Abstract
Wireless Sensor Networks (WSN) possess vulnerable characteristics such as outdoor unattended operations for transmission and self-organized behaviour without a fixed infrastructure leads them to suffer from various challenges including lower processing power, low battery life, small memory and wireless communication channel. Security of a communication channel is the major concern to handle such kind of networks. Due to well-known and accepted limitations, overall security of WSN becomes the main concern to deal with. Intrusion Detection Systems (IDSs) can play an important role in detection and prevention of security attacks. In this paper, we propose a hybrid framework for intrusion detection system for wireless sensor networks that would take advantage of cluster-based architecture to optimize energy consumption. Proposed hybrid model uses anomaly detection based on support vector machine (SVM) algorithm and a set of signature rules to detect malicious behaviours and provide and ideal hybrid framework for IDS in WSN. Simulation result shows that the proposed hybrid model can detect abnormal events with higher and efficient detection rate with lower false alarm. Keywords: Wireless Sensor Network, Hybrid Intrusion Detection System, Support vector machine, Signature attacks, false alarm, detection rate, Frameworks
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call