A Novel Spectrum Sensing Mechanism Based on Distribution Discontinuity Estimation within Cognitive Radio

Abstract

A novel spectrum sensing approach based on distribution discontinuity estimation facilitated by change-point (CP) detection algorithm has been presented in this work. Specifically, a Markov Chain Monte Carlo (MCMC) based CP detection algorithm to estimate the variations in the distribution over the received signal is developed. Primary User (PU) activity, autonomous classification of modulation and detection of PU emulation attempts are detected using the CP- Maximum Likelihood Estimation (MLE) framework. Specifically, the CP based approach facilitates PU activity detection and initiates the MLE based mechanism to discern or reveal the underlying modulation scheme within the received signal. Thus, the proposed joint CP-MLE framework not only aims at detecting the discontinuity or the variations in the underlying distribution over the sensed signal, but also helps in attributing those variations to distinct modulation schemes, in an effort to identify PU emulation attempts. This CP- MLE framework has been extensively validated using the Universal Software Radio Peripheral (USRP) devices for various types of scenarios involving real-time modulation classification and identification of PU emulation type attacks.