What is it about?

Ultra reliable low latency communication has been there for many years in 5G standard. But is it really feasible and is it really implemented now? This paper examines these two question in current 5G standard and deployments. We claim that the latency should be looked at in an end-to-end manner and as a system. We uncover the latency bottlenecks in an accurate system level analysis. We categorize the latency sources in the system and show their interdependency. We find all the possible configuration that can on the paper achieve the URLLC requirements and we discuss them in the practical scenarios. We acknowledge that even those configurations would have scalability, availability, and reliability issues. We also study the non-determinism of sources of latency in 5G systems and their impact on overall latency.

Featured Image

Why is it important?

We do the system level end to end analysis of the latency in 5G networks for the first time through a system level analysis while considering all the interdependencies and the practical considerations. The analysis done in the paper can help the community to understand the URLLC systems and the nature of latency in cellular networks. We claim that the current standard has unrealistic requirements for latency (0.5 ms one way latency) and reliability. We warn about the forthcoming 6G standard which aims an order of magnitude less latency (0.1 ms one way).

Perspectives

I believe we might have been too optimistic when we were defining our expectations of the latency in 5G systems. Today, 6G discussion is talking about an order of magnitude lower latencies. It is important to be conscious when defining our expectation of future 6G networks and have a concrete plan of how this level of latency can be achieved in practice.

Arman Maghsoudnia
Ecole Polytechnique Federale de Lausanne

Read the Original

This page is a summary of: Ultra-Reliable Low-Latency in 5G: A Close Reality or a Distant Goal?, November 2024, ACM (Association for Computing Machinery),
DOI: 10.1145/3696348.3696862.
You can read the full text:

Read

Contributors

The following have contributed to this page