Study On Equalization Techniques Of Satellite Channel With Group Delay

In a satellite communication system with high data rate transmission, the presence oflarge numbers of group delay distortion devices, produces severe group delay distortion,resulting in aggravating inter symbol interference and the increased bit error rate.Equalization technology is one of effective measures in compensating signal distortionfrom the transmission channel. For group delay characteristic of high data ratetransmission satellite channel, this paper designed a satellite channel mode with lineargroup delay characteristics, then researched by adaptive equalization and blind adaptiveequalization technology, and did the following work:Firstly, this paper.analysed and studied the reasons and sorts of the satellite channelgroup delay distortion generated, and selected the most basic linear group delay distortionas the research object. According to the given criterions of group delay distortion,thispaper designed a satellite channel model with linear group delay characteristic, under thetheories of design finite impulse respons filter by the frequency sampling mothod.Secondly, based on the designed satellite channel model, according to the adaptiveequalization Least Mean Square (LMS) algorithm’s contradiction between convergencespeed and steady-state error, this paper carried on the thorough research.on variable stepsize of the LMS algorithm. Based on the analysis of existing algorithms, this paper putsforward a multi- parameter adjustable variable step size LMS algorithm. And through theMatlab simulations of each algorithm, verified the new algorithm effectively performance.Thirdly, the blind equalization algorithm introduced by the LMS algorithm occupyingbandwidth resource problem. And for the slow convergence speed and unsensitive withphase of the constant modulus algorithm (CMA) blind equalization , this paper analysedtwo classes of improved CMA : on improvement of convergence speed of CMA and onadjustment the phase of the modified CMA . On the basis of before, this paper put forwardthe modified CMA with variable step size, and compared the equalization performance foreach algorithm by application of Matlab simulation platform.Finally, according to the defect of CMA blind equalization algorithm for high order QAM signals in equalization, based on the existing blind equalization algorithm and theconstellation features of the high-order QAM signal, this paper proposes a new weightedmulti-modulus blind equalization algorithm. Through Matlab simulation proves that thenew algorithm has a highly efficient equalization ability under the premise of not need toset the threshold.