Reducing the Sampling Complexity of Energy Detection in Cognitive Radio Networks under Low SNR by Using the Optimal Stochastic Resonance Technique

We propose a novel noncooperative technique in cognitive radio (CR) networks, which is based on the optimal stochastic resonance (SR) technique. By introducing the dynamic system approach of SR into the noncooperative spectrum sensing process, the defect of high sampling complexity of traditional energy detector can be reduced efficiently and thus can guarantee the applicability of the optimal SR-based energy detection method. The optimization of the signal-to-noise ratio (SNR) improvement of the system ensures the lowest sampling complexity needed to reach certain performance requirement. Computer simulations show that it can reduce the sampling complexity compared with traditional energy detector used in IEEE 802.22 draft especially under low SNR environments. It can certainly be extended to other wide application areas.