Study On Weak Signal Detection And Recognition Technology By Using Dsp System

In spectroscopy, physics, chemistry, medicine, optical communications, radar, astronomy and many other areas of research and application, the precise detection of weak signals always plays an important role. Noise inherent in any system, and when the signal to be measured is very weak, how to drown the noise of the useful signal in the high-quality extract weak signal detection must be a key task. Therefore, the core of weak signal detection is how to denoising and explore new technologies and methods to improve SNR of the detection system. DSP is greatly admired for its high precision, high processing speed and low power consumption in the field of signal processing. As the advantages of DSP, the thesis has deeply researched the detection of weak signal based on wavelet transform, and simulated and implemented the algorithms on DSP.After knowing the core mission of weak signal detection technology is denoising and a brief comparison of the traditional de_noising technology, the denoising technology based on wavelet transform is chosen. On the topic, study includes two, on one hand, after deep understanding of theoretical basis of wavelet transform algorithm and Matlat, simulation based on MATLAB of denoising based on wavelet transform verifies the correctness and feasibility; on the other hand, the same program in TMS320C6416DSP based on the simulation finally can process real-time signal. In the simulation some key technologies must be focused on, such as the choice of wavelet, the threshold choice and the establishment of such threshold function. However, the DSP test programs are more concerned about real-time signal processing difficulties, such as Gibbs, Signal interpolation and so on. A wide range in this paper, both the mathematical theory of physical phenomena and the derivation. Firstly, the paper includes not only the mathematical theory of physical phenomena and the derivation. Secondly, it also includes programming from theoretical modeling and simulation which can verifies the program. Thirdly, the paper includes how to build a hardware platform and to achieve the design and then optimization of the real-time signal processing process.