Implementation of 5G Network Slicing Using Open Source Software

In modern communications, network slicing is essen-tial network operation for creating and managing an independent logical network which meets the requirement of a service. Net-work slicing enables available network resources to be managed and allocated efficiently and dynamically according to varying demands of users. This paper aims to design and implement 5G network slicing using open source software in order to serve very low latency services. The implementation requires the integration of orchestration, network function virtualization, and software defined network technologies to deploy necessary network functions and to manage the life cycles of network slices. Open source software tools such as Open5GS, Open Source MANO, OpenStack, OpenDaylight, and srsRAN are selected for implementing 5G core, orchestration and management functions, cloud computing platform, software defined network, and radio access network respectively since they offer compatibility in the software integration as well as allow the prototype to be created in a timely and economical manner. To obtain very low latency, User Plane Function and edge computing are implemented based on the latency of internal connections. In the experiment, different scenarios of network slices are deployed and tested to measure the key performance indices such as a delay, download speed and upload speed. The results show that the implemented network slices can offer very low latency at the expense of available bandwidth reduction due to the communication for orchestration and management in the hardware. However, this can be improved by using hardware with higher performance.