Abstract: With the increase of noise power,the characteristics of weak target signal become blurred and difficult to distinguish by noise pollution,which leads to the failure of weak signal detection method.To solve this problem,a quantum signal processing method named Local Semi-Classical Signal Analysis(LSCSA)that could protect the target signal features was proposed.The quantization principle of the LSCSA and the properties of protecting target signal features in quantum domain were introduced.The implementation steps of the LSCSA and the calculation methods of important parameters were given.The LSCSA was combined with Singular Value Decomposition(SVD)and wavelet threshold de-noising method for simulation analysis and experimental verification.The verification results showed that the ability of the LSCSA to protect the features of target signals could help the denoising methods to detect weak signals with very low Signal-to-Noise Ratio(SNR).Combined with other methods,the SNR was greatly improved,and the weak target signals(SNR=-30dB)was accurately extracted.The performance of the LSCSA was superior.

Key words: weak signal detection, quantum signal processing, protection of target signal features, local semi-classical signal analysis, singular value decomposition, wavelet threshold denoising

摘要： 随着噪声功率的增强,微弱目标信号的特征受噪声污染变得模糊且难以区分,导致微弱信号检测算法失效,提出一种可以保护目标信号特征的量子信号处理方法——局域半经典信号分析算法。详细介绍了算法实现量子化的原理和在量子域中保护目标信号特征的性质;给出算法步骤以及重要参数的计算方式;将所提算法与奇异值分解、小波阈值降噪算法结合进行了仿真分析和实验验证。结果表明,所提算法保护目标信号特征的能力可以帮助降噪算法检测极低信噪比的微弱信号,与其他方法结合可极大改善信噪比,准确提取信噪比为-30 dB的微弱目标信号,算法性能优越。

关键词: 微弱信号检测, 量子信号处理, 保护特征, 局域半经典信号分析, 奇异值分解, 小波阈值降噪