Signal Detecting And Processing Of Millimeter Wave Sensor

This paper is affiliated one of the national defence preparatory study task, based on target recognition theory and signal processing. On this paper, the author designs the target recognition system of MMW detector using the universal DSP.Based on the martial application of MMW detector, this design combines the theory of target recognition and the signal processor and make the deeply study on both hardware design and software programming. Considering the highly request of real time in the system, in this paper we choose the programmed fixed-point DSP: TMS320C54x catena chip of TI inc. to design the target recognition system. This system can be fit for the real time job, further more it adapts to more complex target recognition software.In the hardware design, TMS320VC5410 is the core of this system, TLC5510 and AD7824 will accomplish the AD sampling, SST39VF200 is made use of realizing the date memory, Using these chips and other components we buildup an integrated independent digital process and control system. In the software design, the whole software system has been designed.Neural Networks is a large-scale parallel distributed signal process system, ability of the Neural Networks, and consequently acquire accurate classified precision. In this paper, we build the self-adapted target recognition arithmetic, and apply this arithmetic into the target recognition task of MMW detecting system, finally discuss the corresponding theoretical model and realization precept.