Research On Self-interference Suppression Technology Based On Fractional Order LMS In Linear Frequency Modulation Continuous Wave Radar

Linear frequency modulation continuous wave(LFMCW)radar has been used in automobile anti-collision,pipeline laying,range finding,naval tactical radar,mine sweeping detection and other fields.However,the self-interference(SI)affects the performance of the radar and limits its application in other fields.Therefore,how to suppress the SI more fully is very important to improve the isolation between transmitter and receiver in radar.At present,the SI suppression method based on the least mean square(LMS)algorithm is widely used in LFMCW radar because of its efficient implementation structure and excellent cancellation performance.In order to improve the calculation accuracy of LMS in nonstationary signals and further improve the SI suppression performance,the fractional order gradient is used to calculate the cost function of LMS to replace the integer order gradient of LMS,so as to obtain the fractional order LMS(FOLMS)algorithm.Then,as for the problem of less information utilization and unstable calculation of FOLMS,the FOLMS algorithm is modified by the data reuse idea of the affine projection LMS(APLMS)to get the fractional order affine projection LMS(FOAPLMS)algorithm.Finally,the FOLMS method and the FOAPLMS method are utilized to SI suppression in LFMCW radar,respectively,and the performance of them are analyzed.Finally,it is verified that the performance of the proposed method is better than that of LMS method under the background of LFMCW radar application.The specific research of this paper is as follows.Firstly,the SI suppression model of LFMCW Radar is established.The basic principle of LFMCW radar,the influence of SI and the basic idea of SI suppression are introduced.Then,according to several typical SI suppression methods,the realization,advantages and disadvantages of different methods are analyzed.Secondly,the SI suppression algorithm based on classic adaptive signal processing is studied.The SI suppression algorithms based on LMS,NLMS and APLMS are studied respectively.The implementation steps of each algorithm are given and the stability conditions of each algorithm are analyzed.The evaluation factor of SI suppression performance is provided and the simulation and performance analysis of each algorithm are completed in LFMCW radar.Thirdly,the fractional order gradient is used to improve the traditional LMS to obtain the FOLMS,and the FOLMS is applied to the SI suppression process of LFMCW radar.First of all,according to the Caputo and Riemann-Liouville fractional order calculus theory,the fractional gradient descent method is derived.Combined with the derivation process of traditional LMS algorithm,the implementation of FOLMS is obtained.Secondly,FOLMS is applied to SI suppression in LFMCW Radar.From the non-stationary and time-varying characteristics of the adaptive SI suppression process,the influence of fractional order and step size on the FOLMS method is analyzed,which provides reference for the parameter selection of SI suppression in LFMCW radar.Then,we use the idea of data reuse of APLMS to improve the FOLMS,and propose a SI suppression algorithm based on FOAPLMS,so that the algorithm can update the complex weight value from different directions of the cost function quadric surface.This method can use more historical information of signal to estimate algorithm parameters more accurately and suppress SI more fully.Finally,the simulation results of FOLMS method and FOAPLMS method verify the correctness and effectiveness of the method respectively.The results of numerical simulation and actual measured data show that under the same simulation environment,although the computational complexity of FOAPLMS method and FOLMS method is higher than that of LMS method,the SI suppression using FOAPLMS method is the best,and the SI suppression performance of FOLMS method is higher than that of LMS method,which makes the echo signal is detected easily.