Research On Signal Processing Technology Of Real-Time Spectrum Analyzer

With the rapid development of wireless communication technology,time-varying and transient RF signals have become ubiquitous.In order to solve the problem of parameter measurement of these signals,real-time spectrum analyzers have emerged.At present,due to the late start of real-time spectrum analysis technology in China,the core technology is still subject to limitations.Although the relevant equipment can realize the basic real-time spectrum analysis function,there are still some improvements in the processing of the collecting signals.Therefore,the research on signal processing technology of real-time spectrum analyzer is of great significance and application value to promote the development of domestic real-time spectrum analysis equipment and improve the corresponding performance.The main research contents of this thesis include:First,in order to improve the flexibility of the real-time spectrum analyzer and solve the problem of too much data in baseband signals after digital down-conversion,this thesis proposed a down-sampling processing method based on variable analysis bandwidth.To begin with,according to the requirement of variable analysis bandwidth,the cascaded structure based on the CIC filter and the multi-stage HB filter is used to realize different multiples of decimation filtering.Besides,in order to ensure that signals in the analysis bandwidth is attenuated as little as possible,the designing criteria for the bandwidth of decimation filtering is presented.What’s more,an FIR low-pass filter is designed under different analysis bandwidths to filter out the out-of-band spurious signals.This processing method simplifies the designing process of the filter and saves the corresponding resource consumption.Second,the reasons of spectrum leakage in the real-time spectrum analysis are analyzed in this thesis.And the effects of different window functions on spectral peak information recovery are compared.Meanwhile,in order to improve the accuracy of spectral peak recovery,a method of phase difference based on the combination of symmetric windows and asymmetric windows is proposed,which can extract peak information quickly.Initially,two methods of constructing asymmetric windows are given.According to the simulation and analysis,it could be concluded that the symmetric window and the corresponding asymmetric window have different phase characteristics.Then,symmetric windows and asymmetric windows are added to the signal respectively.According to the previous characteristics of the asymmetric window,an algorithm for solving the real frequency of signals based on the secant iterative phase difference method is proposed.In addition,the algorithm is simulated and analyzed in different situations,and compared with the traditional windowing method.The results show that the proposed algorithm can obtain high-precision for the spectral peak information recovery with only a small number of iterations.Third,the overlapping frame processing and spectrum compression technologies are selected to compress the magnanimity data of spectrum output by FFT,which ensures the real-time property of signal processing.To start with,the functions and implementation methods of overlapping frames in the real-time spectrum analysis are studied deeply.And the simulation results show that overlapping frame processing can effectively improve the window effect and increase 100%probability of intercept(POI).Additionally,a method of the point display and vector display based on three-dimensional compressing is presented,which effectively solves the problem of difficult signal observation caused by high spectrum refresh rate.The noise brightness of the spectrum display is too high in vector display,in order to change this phenomenon,a multi-mode interpolation processing display algorithm is proposed,which eliminates pixel isolated points with multiple modes under different signal display requirements.The simulation results show that the algorithm proposed can effectively reduce the display brightness of noise while eliminating pixel isolated points.Furthermore,in order to ensure the diversity of spectrum display,according to the principle of video detection,a single spectrum display method under four spectrum frame accumulation modes is given.Finally,the simulations and comparisons of the display of instantaneous signals are accomplished by the above methods.The corresponding results show that the effect of combining overlapping frames with three-dimensional compression is better.