Detection Algorithm Of Modulated Signal Based On Spectrum Variance

Signal detection can be applied to the spectrum sensing, which is a key technology of cognitive radio (CR). The conventional signal detection algorithms calculate the sum of energy in interested frequency band to recognize whether a modulated signal is present or not. However, the noise power level is volatile in different cases, which deteriorates the detection performance. In this paper, we propose a new modulated signal detection algorithm by dividing the analyzed spectrum band into several blocks and then calculating the sum of their energy variances. Numerical results indicate that about 5dB lower signal-to-noise ratio (SNR) is needed to detect modulated signal in provided algorithm compared with traditional method.