NOMA-based RIS-IoV networks: Perspective of power allocations and RIS deployments

The paper compares the system performance of reconfigurable intelligent surface (RIS)-assisted non-orthogonal multiple access (NOMA) internet of vehicles (IoV) systems with different RIS deployment locations. Firstly, we distinguish two RIS deployment locations and two decoding schemes based on their corresponding NOMA power allocation coefficients. Then, using Jensen’s inequality, the upper bounds on the channel capacity of the two decoding schemes are derived. By considering the RIS location (from the base station to the RIS) and the geometric included angle of the near user-base station-RIS, as well as using the Taylor expansion of the binary function, we obtain an approximate expression for the RIS location that equalizes the upper bound of the two decoding schemes in the situation of high signal-to-noise ratio (SNR). In particular, when the channel gain difference between the two users’ direct links is too large, the gain provided by the RIS will be insufficient to bridge the gap in channel quality between the near and far users. Moreover, simulation results show that when both the distance and the angle are not fixed, the achievable sum-rates of the two decoding schemes will be equal when the RIS is at certain positions, which form a three-dimensional curve.