Research On Signal Detection Method Of Millimeter Wave Radiometer Based On Stochastic Resonance

With the development of monolithic microwave integrated circuit(MMIC)and the miniaturization of weapon system,millimeter wave radiometer with many advantages such as small volume,light weight and good concealment has been paid attention to in military application.However,the output signal of millimeter-wave radiometer with quasi-real-time performance is often in the strong noise background due to its complicated working environment,noise jitter and measurement accuracy.The signal-to-noise ratio is low and affects the system performance.In this paper,the output signal of millimeter wave radiometer is detected and processed by using the method of stochastic resonance to improve the output signal-to-noise ratio and improve the detection performance of the system.The main work of this paper is as follows:(1)The stochastic resonance dynamic model,basic theory and common measurement method are introduced.(2)The effects of stochastic resonance parameters on periodic signal detection are analyzed.The normalization and the best matching principle of the structural parameters are introduced.An adaptive random resonance weak periodic signal detection method based on the best matching principle is proposed.The method is validated by simulation experiments.(3)The effects of stochastic resonance parameters on the detection of aperiodic signals are analyzed.The influence of the choice of iterative initial values on the detection results is studied.Based on this,the method of selecting the structure parameters and the iterative initial value when dealing with the aperiodic signals is improved,and the different types of aperiodic signals are simulated and verified.(4)The improved millimeter-wave radiometer output signal under strong noisy background is processed and the DSP hardware platform is realized by the improved method of weak aperiodic signal detection.And it is verified in the field experiment.