Segment Channel Modeling and Ricean <i>K</i>-Factor Estimation for RIS-Assisted NLOS Communications

Reconfigurable intelligent surface (RIS)-assisted communications have received widespread attention recently, whose channel modeling is the cornerstone for its future practical application. Nevertheless, the RIS-related channel modeling stays at the level of either the theoretical free-space derivation or the end-to-end channel model, yet its general segment channel model applicable to real environment is still rarely reported. In this paper, we formulate a temporal segment channel model for RIS-assisted non-line-of-sight (NLOS) communications, which separates the global channel into three parts, i.e., transmitter (Tx)-RIS, RIS-receiver (Rx), and Tx-Rx. Then we bridge the relationship between the proposed model and the existing models. In addition, we illustrate the Ricean K-factor characteristics of the global channel and the sub-channel in the proposed model. The lower boundary for the Ricean K-factor of sub-channel is also deduced. Two low-complexity estimation algorithms for the Ricean K-factor of sub-channel are proposed, whose accuracy is verified by the measured channel data.