基于分形滤波的能量检测方法

Study of Energy Detection Based on Fractal Filtering

能量检测是认知无线电频谱感知的有效方法,但在低信噪比情况下检测概率明显下降。提出了基于分形滤波的频谱感知能量检测方法。基于分形滤波可以有效抑制噪声的特点,该方法先对接收信号进行分形滤波,以其能量值作为检测统计量,进行能量检测。仿真结果表明,通过基于分形滤波的能量检测和直接进行能量检测方法对比,滤波后信噪比在下降7 d B时仍然满足90%以上的检测概率。此外,该方法对不同调制信号均具有较好的适用性,在噪声不确定情况下仍然具有较高的检测概率。

The energy detection (ED)is an effective method for spectrum sensing in cognitiveradio,but the probability of detection decreased significantly in low SNR. An energy detection methodof fractal filtering based is proposed in this paper. Fractal filter can effectively suppress noise. So thesignal with noise is processed by fractal filter,and then the energy value of the filtered signal iscalculated and taken as test statistic for energy detection. Simulation results show that compared withenergy detection based on fractal filter and direct ED ,SNR for filtered signal at the time of dropping7 dB still meet more than 90 % probability of detection. Moreover,the proposed method has betterapplicability with different modulation signal and more reliable detection performance in the noise ofuncertainty than ED.