Intrusion Detection System Attack Classification with Optimization Model for WSN Security

Abstract

Wireless Sensor Network (WSN) subjected various challenges during data transmission between nodes deployed in a network. To withstand those security challenges Intrusion Detection System (IDS) is designed. IDS is involved in attack detection and classification but is subjected to a lack of effective classification techniques for attack prevention. To overcome those challenges associated with security this research presented an effective clustering technique known as Centred-Order Node Clustering (CONC). Also, Cluster Head (CH) is elected based on the Improved Flower Pollination Algorithm (IFPA) with multi-objective characteristics. By this proposed method lifetime of the network is improved. Additionally, a supervised classification technique called AdaBoost Regression Classifier (ABRC) is developed with the Intrusion Detection System (IDS). The developed ABRC is constructed for malicious node detection with the prediction of several attacks using IDS. Through improved security mechanisms sensor nodes are involved in effective data transmission between sensor nodes. The simulation analysis stated that the proposed mechanism provides better results rather than the existing technique.