A study on application of wavelets and filter banks for Cognitive Radio spectrum estimation

Spectrum sensing is important in cognitive radio (CR). Accuracy and speed of estimation are keys to select appropriate spectrum sensing technique. Conventional spectrum estimation techniques based on short time Fourier transform suffer from familiar problems such as low frequency resolution, large variance of estimated power and large side lobes. In this article, the possibility of employing wavelet packet decomposition for new spectrum sensing is investigated. In order to investigate the performance of the proposed approach, four types of sources (transmitters) are used, namely partial band, single tone, multi tone and swept tone. To simulate the temporal dynamics in CR, a special kind of source called swept tone source is devised. The novelty of this work relates to the development of wavelet packet decomposition using filter bank approach as a new spectrum sensing tool recommended for CR. The results of this study show that the wavelet based spectral estimation is a promising technique for CR applications.