Research Of DS-SS Detection And Frequency Extraction Technology Based On Stochastic Resonance

Thiswork was supported by the Major National Science and Technology Projects (2013zx03001015), National Natural Science FoundationofChina(61379016、61271180).Signal detection and parameters estimation are the base of correct reception and demodulation of signal. They are also the prerequisite of spectrum management and help guarantee the system performance. Noise is generally considered harmful, and therefore most of the signal detection methods are designed to suppress noise. However, the process of suppressing noise will inevitably impair useful signal. The occurrence of stochastic resonance changes the traditional view. It puts forward that via stochastic resonance, noise can be transformed into signal power and enhance the useful signal. Because of the adoption of spread-spectrum technology, the power spectrum density of the direct sequence spread spectrum signal is usually very low and may submerged in strong noise. Consequently, the signal detection and parameters of DS signal is still challenging. Stochastic resonance, because it can use noise to enhance signal through nonlinear system, has advantages over traditional signal detection methods in processing weak signal in strong background noise. This thesis proposes an improved stochastic resonance model to realize DS signal detection and carrier frequency extraction.First of all, this paper introduces the development and application of stochastic resonance and analyzes the problems in its practical application. What’s more, based on Langevin equation, we explain the basic principle of nonlinear bistable stochastic resonance, introduce several classical stochastic resonance theory and some popular measurement methods and make a detailed analysis to the influence of each parameter on system performance. Secondly, this paper introduces high frequency DS signal into stochastic resonance processing. After the analysis of the basic principle and advantages of direct sequence spread spectrum technology, we use simulation to validate the effectiveness of stochastic resonance in processing DS signal. To address the problem of using traditional stochastic resonance to process high frequency signal, we adopt two high frequency stochastic resonance methods:modulated stochastic resonance and scale-transformation stochastic resonance.At last, based on nonlinear bistable stochastic resonance model, this paper proposes an improved stochastic resonance model and present simulation results to validate its performance. The detail procedure of the proposed method is:First, power modification module adjusts the power of input signal to satisfy the requirement of stochastic resonance. Next, we introduce ensemble average and relative detection into stochastic resonance systems to help suppress noise and make a further improvement to the signal detection performance. Finally, combine the modulated stochastic resonance and scale-transformation stochastic resonance to address the stochastic resonance of high frequency signal. At the end, we validate the effectiveness and improvement of the proposed model through theory analysis and simulation.