June 22, 2017 —
It is no secret that edge computing is going to be a vital component in 5G. In fact, edge computing is being seen as one of the top enabling platform for 5G, Why? There are several reasons but one big one is that much of 5G is expected to run in the mmWave bands – at least fixed point-to-point (P2P) and point-to-multipoint (P2MP). But these two scenarios are ideal for the edge.
This is significant for the Internet of Everything/Everyone (IoX). Many of these devices are low power and need local nets to connect. Edge cells are idea for that. So far there has been a lot of talk about edge computing but little implementation. There are many reasons for that as well, but one big one here is putting enough intelligence into edge networks. That has been a bit of a challenge. Now, as small cells and DAS have seen significant technological progress in the last couple of years, that technology bump is coming to the edge.
One need only to look at some of today’s tech leaders. In this case, Amazon. The dominant online retailer has developed a platform, called Greengrass, which is an Amazon Web Services (AWS) platform that enables running run code at the network edge just like running it on the AWS cloud. This, of course, is proprietary and only given to AWS clients, but the fact that it is available speaks volumes.
This is, kind of, the model we can expect to see. Amazon does it with cloud servers, but in the 5G realm, this kind of approach should work extremely well. Edge clients, regardless whether they are part of the cellular network, an iteration of Wi-Fi, MulteFire or whatever, will benefit from the bump in technology. The theory is that, as it becomes possible and practical to embed more functionality at the edge, this will allow IoX devices to perform more complex, data-intensive operations. This, in turn, will ease the load on the core (because data won’t have to make the round trip for processing there) and bring communications to near real-time in many cases.
Of course, this won’t work for all scenarios, but developers say that a significant proportion of current functions, that are routed to the core and back, can be handled by intelligent edge platforms.
One of the big advantages to this is the real time element, which can have significant implication for things like medical emergencies and other critical use cases. A second big advantage is cost. Edge computing significantly reduces the hardware requirements of the core as well as improving efficiency at the edge, which reduces costs.
A second benefit is the ability to use the edge to manage IoX devices. Functions such as updating can be performed with minimal, or no downtime or use of core resources. This is especially advantageous with large numbers of devices by running updating functions locally.
There is no doubt about the significance that edge computing will bring to both the IoX and 5G. All that is left is to see how it will come together.