General Information

Network Slicing is one of the key technologies in 5G networks; this can establish an end-to-end logical network comprising dedicated and shared resources to satisfy diverse QoS requirements for heterogeneous services. In this work, using Real-Encoded Genetic Algorithms (GA), we have developed a novel approach to achieve the optimal inter-slice resource management, the objective being to minimize the number of unqualified users who have not obtained the required bitrate to meet their desired QoS, while minimizing the inter-slice radio resource isolation as estimated by a factor we term “Isolation Factor” (IF). IF is a novel concept proposed in this work to estimate the degree of inter-slice radio resource isolation caused by intercell interference between slices in different cells. The proposed real number encoding scheme for this method requires no prior knowledge about network traffic models and provides high scalability and flexibility.