High-Resolution Channel Parameter Extraction Method And Its Application In Massive MIMO Channel Measurement

With the development of society,the demand for wireless data transmission increase dramatically,so it demands higher for the transmission performance of the wireless communication systems.In order to derive superior transmission performance,it is necessary to fully consider the effect of wireless propagation environments on the wireless communication systems.Channel models are usually used to characterize wireless propagation environments in the communication field.The establishment of effective channel models depends on the acquisition of channel parameters(delay,incidence azimuth,Doppler frequency,complex amplitude and so on).These channel parameters are obtained from comprehensive channel measurement data using channel parameter extraction method.Therefore,channel parameter extraction method and channel measurement are of great significance in establishing the channel models.Now,the research of 5G attracts much attention.As a key technology in 5G,Massive MIMO has superior performance such as improving system capacity dramatically and so on.In order to lay foundation for Massive MIMO channel modeling,the high resolution channel parameter extraction method SAGE and its application in Massive MIMO channel measurement are mainly studied in this thesis.First,we construct a simulation platform of radio channel parameter extraction based on the SAGE algorithm.The delay,incidence azimuth,Doppler frequency and complex amplitude of impinging waves can be accurately estimated under this platform.Second,we propose a cyclic maximization SAGE algorithm.When reestimating the delay and incidence azimuth,the proposed method cycles between the maximization of objective function with respect to delay for given incidence azimuth and the maximization of objective function with respect to incidence azimuth for given delay until the calculated sequences of delay and incidence azimuth converge.Simulation results show that the proposed scheme has smaller estimation deviation than that of SAGE.In addition,we optimize the maximization process of the SAGE by a multilevel search algorithm,ie.,decreasing the search step gradually in reestimating the incidence azimuth and Doppler frequency.Simulation results show that compared with the conventional SAGE,the computation load of the multilevel search SAGE dramatically reduce.Finally,we perform Massive MIMO channel measurements in Xidian University,use the SAGE algorithm for extracting angular power spectrum of impinging waves over the large antenna array and analyze non-stationary properties on large antenna array.The above work can provide reference for Massive MIMO channel modeling.