Outage Assessment of Mixed FSO-RF NOMA System Using UAV and IRS with 1-Bit Coding

Abstract

This paper characterizes the performance of a nonorthogonal multiple access enabled mixed free-space optical-radio frequency communication system assisted by an unmanned aerial vehicle relay and an intelligent reflecting surface (IRS) with 1-bit coding mechanism. Analytical expression of the outage probability (OP) is derived, and verified using Monte Carlo simulations, to assess the performance of the underlying system with 1-bit coding technique at the IRS. In addition, an asymptotic OP analysis is performed to determine the influence of critical parameters on the overall system performance. Furthermore, given the complexity involved in deriving the analytical expressions of the exact OP for each user under various channel conditions, we propose a deep neural network model for OP prediction. Moreover, numerical results reveal the efficacy of the proposed deep learning framework in predicting the OP performance.