Intelligent Reflecting Surfaces in VL-NOMA: A Network Signal Enhancement Perspective

This paper proposes an intelligent reflecting surface (IRS) in visible light non-orthogonal multiple access (VL-NOMA): a network signal enhancement perspective. A light-emitting diode (LED) with advantages such as illumination, programmable nature, and flexibility is used as a carrier to transmit the signal to two users. The LED employ the technique of power allocation (PA) to assign power signal properly to the two users. The potential challenge is on how to redirect the reflected signal from the LED-IRS-to the two users. The IRS has been deployed with deep reinforcement learning (DRL) approach to direct the reflected signals in an indoor system to route toward pointing to the two users which resulted in an improved signal-to-noise-ratio (SNR). This process network the LED-IRS-two users considering the polar angles. The proposed communication system is synchronized to yield an improved SNR with a significantly high channel capacity, and superimposed signals. We optimized the system to see how the reflected signal behaves. We observe that the signal with optimization performs better than ones without. We illustrated the synchronized LED-IRS-two users channel performance and superimposed signal employing the MATLAB tool.