Bistable optimal stochastic resonance single-frequency weak signal detection method based on frequency conversion

Abstract

The invention relates to a bistable optimal stochastic resonance single-frequency weak signal detection method based on variable frequency in the technical field of signal processing. The method comprises the following steps of: multiplying a single-frequency received signal r(t) and a local signal cos([omega]st+2[pi][delta]f*t) to perform frequency conversion; performing weighted summation of the received signal r(t) cos([omega]st+2[pi][delta]f*t) after frequency conversion and a locally generated zero-mean unit power resonance white Gaussian noise nSR(t); inputting the weighted sum signal [k1r(t)cos([omega]st+2[pi][delta]f*t)+k2nSR(t)] into a bistable stochastic resonance system to obtain an output signal-to-noise ratio SNRo of the bistable stochastic resonance system at a frequency [delta]f; performing maximum likelihood optimization on a weighting coefficient to obtain an optimal weighting coefficient; bringing the optimal weighting coefficient into the bistable stochastic resonance system to obtain a state variable output sequence of the system, and inputting the output sequence into an energy detector to obtain the energy of an output signal; and judging that a signal to be detected exists when the energy of the output signal is greater than the set energy threshold,. The invention has advantages of low calculation complexity, good robustness, high detection accuracy and strong feasibility and practicability.

Description

technical field [0001] The invention relates to a method in the technical field of signal processing, in particular to a frequency conversion-based bistable optimal stochastic resonance single-frequency weak signal detection method. Background technique [0002] In communication systems and signal processing systems, the detection of single-frequency weak signals under low SNR conditions is a common problem, especially when the signal spectral lines are submerged by background noise, the detection problem becomes more difficult. Some commonly used detection methods in the case of high signal-to-noise ratio, such as spectrum analysis, energy detection, etc., will seriously affect their detection performance or even become unusable as the signal-to-noise ratio decreases. [0003] After searching the existing literature, it was found that the relevant literature is as follows: [0004] 1. Least square method (O.Besson and P.Stoica, "Sinusoidal signals with random amplitude: le...

AI Technical Summary

Problems solved by technology

This method has good detection efficiency under the condition of high signal-to-noise ratio, but under the condition of low signal-to-noise ratio (such as the situation below -10dB), its detection accuracy is very low, and the practicability is not good.

[0005] 2. Higher-order statistics method (K.M.Hock, "Narrowband weak signal detection by higher orderspectrum (using high-order spectrum method to realize narrowband weak signal detection)," IEEE Trans.Signal Processing, vol.44, no.4, pp.874 -879, Apr.1996.) is another weak signal detection method that has received widespread attention. It is a detection technology based on the theoretical method of spectral estimation above the second order, but this type of method often has high computational complexity. , so that its application range is greatly limited, especially when the signal detection has real-time requirements, its weakness is particularly prominent

[0006] 3. Signal subspace method (A.Eriksson, P.Stoica and T.Soderstrom, "Second-order properties of MUSIC and ESPRIT estimates of sinusoidal frequencies in high SNR scenarios (MUSIC method and ESPRIT method second-order characteristics), "IEE Proceedings Radar and Signal Processing, vol.140, no.4, pp.266-272, Aug.1993.) is a method that uses the independent characteristics of signal and additive noise to emerge as the times require The method, combined with the eigenvalue decomposition of the covariance matrix of the received signal to estimate the signal subspace, has better detection accuracy than the previous two methods, but because this kind of method generally needs to pass through the signal in its implementation The process of eigenvalue decomposition also has high computational complexity, and the practicability is affected to a certain extent.

However, there is a serious problem in the energy detection method, that is, the detection probability is very low in the case of low signal to noise (such as the situation below -10dB), so the detection effect is also poor when the channel environment of the communication system is poor, and it has become a constant An important aspect restricting its application

Images