Spectrum Sensing under Spectrum Misuse Behaviors: A Multi-Hypothesis Test Perspective

Spectrum misuse behaviors, brought either by illegitimate access or by rogue power emission, endanger the legitimate communication and deteriorate the spectrum usage environment. In this paper, our aim is to detect whether the spectrum band is occupied, and if it is occupied, recognize whether the misuse behavior exists. One vital challenge is that the legitimate spectrum exploitation and misuse behaviors probabilistically coexist and the illegitimate user may act in an intermittent and fast-changing manner, which brings about much uncertainty for spectrum sensing. To tackle it, we first formulate the spectrum sensing problems under illegitimate access and rogue power emission as a uniform ternary hypothesis test. Then, we develop a novel test criterion, named the generalized multi-hypothesis Neyman-Pearson (GMNP) criterion. Following the criterion, we derive two test rules based on the generalized likelihood ratio test and the Rao test, respectively, whose asymptotic performances are analyzed and an upper bound is also given. Furthermore, a cooperative spectrum sensing scheme is designed based on the global GMNP criterion to further improve the detection performances. In addition, extensive simulations are provided to verify the proposed schemes' performance under various parameter configurations.