A low-complexity spectrum sensing technique for cognitive radios based on correlation of intra-segment decimated vectors

Efficient spectrum sensing plays a vital role in opportunistic and dynamic spectrum access for cognitive radios. Several spectrum sensing algorithms based on energy detection, matched filtering and autocorrelation based feature detection have therefore been proposed in the literature. All these algorithms tend to have higher complexity for achieving better sensing performance, which requires higher computing power and/or more sensing time. Moreover they are not flexible enough to dynamically trade complexity with sensing performance. In this paper, we present a novel low-complexity spectrum sensing technique based on correlation of intra-segment decimated vectors. The primary signal to be detected is sampled at a particular frequency, and digitized samples are segmented and decimated by suitable decimation index. The means of resulting vectors are correlated to detect the presence of the signal. MATLAB simulations of the proposed technique have been carried out to validate the proposed technique. Single-tone and Multi-tone signal at a 25MHz IF frequency and ATSC DTV (in UHF) signal and PAL-B Analog TV (in VHF) signal are down-converted to 25MHz IF frequency, and used to validate the proposed detection technique. Not only the proposed scheme is found to able to detect the presence of a signal effectively, but also the complexity of the proposed scheme is only nearly 1/10th of the complexity (for vector length n = 10) of the simplest of the existing correlation based technique.