Improved Consecutive Mean Excision Algorithm Based Spectrum Sensing for Dynamic Spectrum Access

This paper deals with spectrum sensing for dynamic spectrum access (DSA) via channelized radiometers. Consecutive mean excision (CME) algorithms have been employed in the literature to identify the presence or absence of licensed user signals. To obtain better control of false-alarm rate and detection rate, an improved CME algorithm is proposed in this paper, in which two thresholds are incorporated to perform the channel cluster based licensed user signal detection. It is verified through numerical experiments that with appropriate compromise on setting the two thresholds, the improved CME algorithm can significantly boost the detection rate while ideally maintain the false-alarm rate as sufficiently low as desired.