LiDAL-Assisted RLNC-NOMA in OWC Systems
Ahmed A. Hassan, Ahmad Adnan Qidan, Taisir Elgorashi, Jaafar Elmirghani
TL;DR
The paper tackles interference and CSI uncertainty in dense indoor OWC by proposing a dual-function architecture that fuses RLNC-NOMA with LiDAL-based localization. It develops a comprehensive system model, including RLNC encoding/decoding, a LOS-dominant channel framework with imperfect CSI, and a MIMO-LiDAL sensing subsystem for passive localization. Key contributions include analytical expressions for decoding-success probability, LiDAL localization probability, and a CRLB-based localization bound, plus a LiDAL-assisted user-grouping scheme to improve throughput. Simulations show that RLNC-NOMA outperforms conventional schemes in decoding reliability, while LiDAL-enhanced localization substantially improves CSI estimation and sum-rate, highlighting the practical potential for integrated sensing and communication in next-generation indoor networks.
Abstract
Optical wireless communication (OWC) is envisioned as a key enabler for immersive indoor data transmission in future wireless communication networks. However, multi-user interference management arises as a challenge in dense indoor OWC systems composed of multiple optical access points (APs) serving multiple users. In this paper, we propose a novel dual-function OWC system for communication and localization. Non-orthogonal multiple access (NOMA) with random linear network coding (RLNC) is designed for data transmission, where NOMA allows the serving of multiple users simultaneously through controlling the power domain, and RLNC helps minimize errors that might occur during signal processing phase. This setup is assisted with a light detection and localization system (LiDAL) that can passively obtain spatio-temporal indoor information of user presence and location for dynamic-user grouping. The designed LiDAL system helps to improve the estimation of channel state information (CSI) in realistic indoor network scenarios, where the CSI of indoor users might be noisy and/or highly correlated. We evaluate the performance of NOMA combined with RLNC by analyzing the probability of successful decoding compared to conventional NOMA and orthogonal schemes. In addition, we derive the Cramer-Rao Lower Bound (CRLB) to evaluate the accuracy of location estimation. The results show that the proposed RLNC-NOMA improves the probability of successful decoding and the overall system performance. The results also show the high accuracy of the unbiased location estimator and its assistant in reducing the imperfection of CSI, leading to high overall system performance.