Characterizations on Non-stationary Channel Statistics for High-Speed Train Communications in Viaduct Scenario

In this paper, results for non-stationary channel statistics based on the wideband high-speed train (HST) channel measurement in the rich-scattering viaduct scenario are presented. Due to the fast movement of the train, both the fading characteristics and the Doppler-delay statistics change rapidly with time. The measurement campaign is implemented on the Guangzhou-Shenzhen passenger-dedicated railway at a central frequency of 2.4 GHz. The parameters of multi-path components (MPCs) are estimated by the Subspace-Alternating Generalized Expectation-maximization (SAGE) algorithm. The large-scale fading properties and the small-scale Doppler-delay statistics, including path loss, shadow fading, Rician K factor, number of MPCs, root-mean-square (RMS) delay spread, RMS angular spread, power-delay profile (PDP), and Doppler-power spectrum are analyzed. The non-stationary statistical models of the above parameters are given based on the measured results. The correlation property is characterized by the correlation matrix distance (CMD). It is shown that the HST channel in the rich-scattering viaduct scenario has strong non-stationarity characteristics. This work provides references for the design of HST communication systems and promotes technical developments.