Research Of FMCW Signal Linearization Technique

FMCW radar in detection of hierarchy within the glacier, high resolution is needed todistinguish structure of each layer of the glaciers, and the key of system with high resolutiondirectly connected with the linear FMCW signal source. In actual FMCW radar systems, dueto the impact of device, environmental factors resulting signal source in the system isnon-linear, and launches a small nonlinear signal there will be beating the spectrum main lobestretching, side lobes increases, deteriorating system resolution. Since the accuracy of thedetected non-linear FMCW signal determines the accuracy of the linear, then studying indetecting precisely the nonlinear of signal source and nonlinear correction.Firstly, the paper introduces the principles of FMCW system, discussing the effects ofnonlinearity on the resolution of the system, analyzing the main factors of resolutiondeteriorating in the system derived from the nonlinear of signal in the FMCW system. Directhardware is an indirect method of nonlinear circuit detection, the paper uses direct nonlineardetection method, which uses dynamic detection of travel signal nonlinear frequencyestimation method and uses repetitive control theory to achieve correction. Applying twomethods for estimating signal frequencies: aPFFT/FFT integrated phase frequency estimationmethod, and sampling frequency of zero gravity estimation method, both methods can be highprecision estimation of signal frequency. Then establishing the linearity of LFM signalpolynomial model, analysis of derived mathematical relationship between the beat and LFMsignals through frequency estimation method for estimation on business determine the actualnonlinear LFM signal time-frequency relationship. Due to the heterodyne frequencyestimation of signal frequency estimation is a dynamic, it analyses the effect of dynamicfrequency estimation for frequency estimation method. Finally set a suitable samplingfrequency using the sampled zero-energy heterodyne signal nonlinear frequency estimationmethod for detection of gravity derived from actual LFM signal linear, based on repetitivecontrol system for nonlinear correction theory, experimental simulation results show thefeasibility of the nonlinear correction.