Software Defined Network Architecture Based Network Slicing in Fifth Generation Networks

Abstract

Network slicing is a promising approach that can cater to the need for differential services. It can be employed to improve the performance of users. In 5G networks, it is adopted to enhance the Quality of Experience (QoE) of the clients of differential services. In this research, 5G clients of five distinct services in different combinations in two scenarios are considered. They are eMBB, mMTC, URLLC, voice, and eMBB_p clients. This study develops SDN architecture-based network slicing in a 5G network. It is simulated using Python 3.7. Whenever the bandwidth available on the client-associated Base Stations is not sufficient to meet their service requirements, the process of network slicing is invoked. It consists of slice creation and slice allocation. Both of these modules are governed by Network slicing management. The amount of slice depends on the service requirements of the active clients. In scenario 1, the active clients are assumed to be 96. There are some combinations: 70 eMBB, 10 mMTC, 10 URLLC, 2 voice, and 4 eMBB_p. while in scenario 2, it is 52 eMBB, 21 mMTC, 15 URLLC, 24 voice, and 5 eMBB_p. As per the 3GPP standard, the bandwidth required for eMBB_p is 10 Gbps. The bandwidth available per Base Station is 6 Gbps; however, 1 Gbps is assigned for signaling, and hence only 5 Gbps are available per Base Station to serve the clients. From the simulation results, it is found that scenarios 1 and 2 achieve bandwidth utilization of 24% and 38%, respectively.