On SNR Wall Phenomenon under Cooperative Energy Detection in Spectrum Sensing

Cognitive radio is a promising scheme of 5G to efficiently utilize the spectrum and greatly gain in capacity, and energy detection algorithm is widely adopted in spectrum sensing for the low complexity of the receiver structure and no need of primary signal information. Single user energy detection in cognitive radio system has been proved vulnerable to the noise uncertainty, and suffered from the SNR wall. Compared with single user detection, cooperative energy detection algorithms can increase the detection probability significantly. This paper expounds the SNR wall in a new perspective and analyzes the SNR wall phenomenon under typical cooperative energy detection algorithms such as AND/OR hard decision and EGC soft decision. To keep a constant detection probability, three kinds of detection threshold, conservative/traditional/self-adaptive threshold, are studied and their influence are compared to find out the impact of SNR wall phenomenon. This paper also introduces a method to estimate the signal power, and analyzes the effect of possible estimation error of signal power on SNR wall, while MRC algorithm may decrease this limitation. Numerous simulations have been executed to verify the theory derivation, and results show that cooperative sensing and suitable threshold setting are hopeful to decrease the limitation of SNR wall. In cooperative sensing environment, MRC algorithm can low down SNR wall with the diversity increase of cooperative sensing users.