Performance Analysis And Key Technology Research Of Direct Sampling Microwave Radiometer

The direct sampling microwave radiometer (DSMR) is a novel microwave radiometer with simple structure, which substitutes wide-band, high-speed ADC of conventional microwave radiometer for mixer, local oscillation analog circuit and so on, direct samples the radio-frequency signals. Due to trading analog complexity for digital ditto, traits such as light weight, high stability, low power consumption, easily processing and so on make the DSMR very suitable to apply in the remote sensing as a space-borne microwave radiometer. The DSMR is one developing sub-areas of microwave radiometer.Due to the impacts of low SNR and nonideal ADC quantification, the sensitivity and measurement accuracy of a total power DSMR are low. Furthermore, taking phase error of RF front-end, sampling error of high-speed ADC and the size of array into consideration, the image quality of synthetic aperture DSMR is still poor. Therefore, it is very significant to improve performances of DSMR by study in the signal receiving mechanism, signal statistics of the DSMR receiver, analysis performance of the DSMR, and study in key technology. The main research aspects of this thesis are:The paper establishes the sensitivity expression of total power DSMR based on the nonideal quantization model of ADC. The utilization of digital filter reduces system noise and that of over sampling and spectral estimation increase equivalent integral time, respectively, so as to improve the sensitivity of total power DSMR.My research considers the actual ADC nonideal quantification, equals ideal quantification and the polynomial nonlinear model to be nonideal quantification. It infers the nonlinear calibration equation of the total power DSMR based on the equivalent model; carries on the calibration to the total power DSMR by the multi-point calibration method, and enhances the measuring accuracy of the total power DSMR.According to the imaging characteristics of synthetic aperture DSMR, the paper presents a simplified method to produce quardure signals using digital Hilbert transform of the in-phase data collected, which can eliminate the quadrature errors. It analyzes how the in-phase errors, clock jitter and ADC nonideal quantification affect the cross correlation and how the nonideal ADC affects microwave image performance of synthetic aperture DSMR; uses an adapted signal processing method to calibrate sampling errors of ADC to improve image performance of synthetic aperture DSMR.Since the imaging model of synthetic aperture DSMR is the undetermined linear equations, and both "observation space" and "system response matrix" are corrupted by noises, the image formation model does not have the unique solution. It obtains the optimization solution by using the total least square method to reconstruct microwave radiation image; carrying on microwave image processing by image partition and the median value filter to enhance the quality of microwave radiation image.