Wormhole Attack Detection in Wireless Sensor Network using Discrete Wavelet Transform

The Wormhole attack is a critical and challenging security threat for wireless sensor networks (WSNs). Although launching a wormhole attack is a relatively trivial task for the attacker, detecting it from an infected WSN is a tough job, because it can be launched from a compromised legitimate node. In recent years, signal processing methods, like the short term Fourier transform (STFT) and wavelet transform (WT) have been applied to find anomalies in computer network traffic [1]. According to the literature, the “number of neighbors” is an important detection feature to identify wormhole attacks from infected networks [2]. In our research, the discrete wavelet transform (DWT) is used to analyze a time-series of neighborhood counts to detect wormhole attacks. This proposed scheme is able to detect wormhole attacks in networks which contain both uniform and non-uniform sensor distributions. Also, the proposed scheme does not require special hardware, nor does it require significant network overhead. The simulation results show that this innovative scheme can detect wormhole attacks with high detection accuracy and negligible false positive rates for both uniform and non-uniform sensor distributions. Keywords—discrete wavelet transform; wormhole attack; wireless sensor networks; uniform sensor distribution; Nonuniform sensor distribution.