A Network Slicing-Oriented Model for Low-Latency 5g Communication

Abstract

The advent of 5G networks brings the challenge of effectively managing limited network resources over multiple virtual network slices with different Quality of Service (QoS) de- mands like low latency, high reliability, and assured throughput. This paper suggests a simulation and optimization framework implemented in Python that simulates major 5G components like clients, slice containers, and base stations. The architecture allows real-time setup and dynamic resource allocation to keep up with varying network loads. A literature review suggests that current systems usually function with static usage of resources ranging from 70% to 80%, and latencies between 40 and 60 milliseconds, which are not very flexible in terms of adjusting to real-time variation in traffic. Conversely, our solution showcases a 12.5% improvement in resource usage efficiency and a 0.8% decrease in mean latency. While the latency reduction seems marginal, it represents large performance increases on microsecond-order level accuracy in latency-constrained settings.