Recovering an unknown signal completely submerged in strong noise by a new stochastic resonance method

Unknown signal recovery plays always a crucial role in the discipline of signal processing. Especially, a signal completely submerged by a strong noise is more difficult to be restored and identified in the engineering fields. Here, we provide an effective method to recognize the types and related parameters of an unknown signal in a strong noise background. Firstly, the nonlinear vibration approach is adopted to enhance an unknown weak signal with the assistance of proper noise, in which a new quantitative indicator is designed to keep the resonance response to follow the unknown signal features. Subsequently, the polynomial fitting and the variance of the time difference sequence are implemented to estimate several important signal parameters. Finally, the frequency spectrum of the recovered signal is compared with that of the original signal to verify the correctness of the restored signal. Recovery results of three typical signals indicate that the proposed method is effective. Moreover, unknown weak signals are obviously enhanced and signal features are completely preserved. The proposed method successfully takes advantage of the energy of the complex noise components. This work may pave the way for recovering unknown signal from a strong noise background.