Enhanced Wideband Spectrum Sensing Algorithm for Analysis of GSM Band

Dynamic variations in noise level and degradation of detection performance at low SNR are the main obstacles in sensing a wideband spectrum. The techniques based on energy detection estimate the energy of the sensed signal and compare it with a fixed predefined threshold. Such a general approach lacks adequate reliability and may cause inaccurate detection in sensing a wideband spectrum. This paper proposes an ‘Enhanced Wideband Spectrum Sensing Algorithm using Gradient and Double Threshold’ that aims at improving the detection performance in circumstances fraught with variations in noise level as well as low SNR. First, wideband sensing of GSM band is performed to determine spectrum occupancy and vacancy, by using the first algorithm. The same band is then scanned using the second algorithm. Experimental results obtained from the hardware platform are in favor of the second algorithm and prove that this proposed algorithm delivers improved detection performance in the presence of variations in noise level and low SNR.