Channel Estimation with New Basis Expansion Model for Wireless Communications on High Speed Railways

Time-varying channel estimation is a well-known challenge for wireless communications on high-speed railways. Existing estimation algorithms include interpolation, basis expansion model (BEM) and autoregressive models. These methods remain unchanged for every scenario on high speed railways (HSRs), such as viaducts, tunnels, cutting and open areas. Thus their estimation performance are not consistent and also not satisfactory in many cases. Noting that every train on one high-speed railway follows the same track and therefore the channel parameters are correlated at each fixed position. Based on this observation, we propose a new BEM that explores the correlation between current channels parameters and past ones. More importantly, we analyze its channel estimation performance and justify our suggestions. Finally, simulation results are provided to corroborate our proposed studies.