Indian Journal of Science and Technology

Abstract

Cognitive radio is one of the promising technique for dynamic spectrum sharing to solve the spectrum scarcity problem. Spectrum sensing is the key process in cognitive radio for the dynamic spectrum usage. But security in sensing is still in research issues .there are many security threads in cognitive radio but the primary user emulation attack is the predominant one which will not allow the dynamic spectrum sharing in effect. This primary emulation attack is the major drawbacks in energy detector based spectrumsensing. There are various methods to detect the emulation attack like RSSI based, location based etc. But the accuracy of detection is not good because of the uncertainty in the received power due to the random behavior of the channel condition. Most of those methods assume that primary users are stationary and channel variations are not significant. But in real time this assumption is not valid one. Here in this paper we propose higher order statics based primary user attack detection is proposed which can model the features of the primary user and primary user emulator attackers very well and we are able to detect the primary user emulation attack accurately. Here various fading channel scenarios like Rayleigh, rician and nagagammi is considered between the primary user and the attackers on that the performance of the detection of primary user emulation is analyzed.