Enhancing Non-Terrestrial Network Performance with Free Space Optical Links and Intelligent Reflecting Surfaces

TL;DR

This paper investigates a novel non-terrestrial network architecture combining free-space optical links, intelligent reflecting surfaces, and high altitude platform relays to enhance connectivity and performance in the post-5G era.

Contribution

It provides a comprehensive performance analysis of an innovative NTN architecture with detailed modeling of optical and RF links, including new asymptotic expressions and validation through simulations.

Findings

The system achieves improved ergodic capacity and lower BER with the proposed configuration.

Analytical expressions accurately predict system performance, validated by simulations.

High SNR asymptotic analysis offers simplified performance insights.

Abstract

The integration of non-terrestrial networks (NTNs), which include high altitude platform (HAP) stations and intelligent reflecting surfaces (IRS) into communication infrastructures has become a crucial area of research to address the increasing requirements for connectivity and performance in the post-5G era. This paper presents a comprehensive performance study of a new NTN architecture, which enables communication from the optical ground station (OGS) to end users through the utilization of HAP and terrestrial IRS nodes. In this configuration, the HAP acts as an amplify-and-forward (AF) relay terminal between the free-space optical (FSO) link and the RF links. Specifically, the RF links are modeled using the Shadowed Rician and the generalized Nakagami-$m$ models, where the FSO link is characterized by the Gamma-Gamma distribution with generalized pointing errors. The FSO system operates under either intensity modulation with direct detection or heterodyne detection. Using the mixture Gamma model, we approximate the non-centered chi-square distribution that describes the total fading of the RF link, and we assess the performance of the end-to-end system by analyzing the ergodic capacity, the average bit-error rate (BER), and the outage probability, calculated using the bivariate Fox-H function. We also provide simple asymptotic expressions for the average BER and the outage probability at high signal-to-noise ratio (SNR). Finally, the proposed analysis is validated with numerical and Monte-Carlo simulation results, showing an exact match.