基于循环平稳的图样特征识别频谱检测方法

Pattern recognition spectrum detection method based on cyclostationarity

频谱空穴检测是认知无线电研究的关键技术。针对认知无线电系统中甚低信噪比环境下频谱检测的困难,基于循环平稳理论,提出对接收信号的循环谱密度函数俯视图特殊图样的模式识别来进行频谱检测判决的方法。由此设计出谱峰搜索检测与图样特征识别相结合的频谱检测判决机制。实验仿真结果表明,所提出的检测方法性能良好,可以有效地对弱主用户信号进行频谱检测。

Spectrum hole detection is the key technology of cognitive radio. For the difficulties of spectrum detection in lower SNR, a new spectrum detection method based on cyclostationary theory is proposed. This method is to carry out pattern recogni- tion on the unique planforrn of the received signals＇ cyclostationary spectral density function. And the detection mechanism using peak searching combined with pattern recognition is designed. Simulation results show that, the detection performance of the proposed method is better than that of peak searching. And it can detect the weak primary users＇ signals effectively.