Wideband Spectrum Sensing using Modulated Wideband Converter by Revised Orthogonal Matching Pursuit

Abstract

Spectrum sensing is a major function of Cognitive radio. Wideband spectrum sensing has been a focal area of research and development for the past few decades. Conventional wideband spectrum sensing techniques used Nyquist sampling rates. The major drawback associated with these techniques is its high computational complexity attached to the required high sampling rates.With the advent of sub-Nyquist sampling techniques, a breakthrough is achieved in wideband spectrum sensing. Modulated wideband converter is one such technique. Orthogonal matching pursuit is a widely used compressive sensing algorithm for recovering sparse signals in noisy linear regression models. Recovering anN-sparse signal from $N$ iterations is a sheer restriction to orthogonal matching pursuit algorithm. Researches on orthogonal matching pursuit suggests that more iterations will improve its performance. Extended orthogonal matching pursuit (OMPα) runs orthogonal matching pursuit algorithm for (N+[αN]) iterations instead of $N$ iterations, where α. ∊[0, 1] and can recover the same signal with reduced number of measurements.However, both orthogonal matching pursuit and OMPα requires a priori knowledge about the sparsityN. Sparsity unaware orthogonal matching pursuit (OMP∞)is introduced to overcome this limitation. OMPα run iterations until the signal residue vanishes and achieves l 0 -norm recovery without the sparsity information.In this paper, we analyzed the performance of modulated wideband converter based wideband spectrum sensing technique using various orthogonal matching pursuit algorithms.