Collaborative Spectrum Sensing Scheme: Quantized Weighting with Censoring

Abstract

Spectrum vacancies that stem from current non-usage of the spectrum band by legacy primary users can be detected using various spectrum sensing techniques. These techniques depend on the actual knowledge of the radio environment being inspected, i.e. signal characteristics, noise levels etc. The simplest and most common spectrum sensing technique is energy based detection that needs no a priori knowledge about the monitored spectrum, but may lead to imprecision when assessing the possible presence (or absence) of a primary user. Therefore, possible collaboration among the nodes performing the energy based spectrum sensing (in terms of sensing reports exchanges) improves the reliability and avoids the hidden terminal problem caused by the shadowing from large obstacles. This paper introduces a novel collaborative spectrum sensing scheme with light communication overhead called Quantized Weighting with Censoring (QWC). The scheme includes censoring of the unreliable sensing reports in some range of uncertainty and introduces weighting coefficients for different quantization levels. The performances of the QWC scheme are compared with the Majority Voting (MV) and Equal Gain Combining (EGC) schemes. The results show that the QWC scheme outperforms the well known EGC scheme.