Measurement-Based Markov Modeling for Multi-Link Channels in Railway Communication Systems

Multi-link transmission is one of the promising communication techniques capable of improving system capacity and cell-edge user spectral efficiency in railway communication networks. The performance of multi-link transmission heavily depends on the propagation characteristics of radio channels, especially the correlation property between multiple radio channels. Thus, it is important to characterize the multi-link channel in railway communication networks. In this paper, extensive wideband measurements in a viaduct railway environment at 460 MHz with two base stations and one mobile station are performed. Large-scale parameters (LSPs), including large-scale fading, Ricean K-factor, delay spread, and angle spread, are extracted from the measurement data. Based on the measurements, auto-correlation and cross-correlation properties of each LSP are investigated. A Markov-based multi-link tapped-delayline model for railway communications is established, where the Markov chains are introduced to model the birth and death state of multipath components in multi-link scenarios. Using the relationship between the correlation coefficients of complex random variables (RVs), the amplitude and the phase of taps with different delays are modeled as correlated RVs. The proposed channel model is implemented and validated with measurements.