Improved wavelet transform based edge detection for wide band spectrum sensing in Cognitive Radio
Abhishek Kumar, Seemanti Saha, Rajarshi Bhattacharya · 2016
In Cognitive Radio Network an efficient fast and accurate wideband spectrum sensing technique is highly required to identify the spectrum holes in wireless environment to achieve efficient spectrum utilization. The wavelet transform being a multi-resolution analysis tool has been proposed for the edge detection of sub-bands in wideband spectrum sensing [1]. In this paper we proposed an improved wavelet transform (WT) based algorithm where we perform non-linear logarithmic scaling of the WT coefficient after doing normalization of the PSD of the wideband spectrum. Logarithmic scaling along with the normalization is done to emphasis the small single scale modulus maxima at the edges to achieve more effective WT based edge detection in order to get better accuracy. Comparative studies show that the proposed algorithm performs better than the conventional WT based edge detection algorithm as well as edge detection multi-scale algorithms.