Efficient Filtering and Location Detection against Insider Attacks in WSN
Abstract
Improving the security in wireless sensor networks(WSNs) is a necessary and challenging task in recent days. There are several techniques proposed for attaining security in data aggregation. Among those system monitoring modules(SMM) and intrusion detection modules(IDM) are the two techniques which perform the detection of external and internal threats of WSNs. This work illustrates how local detection approaches work together with the SMM to differentiate between internally malicious events and emergency events to overcome the limitations of local detection mechanisms. Preserving privacy is also an important concern when data aggregation takes place in military applications. An Extended Kalman Filter(EKF) based mechanism is used to detect false injected data. Specifically by monitoring behaviours of its neighbours and using EKF to predict future states. Each node aims at setting up a normal range of neighbours’ future transmitted aggregated values. An algorithm combining cumulative summation and generalized likelihood ratio(GLR) algorithms are used to create effective intrusion detection.
Keywords
System Monitoring Modules (SMM), Extended Kalman Filter (EKF), Cumulative Summation(CUSUM), Generalized Likelihood Ratio(GLR).