Multifractal slope feature based wireless devices identification

Cognitive Radio (CR) is a promising technology for alleviating spectrum shortage problem. However, with the realization of CR, new security issues are gradually emerging, like Primary User Emulation (PUE) attack. This paper incorporates transient-based identification into cognitive radio network (CRN) to defend against the PUE attacks. A method of extracting transient envelope features for wireless devices identification based on multifractal has been presented. Utilizing the joint fingerprint features of multifractal slope and polynomial fitting for wireless devices identification, the results show that the recognition performance is greatly improved.