基于功率谱几何平均的频谱感知算法

Spectrum Sensing Algorithm Based on Geometric Average of the Power Spectral

利用功率谱极值和几何平均的频谱感知算法复杂度低,抗噪声功率不确定性和抗载波频偏的能力较强,但该算法的主用户信号功率估计不够准确。为了提高频谱感知性能,提出基于功率谱几何平均的频谱感知算法,算法用信号功率谱最大值平均减去信号功率谱几何平均来估计主用户信号功率,用信号功率谱的几何平均来估计信道噪声功率,利用两者的比值作为算法的判决统计量,并推导出检测门限。在AWGN信道、信噪比为-13 dB和虚警概率为0.07的条件下,进行5 000次仿真,改进算法的检测概率为0.975 6。理论分析和算法仿真均表明,改进算法提高了主用户信号感知准确度。

The spectrum sensing algorithm based on power spectrum extremum and geometric average has low complexity, strong ability of anti noise power uncertainty and anti carrier frequency offset, but the main user signal power estimation of the algorithm is not accurate. In order to improve the spectrum sensing performance, this paper proposes a spectrum sensing algorithm based on the geometric average of power spectrum. The algorithm estimates the signal power of the main user by subtracting the geometric average of the signal power spectrum from the maximum of the signal power spectrum, estimates the noise power of the channel by the geometric average of the signal power spectrum, uses the ratio of the two as the decision statistics of the algorithm, and deduces the detection threshold. Under the condition of AWGN channel, SNR of-13 dB and false alarm probability of 0.07, 5 000 simulations have been carried out. The inspection probability of the improved algorithm is 0.975 6. Theoretical analysis and algorithm simulation show that the improved algorithm improves the accuracy of the main user’s signal perception.