Joint optimization of a mirror array and power allocation in an OIRS-aided NOMA/OFDMA VLC system with mobile users and dynamic blockages

In this work, optical intelligent reflecting surface (OIRS) is introduced into the hybrid non-orthogonal multiple access (NOMA)/orthogonal frequency division multiple access (OFDMA) visible light communication (VLC) system with mobile users and dynamic blockages for the first time, to our knowledge. Specifically, the dynamic blockages are modeled by the random way-point (RWP) model, and two cases of movement patterns are considered for the mobility of users. The theoretical symbol error rate (SER) expressions for different users are derived and validated by Monte Carlo (MC) simulation, and one optimization problem is formulated with four critical parameters. A joint mirror array and power allocation (JMAPA) algorithm is then proposed to minimize the average SER. The superiority of the JMAPA algorithm is demonstrated by comparing the SER performance with other power allocation algorithms and different OIRS designs. Then, the impact of distance separation between APs on the SER performance is investigated with the JMAPA algorithm. Results show that with the assistance of OIRS, the SER performance of this hybrid system will be markedly enhanced, and compared to other bench algorithms and different OIRS designs, the JMAPA algorithm achieves superior SER performance and shows the best adaptability to different transmit power, while its effectiveness in optimizing the orientation of OIRS units is also validated. Moreover, the increases in the radius, height, and number of dynamic obstacles will degrade the overall SER performance with this algorithm while an increase in the velocity of the obstacles will enhance the SER performance in the dynamic environment. Furthermore, the distance separation between APs has an obvious effect on the SER performance.