Spectrum Management for Wireless Resources in 5G Networks using Gated Graph Convolutional Network and Whale Optimization Algorithm

Recently, Artificial Intelligence (AI)-driven spectrum management in 5G networks leverages learning techniques to reduce allocation of resources and identify the patterns in traffic that leads to effective communication of wireless devices. However, it faced challengers include complex network and providing security to data. To overcome these challenges, various Machine Learning (ML) techniques are used to recognize patterns from complex data and make accurate predictions and Deep Learning (DL) techniques used to handle high dimensional data. Initially, the Multiple Input Multiple Output (MIMO) signal transmission is processed through Proximal Policy Optimization (PPO) for dynamic beamforming. Further, channel estimation is performed using Gated Graph Convolutional Network (GGCN) to enhance multi-user spectrum efficiency. Finally, Improved Whale Optimization Algorithm (IWOA) is utilized for spectrum optimization. The proposed GGCN-IWOA model acquired better results with 5.3 Mean Square Error (MSE), 4.2 Mean Absolute Error (MAE) and 2.8 Root Mean Square Error (RMSE) when compared to Convolutional Neural Network with Gated Recurrent Units (CNN-GRU) model respectively.