A hybrid \({Q}_{L}\) ANN model designed to improve the Quality of Transmission of optical communication network

Optical communication networks represent a fast-growing field in communication technology with numerous optimization and advancements to improve communication efficiency between systems. Introducing AI into various fields plays a key role in advancing optical network communication (ONC), comprising physical, data transmission, network, application, and transport layers. All these layers consist of several parameters, which should be optimized to provide better overall ONC. These layers must be dynamically optimized as they provide real-time data. AI algorithms can be programmed to optimize the parameters of different layers in ONC, like those, which involve the Quality of Transmission (QoT) parameters, and provide a better network. The QoT of OCN entirely depends on the network parameters and determined traffic data. The traffic data is determined by optimizing the network parameters. Thus, a Hybrid Q-Learning Artificial Neural Network (\({Q}_{L}\) ANN) model is proposed in this study to improve the QoT of OCN. The Q-Learning model preprocesses the network parameters and determines the traffic data. Based on the optimized traffic data, the ANN improves the QoT of OCN. The output of the Q-Learning model feeds into ANN as an input. In comparison, the ANN is trained with traffic data and network parameters and improves the QoT of the network. The performance of \({Q}_{L}\) ANN model is evaluated by comparing and verifying its performance with the existing models. The comparison results showed that the \({Q}_{L}\) ANN model provides better QoT than the traditional models, and it uses AI to improve the network’s performance more efficiently. The \({Q}_{L}\) ANN model obtained 99.5% accuracy with 96% of success transmission and proved efficient.