Spectrum sensing based on fractional lower order moments for cognitive radios in α-stable distributed noise

• We propose a novel FLOM based detector for the detection of a primary user in the SαS noises.
• We derive the detection performance of the FLOM detector in the low GSNR regime for non-fading sensing channels.
• We also analyze the detection performance with respect to the degree of non-Gaussianity and the number of samples.
• We analyze the global detection and false alarm probabilities of the proposed FLOM-based cooperative sensing for Rayleigh fading channels.
• We show that the FLOM detector can significantly enhance the detection performance over the Cauchy detection in the SαS noise.

The traditional spectrum sensing methods based on second order statistics are in general not applicable to detecting a primary user with unknown parameters in non-Gaussian noises. This paper presents a novel spectrum sensing scheme based on fractional lower order moment (FLOM) for the detection of a primary user in non-Gaussian noise that are modeled by the α-stable distribution. The new detector does not require any a priori knowledge about the primary user (PU) signal and channels. The statistics of the proposed FLOM detector are defined in a multi-user cooperative framework and its detection and false alarm probabilities as well as deflection coefficient are analyzed for both non-fading and Rayleigh fading communication channels between the primary and secondary users. The detection performance of the proposed method versus the generalized signal-to-noise ratio, the characteristic exponent α and the number of cooperative users is also studied along with comparison to the Cauchy detector through computer simulations. Analytical and simulation results show that the proposed FLOM detector has a much better performance than the Cauchy detector in the α-stable distributed noise environment. It is also shown that multi-user cooperative sensing leads to a significantly higher probability of detection than the single user version.