The Detection Of Weak Periodic Signal And DOA Estimation

Since people recognized that noise can be effectively utilized and transformed into a useful signal under certain conditions, the research on how to effectively use noise or convert part of noise into a useful signal has attracted widespread attention. The study ranges from the selection of stochastic resonance model, the adjustment of optimal stochastic resonance status indicators to the extraction and recovery of signal, and so on. This paper makes further research on the detection of the periodic Signal, firstly, this paper studies the self-adaptive stochastic resonance detection algorithm for weak periodic signals; Then, designning the array bistable stochastic resonance systems on the basis of this algorithm and combining with external periodic modulation signal method, it further studies the self-adaptive stochastic resonance detection of high frequency signals under the interference of color noise; Finally, aiming at practical engineering applications, it explores DOA estimation based on the self-adaptive stochastic resonance algorithm.According to the given system input signals of low or high frequency and the number of input signals, for the detection of low frequency signals, select the adjusting range of the system parameters by the threshold analysis and variable step size LMS algorithm, fix their step size and regulating system parameters automatically, and find the most optimized system parameters which results in optimal stochastic resonance effect; For the detection of middle or high frequency signals, first of all, preliminary processing of the signals is carried out by employing the array bistable stochastic resonance system and the autocorrelation function, so that the frequency of multi-frequency signals to be measured can be centralized and highlighted, then, approach the frequency of the signal to be measured and produce the difference frequency which satisfied stochastic resonance effectction by adjusting the frequency of the periodic signals, select the most appropriate indicators to measure the stochastic resonance phenomenon, finally, extract the frequency of the given signals in the frequency domain of the stochastic resonance output signals. Compared with the traditional self-adaptive stochastic resonance signal detection methods, it not only improves the work efficiency by considerable reduction of system parameters adjustment range, but also uses the color noise which is more consistent with the actual background; It is an effective solution to solve the problem of odd multiples of frequency which exists in the detection of multi-frequency signals, and thus results in more accurate extraction of the measured signal frequency and weaken the waveform distortion.Combined with low-frequency and mid or high-frequency self-adaptive stochastic resonance algorithms and DOA estimation, it improves the signal-to-noise ratio of the weak signal source by the stochastic resonance technology, and then estimates the spatial position of the signal source by DOA estimation. The simulation results show that it greatly increases the signal-to-noise ratio of the signal to be measured, and thereby extracts to the orientation of the target accurately. Simulation results are consistent with the theoretical analysis, which shows that the stochastic resonance technology of multi-frequency signals will have broad application prospects in the future.