Weighted energy detection method for spectrum sensing in cognitive radio

In order to improve the average performance of detecting spectrum holes for second user (SU), we propose the weighted energy detection method. In this paper, we firstly classify the detection situations into two modes. One corresponds to the situation that the primary user (PU) signal takes up the whole detection window if PU signal exists. The other corresponds to the situation that the PU signal only appears in part of the whole detection window. Then we introduce the weighted factor for each received signal sample in detection window. The weighted factors follow the tendency that the samples closer to the end of detection window are more important for judgment. The simulation results show that the weighted method works better than the conventional method and hybrid method in the second mode. The weighted method has the lowest detectable signal-to-noise ratio (SNR) in second mode. And the performance improvement in second mode leads to only a little performance loss in the first mode. The weighted method makes a performance compromise between the first mode and the second mode.