Distributed RIS-Assisted Terahertz Communications: Performance Analysis and Optimization

Abstract

Terahertz (THz) communications are envisioned as a promising technology to meet the ultra-high-speed data transmission requirements of future wireless networks. However, severe molecular absorption and misalignment fading can significantly degrade system performance. To address these challenges, this paper investigates distributed reconfigurable intelligent surface (RIS)-assisted THz communication systems. To evaluate network performance, we first derive closed-form expressions for the cumulative distribution function (CDF) and probability density function (PDF) of the cascaded channel under both exact-case and worst-case conditions. Subsequently, we obtain accurate approximate expressions for the ergodic capacity and outage probability for both scenarios. To deepen the theoretical understanding, we further derive the corresponding diversity orders in the high signal-to-noise ratio (SNR) regime. Moreover, leveraging the derived theoretical results, we propose a low-complexity determinantal point process learning (DPPL)-based algorithm to optimize the RIS-destination (RIS-D) association and maximize the system utility. Extensive simulation results demonstrate that: i) expressions remain accurate for a wide range of system parameters, including the number of distributed RISs and reflecting elements; ii) cooperation among distributed RISs can effectively mitigate the performance loss caused by misalignment fading in THz systems; and iii) the proposed RIS-D association strategy significantly outperforms existing benchmark schemes while maintaining low computational complexity.