An Efficient Architecture for Latency Optimisation in 5G Using Edge Computing for uRLLC Use Cases

Fifth generation wireless (5G) technology includes a range of service categories, notably ultra-reliable low-latency communications (uRLLCs). uRLLCs use cases are distinguished by their exceptional reliability in delivering packets quickly. However, fulfilling the stringent low latency demands of uRLLC applications poses a significant challenge for 5G networks. Traditional central cloud-based architectures struggle to achieve these tight requirements, necessitating creative latency optimisation methods. Therefore, Edge computing emerges as a viable paradigm, with the potential to bring computational resources closer to end users and applications to achieve end-to-end latency of around 10 milliseconds (ms). In this paper, we propose an efficient architecture for minimising latency in 5G networks by moving the user plane function (UPF) to the network edge closer to the users, leveraging the control and user plane separation (CUPS) strategy. In addition, this paper further proposes the software-defined networking (SDN) based backhaul. Leveraging SDN in the backhaul network of 5G deployments allows operators to create dynamic, scalable, and efficient networks capable of serving a diverse variety of services and applications with varying performance needs. The experiment results from the 3rd generation partnership project (3GPP) compliant 5G testbed prove that the proposed architecture optimises the latency by 60%. Average throughput is also improved by approximately 40%.