DI5GUISE: A highly Dynamic Framework for Real-Time Simulated 5G Evaluation

Abstract

As the next generation of cellular networks (5G) nears deployment, autonomous smart nodes, such as Internet of Things (IoT) and vehicular devices, are being developed to take advantage of increased throughput and improved resiliency offered by the new link layer protocols of 5G. Evaluating the impact of real-time 5G air-interface scheduling algorithms on the stringent real-time delivery needs of these devices and their new applications is no small feat. In this short paper, we expand upon the mmwave (5G) module for NS-3 and present DI5GUISE, a configurable and highly dynamic real-time simulated testbed framework, through which the impact of these smart nodes can be investigated. We utilise real-time video delivery as an example of applications with high throughput requirements. The experimental results illustrate that even on low cost commodity hardware, such as Raspberry Pis, over 80Mbps per client of real-time streaming of 4K content can be achieved.