Deep learning based BER improvement for NOMA-VLC systems with perfect and imperfect successive interference cancellation

Abstract

This paper focuses in the improvement of the BER performance of multiple-input multiple-output (MIMO) systems is investigated utilizing non-orthogonal multiple access-visible light communication (NOMA-VLC). Applying multi-user downlink MIMO-NOMA-VLC system within equal gain combiner at the receiver is used with two types of modulation; On–Off Keying (OOK) and L-Pulse Position Modulation, with L = 4 and 8. The perfect and imperfect successive interference cancellation scenario is used in this system, and the scenario is considered for two and three users. Our proposed framework is divided into two stages. First, data is collected using the MATLAB software. Second, two deep learning models (DLMs); ResNet50V2 and InceptionResNetV2 which are trained and tested. Python software is then used to develop and train the DLMs. The obtained results assures the superiority of ResNet50V2 over InceptionResNetV2, in different cases and for all users. The BER performance is also studied versus α for two and three users OOK modulation single-input single-output (SISO), (2 × 2) and (3 × 2) MIMO-NOMA-VLC systems based on the two DL techniques; ResNet50V2 and InceptionResNetV2. Again, ResNet50V2 achieves better results than InceptionResNetV2. The obtained results are compared with the previously published ones, showing that the proposed system and techniques achieve better results.

Deep Learning-Based Detection Algorithm for the Multi-User MIMO-NOMA System

“…This approach enables the precise decoding of received signals at users through the SIC decoding technique after precoding at the transmitter. Reference [27] applies two deep learning models to the MIMO-NOMA visible light communication (MIMO-NOMA-VLC) scenario. Simulation results demonstrate that the models can achieve low BER in both perfect and imperfect SIC scenarios.…”