非高斯水下噪声中微弱线谱信号的符号函数型随机共振检测

Weak Line Spectrum Detection Based on Sign Function Type Stochastic Resonance in Non-Gaussian Marine Noise

采用符号函数作为随机共振模型,对非高斯水下噪声环境中的信号检测进行研究,分析了其检测性能。与Langevin型的随机共振相比,该模型具有更高的局部信噪比增益和更宽的频率适应范围。数值仿真表明,噪声的非高斯性越明显,符号函数型随机共振系统的性能表现越优异。由实测数据对所采用的随机共振的有效性进行了验证。

Taking sign function as a stochastic resonance （SR） model, we researched a signal detection method in the non-Gaussian underwater noise environment and analyzed the performance of the detector. In contrast with SR based on Langevin function, the proposed model have a higher local signal-tonoise ratio gain and can be used in a wider frequency range. Numerical simulation indicates that, the more evident the non-Gaussian characteristics of noise are, the more excellent the performance of the detection method based on sign function SR is. The real data certify the proposed SR＇s validity.