Implementation of SNR estimation based energy detection on USRP and GNU radio for cognitive radio networks

Development of smart spectrum sensing techniques is the most important task in the design of a cognitive radio system which uses the available spectrum efficiently. The adaptive SNR estimation based energy detection technique has the dual benefit of improving the efficiency of spectrum usage by capitalizing on the underutilization of the spectrum in an adaptive and iterative fashion, as well as reducing the hardware resources leading to easy implementation on a versatile and diverse group of cognitive radio infrastructures. The use of adaptive threshold for energy detection based on SNR estimation improves the spectrum sensing performance and efficiency of the cognitive radio by many folds, especially in low SNR as well as high noise variance situations. The proposed method is implemented on the USRP B200 and results show significant improvement in the detection rate of primary users as compared to conventional energy detection techniques.