Based On The Svm Digital Beamforming Algorithm Research

Digital beamforming technology is core technology in modern intelligentantenna. We apply SVM method to digital beamforming technology of CDMAcommunication system, exploring the new method of the digital beamforming.The main contents and the completed work of this paper have:Firstly, introduce the basic idea and method of SVM, analyze theperformance and characteristics of SVM as a classifier, choose the properkernel function and optimize the parameters of kernel function. Formed thefoundation for established based on SVM digital beamforming technology.Secondly, study the several common optimal weight vectors criteria. Weknow the performance measures are equivalent in theory, but the performanceof the algorithm directed by different performance measure is different.Weknow SVM is based on BER criteri through study. Study the classic adaptivebeamforming algorithm: LMS algorithm and RLS algorithm, then analyze andcompare the performance of the two algorithm.Thirdly, through analysis the performance of the CDMA modulationsystem, digital beamforming can be regarded as classification problem. digitalbeamforming based on SVM method is applied to BPSK and QPSKcommunication system. Through the comparison of the SVM, LMS and RLS,show the SVM method superiority. Digital beamforming based on SVMmethod can applied QPSK communication system through the simulation test.And analyze the performance of SVM algorithm.At last, the geometrical structure of array is studied using SVM algorithmwhich is applied in the CDMA system and obtained the optimum range of thegeometrical structure. Then make sure the optimum geometrical structure ofthe linear and circular array. On this base, the performance of the linear and circular array was studied in the situation of unloading and overloading. WeAnalyse and compare the characteristics of the linear and circular array that inthe situation of unloading and overloading.