An estimation of primary user's SNR for spectrum sensing in cognitive radios

The accuracy for any spectrum sensing technique in cognitive radio (CR) is affected by the transmission channel parameters such as the signal to noise ratio (SNR). In this paper, a special approach is used to estimate the SNR in the received primary user (PU) signal. This approach is based on using filter bank transform which is called hybrid slantlet transform (HST). This transform decomposes the input PU signal into some sub-bands where some of these sub-bands give the approximate signal power with low noise affects and the others give the approximate noise power. The results for this approach are compared with iterated discrete wavelet transform (DWT) approach. Moreover, these results are shown the superior for HST approach in comparison with iterated DWT approach in reduction of the system complexity computational that it is reduced by 9.75%.