Optimizing Network Slicing Through Neural Network and Particle Swarm Optimization

Network slicing permits the development of several virtual networks on a shared physical architecture, allowing for the provision of varied communication services. To fully realize the benefits of network slicing, efficient resource allocation is critical. In this study, an artificial neural network (ANN) is utilized in conjunction with a heuristic technique known as particle swarm optimization (PSO) technology to determine the optimal process parameters for network slicing. The developed technique is instantiated by a two-way functionality that responds to various industrial demands. Typically, the operators are constrained by project specifications or time constraints in an industrial arrangement. As a result, a methodology that has been evaluated exclusively on specific real-time data with limited input parameters consists of three primary phases: (a) data gathering, (b) optimization of optimal weighted features, and (c) categorization of slicing. The established technique frees operators from imposing optimization parameters and allows them to provide the best solution for the goal. The model's validation reveals a strong fit between expected and experimental findings.