Secure Intrusion Detection System for Hierarchically Distributed Wireless Sensor Networks

Wireless sensor networks (WSNs) are a collection of sensors and developing technology applied in military and civilian areas of study. Sensor networks randomly and hierarchically configured in remote and unattended environments. Wireless sensor networks expose to security challenges since sensor nodes are deployed in an open environment and wireless communication between the nodes. WSNs are vulnerable to different security threats such as a black hole, Sybil, sinkhole, wormhole, forwarding, gray hole attacks…, etc. Due to the increase, these attacks on critical infrastructures managed by networked systems, designing robust intrusion detection systems is essential for protecting sensitive information. In this work, A secure intrusion detection system is designed based on decision tree a machine learning classification model. This model is used to build predictive model by training and testing. Decision tree IDS confirms the detection accuracy of 99.8% using MATLAB with NSL-KDD dataset. We examined this work by using NSL-KDD dataset as benchmark for performance comparison detection metrics with different class of attacks.