Cooperative Spectrum Sensing Method Using Sub-band Decomposition with DCT for Cognitive Radio System

Abstract

In this paper a new cooperative spectrum sensing method for cognitive radio systems is presented. The proposed method is based on a new soft decision rule at Fusion Centre (FC) for deciding the primary user (PU) signal presence using the received energy from each Cognitive User (CU) after performing a set of transformations. The sensing process can be summarized as follows: each CU collects the available spectrum and one-level Discrete Wavelet Transform (DWT) decomposition for the input spectrum is taken. After the decomposition, only the first lower frequencies (LL) and the last high frequencies (HH) bands of two-dimensional DWT are truncated. Then Discrete Cosine Transform (DCT) for the truncated sub-bands is computed. The energy of each sub-band with Average Ratio (AR) between LL and HH energies is calculated. Finally, only the energy in LL band and AR are sent to FC via reporting channel. At FC the received energies and ARs from all CUs are averaged. The proposed fusion rule is designed such that if the average energy falls below the half of predefined threshold, PU is declared to be absent. From other hand if the average energy equals or greater than threshold value and average AR is greater than 1, PU is declared to be present. Otherwise, it is declared to be absent. The simulation results in both AWGN and Rayleigh fading channels showed that the proposed method has improved detection probability especially at low SNR, reduced sensing time and reduced energy consumption as compared to censoring hard fusion rules available in the literature.