Transmissive Beyond Diagonal RIS-Mounted LEO Communication for NOMA IoT Networks

TL;DR

This work addresses spectral efficiency optimization for a downlink NOMA IoT network served by a transmissive beyond diagonal RIS (T-BD-RIS) mounted on a LEO satellite. It formulates a non-convex joint optimization problem to maximize the sum rates $R_i+R_j$ by jointly designing the LEO transmit powers $p_i,p_j$ and the BD-RIS phase matrix $\boldsymbol{\Phi}_t$, under QoS and power constraints, and channel/model assumptions including $y_i = h_i \boldsymbol{\Phi}_t x + n_i$, $y_j = h_j \boldsymbol{\Phi}_t x + n_j$ and $x = \sqrt{p_i P_t} x_i + \sqrt{p_j P_t} x_j$. The authors apply successive convex approximation (SCA) to linearize non-convex terms and decouple the problem into P1 for NOMA power allocation (solved in closed form by KKT) and P2 for phase-shift design (solved via SDR with a Taylor approximation, followed by eigen-decomposition to recover $\boldsymbol{\Phi}_t$). Numerical results show that the proposed framework yields higher spectral efficiency than a benchmark with fixed power and optimal phase shifts, with gains growing as transmit power and RIS elements increase; the work demonstrates the practical viability of T-BD-RIS for enhancing satellite-based IoT networks. Overall, the paper contributes a novel joint optimization framework for T-BD-RIS–assisted LEO-NOMA systems and provides insights into the benefits of transmissive BD-RIS in satellite communications. $R_i$ and $R_j$ are defined as $R_\eta = \log_2(1+\gamma_\eta)$ with SINRs $\gamma_i$ and $\gamma_j$ as given, and the optimization utilizes $W = \boldsymbol{\Phi}_t \boldsymbol{\Phi}_t^H$ and SDR-based reconstruction of $\boldsymbol{\Phi}_t$.

Abstract

Reconfigurable Intelligent Surface (RIS) technology has emerged as a transformative solution for enhancing satellite networks in next-generation wireless communication. The integration of RIS in satellite networks addresses critical challenges such as limited spectrum resources and high path loss, making it an ideal candidate for next-generation Internet of Things (IoT) networks. This paper provides a new framework based on transmissive beyond diagonal RIS (T-BD-RIS) mounted low earth orbit (LEO) satellite networks with non-orthogonal multiple access (NOMA). The NOMA power allocation at LEO and phase shift design at T-BD-RIS are optimized to maximize the system's spectral efficiency. The optimization problem is formulated as non-convex, which is first transformed using successive convex approximation and then divided into two problems. A closed-form solution is obtained for LEO satellite transmit power using KKT conditions, and a semi-definite relaxation approach is adopted for the T-BD-RIS phase shift design. Numerical results are obtained based on Monte Carlo simulations, which demonstrate the advantages of T-BD-RIS in satellite networks.

Figures (3)

• Figure 1: System model
• Figure 2: Achievable spectral efficiency versus available transmit power of LEO satellite for T-BD-RIS for the proposed framework and the benchmark framework, where the phase shift elements of T-BD-RIS is set as 64.
• Figure 3: Achievable spectral efficiency versus available phase shift elements of T-BD-RIS for the proposed framework and the benchmark framework, where the transmit power of LEO is set as 20 W.