| Introducing Non-orthogonal Multiple Access（NOMA）technology in indoor Visible Light Communication（VLC）is an effective way to achieve multi-user communication and networking applications.However,the NOMA-VLC system is vulnerable to channel environment interference,mutual interference between users,and difficulties in ensuring fairness among users,which leads to communication performance issues.To address these issues,Turbo coding technology and Compressed Sensing（CS）technology are used to optimize the system’s performance in this paper.Specifically,the research includes:1.The paper conducts research and analysis on the channel gain model of the VLC system in a multi-user scenario.The Fixed Power Allocation（FPA）method is adopted for NOMA power allocation based on the different channel gains of users,which enables multi-user communication.Turbo coding technology is used for channel coding to ensure reliable data transmission.On this basis,a NOMA-VLC system model for a multi-user scenario is established,and the communication performance of the system is analyzed through simulation.2.In order to improve the decoding efficiency of Turbo coding,the paper conducts optimized research on the iteration stopping criterion of the Sign Change Ratio（SCR）to enhance the decoding efficiency.An auxiliary decision criterion is established by introducing an external variable,which is combined with the original criterion to form a new decision criterion that ensures decoding accuracy while improving decoding efficiency.The performance before and after criterion optimization is verified through simulation experiments.3.In order to further improve the communication performance of the system using the CS method,the paper proposes a data compression method that combines Singular Value Decomposition（SVD）and Radial Basis Function（RBF）neural networks to improve the system’s efficiency.Based on the channel sparsity characteristic,the Sparsity Adaptive Matching Pursuit（SAMP）algorithm is used for channel estimation to enhance the communication reliability of the system and reduce performance differences among users.4.Based on theoretical analysis and simulation experiments,a NOMA-VLC experimental platform based on CS and Turbo coding is designed and established.The paper system is experimentally verified and its performance is analyzed in actual indoor environments.Theoretical research and experimental results show that after the introduction of Turbo coding,the paper system can guarantee accurate demodulation of data,and the threshold for the forward error correction rate is reached between two users at an average signal-to-noise ratio of11.76 d B.The improved SCR criterion can effectively enhance the system rate.Data compression using CS methods can improve system efficiency without significantly sacrificing reliability.Channel estimation based on CS can further improve reliability and user fairness.At a BER level of 10-3,user 1’s bit error rate increased by 0.85 d B,user 2’s increased by 0.96 d B,and the performance gap between users was optimized by 0.11 d B.Theoretical research and experimental results show that the system proposed in this paper,after introducing Turbo coding,can ensure accurate demodulation of data.When the signal-to-noise ratio（SNR）of user 1 is 11.47 d B and the SNR of user 2 is 12.04 d B,the bit error rate（BER）performance can reach the threshold of the forward error correction rate.The Turbo decoding algorithm optimized by the SCR criterion can significantly improve the decoding efficiency at low SNR.By using the CS method for data compression,the system effectiveness can be improved without significantly sacrificing reliability,as long as the compression ratio is no higher than 2.0:1.Based on CS-based channel estimation,system reliability and fairness among users can be further improved.When BER is at the level of 10-3,the BER of user 1 is improved by 1.59 d B,the error rate of user 2 is improved by 1.84 d B,and the performance difference between users is optimized by 0.25 d B. |
