human lifetime. Two of the methods are intrusion detection and monitoring in dangerous environments. Generally, WSN are extremely vulnerable to various kinds of network attacks, lifespan of whole network is reduced by interfering with data transmission and communication operations. A novel approach, provides accurate measurement of several obstacles to achieve rapid identification and avoidance of disruptions. In order to more accurately estimate k-barrier number for efficient intrusion detection and prevention, a Particle Swarm Optimization based Artificial Neural Network (PSO-ANN) is developed. The region's size, sensor's detecting field, area where sensor transmits its signals, several sensors constitute the four important features utilized for learning and evaluating PSO-ANN model. The extraction of the different attributes is achieved with the help of simulation through Monte Carlo. It has been inferred from the research that the approach is accurate in predicting number of invasions, correlation coefficient, root mean square error, accuracy, precision are achieved to be excellent in contrast to the remaining approaches.

Keywords

• Intrusion detection
• Classification
• Wireless sensor network
• Network attacks
• Feature importance and feature sensitivity