5G requirements mandate a latency in the region of 1 millisecond (ms), a steep decrease from LTE latencies of 10 ms or more. Will anybody notice? Radio frequency engineers will for sure, but the rest of us will not.

This is because we humans are slower than today’s wireless networks (but faster than 2G and 3G networks) and, unlike technology, we do not get faster as we evolve (in fact we get slower as we age). Even avid gamers do not perceive the difference when latency drops below 10–13 ms, the time it takes for visual stimuli to reach out primary visual cortex, where high-level visual processing starts. It takes us another 20–30 ms to perceive complex motion patterns. We are still not sure how much longer it takes to become aware of what we see (or how that happens), but it takes 150–200 ms to give a motor response to a sensory stimulus; that reaction time presumably involves some level of awareness.

Does this mean we will remain blissfully unaware of the blisteringly fast 5G RAN? Will 5G New Radio (NR) be a big waste of money? Not at all. We will not be able to appreciate the lower latency directly, but the networks will run much more efficiently and support applications, services, and traffic flows that would not work otherwise. These things we will definitely notice and appreciate.

The most salient impact of RAN latencies of 1 ms is on ultra-reliable low-latency communications (uRRLC) applications, such as virtual reality, augmented reality, autonomous driving, some industrial IoT and safety applications, or remote surgery. These applications can’t achieve the required reliability and quality of experience (QoE) unless the latency is very low.

Streamed content will also benefit, even though streaming itself is not highly sensitive to latency compared to, say, voice or interactive video. With lower latency, TCP-driven congestion will be reduced, and TCP optimization will be more effective. This will benefit not only streamed video (which accounts for a large share of wireless traffic), but all traffic, because overall network congestion will decrease. Subscribers will be pleased.

Yet the 5G 1 ms latency in the RAN is not enough to reliably support uRLLC applications or other latency-sensitive applications. Just as humans add latency upon receiving the signal, at the other end the signal accumulates latency as it travels the network, passing through the backhaul and fronthaul, wireless core, and transport to the internet, then all the way to where the content is stored and processing is done. Much of this latency is unavoidable. For instance, fiber adds a 1 ms round-trip latency every 100 km. When we look at the end-to-end latency in a wireless network, humans are no longer slow.

Often what we perceive as a slow network is not the result of capacity or latency limitations in the RAN, but the effect of latency from sources beyond the RAN. In most cases, these sources of latency cannot be eliminated or controlled, but they can ruin Quality of Experience (QOE).

But there is a way around this: move the content and the processing closer to the network’s edge, so the sources of latency are bypassed. This is what edge computing does. By bringing content and processing to the edge, it lowers what the subscriber perceives, while leaving the end-to-end latency untouched.

For instance, the transport latency will remain unchanged, but it will not affect a virtual reality application that is hosted at the RAN edge. Operators will find it easier to optimize the transmission of video content that is stored at the edge.

Does this mean we should move all content and processing to the edge? Most certainly not. Much of the content and processing does not benefit from being located at the edge, and indeed it is more effectively managed in a central location. What it does mean is that operators can select what to move to the edge and what to keep in a centralized location, depending on cost, demand and performance. In addition, the operator can reserve the low latency for the applications that need it.

In this perspective, it is the combination of low latency in the 5G RAN and edge computing that only when brought together can deliver the perceived 5G experience we are looking for. Not just a faster network, but a network that feels and looks faster.

ANALYST COLUMN DISCLAIMER

Statements and opinions expressed in articles, reviews and other materials herein are those of the authors; not the editors and publishers.

While every care has been taken in the selection of this information and reasonable attempts are made to present up-to-date and accurate information, SDxCentral, LLC cannot guarantee that inaccuracies will not occur. SDxCentral will not be held responsible for any claim, loss, damage or inconvenience caused as a result of any information within this site, or any information accessed through this site.

The content of any third party web site which you link to from the SDxCentral site are entirely out of the control of SDxCentral, and you proceed at your own risk. These links are provided purely for your convenience. They do not imply SDxCentral's endorsement or association. The copyright and any other intellectual property right any third party content belongs to the author and/or other applicable third party.