Wide-band Spectrum Sensing for Cognitive Radio by Combining Antenna Signals

This paper presents methods for enhancing blind spectrum sensing of wide-band signals using multiple antennas on a single cognitive receiver. Regulatory bodies in various countries have found that most radio spectrum is inefficiently utilized. Cognitive radio is a paradigm capable of utilizing bands assigned to licensed or primary users when they are not being used. A critical component of a functioning cognitive radio system is the spectrum sensing of the primary users. We present a hidden Markov model based approach to detecting wide-band signals with multiple antennas, contributing (i) simple means of combining signals across multiple antennas, namely pre-detection combining and post-detection combining, and (ii) a modified hidden Markov model algorithm to detect the interferers from the antenna-combined signals. The designs are characterized using both the dependence of missed detection and false alarm error rates on the Interference-to-Noise Ratio (INR) and receiver operator characteristics.