Millimeter Wave Radiometer Signal Preprocessing Algorithm

The Millimeter wave radiometer, which is a kind of wideband radio receivers, works on the millimeter wave. As the wave has the good penetrability and high spatial resolution, millimeter wave radiometer can be used to image guns, knives and explosives hidden under the clothes. Therefore, millimeter wave radiometer has important practical value in military and civil areas such as remote sensing, blind dropping, navigation, security check and so on.In this thesis, the radiometer data acquisition and preprocessing techniques has been studied based on a certain program, including program design, algorithm theory, hardware development, software system and debugging.The main content of the thesis is as follows:(1) The channel equalization mismatch model has been studied specific to the systematic errors caused by the differences of radiometer channel gain. After the comparison, frequency domain filtering produced better results in getting rid of band noise, especially the periodic one.(2) In accordance with the characteristics of the radiometer's output signals and real-time processing, the output signals have been pre-processed by the improved LMS adaptive filter on the basis of linear prediction. Through simulation, the improved algorithm has better convergence than the traditional one.(3) Signal collection circuit has been designed and completed taking the FPGA as the core. Besides, the debugging of the whole system met all expected requirements and got correspond results.