Enhanced Genetic Algorithm with Dispatching Rules and Diploid Multi-parent Evolution
Abhishek NM, Nitish Bhat, M. N. Vijayalakshmi, Akshata Patil, Ayesha Sotakanal · 2021 IEEE Mysore Sub Section International Conference (MysuruCon) · 2021
5G covers a wide range of use cases, all of which will be supported by the same physical network. To make this work, network slicing with customized network capabilities is designed to address the needs of the various users. This in turn faces a growing requirement for effective online algorithms to provide an optimal slicing strategy for efficient usage of resources. Genetic Algorithm (GA), being one such optimization algorithm, helps determine the slicing strategy while providing solid effectiveness, good robustness, and high scalability. In this work, we propose an enhanced GA by applying dispatching rules, a multi-parent Diploid evolution strategy improving the convergence rate of GA. The proposed enhancements in GA efficiently approach the optimal slicing technique with maximized network utility and an overall 71% improvement with respect to convergence rate compared to existing methods.