Edge networking and edge cloud are closely associated with 5G. It will be either impossible or very difficult to deliver advanced low latency services without hosting application logic closer to the user than in today's centralized network cloud architectures. It's not an overstatement to say edge computing is inherent to 5G architecture.
News feeds in recent weeks show how the edge has captured the attention of the larger, more sophisticated operators. Here are just three stories published by my editorial colleagues at Light Reading that demonstrate this focus:
I also recently published the results of an operator survey that asked about the role edge will play in 5G network and service strategies. The results show that the majority of respondents believe ownership of access networks and edge cloud assets gives them an advantage over off-network cloud providers. An excerpt from that survey is shown below.
Does ownership of access networks and edge cloud infrastructure enable mobile service providers to offer superior end-to-end performance guarantees for 5G compared with public cloud providers? (n=127)
(Source: Heavy Reading 5G Operator Survey on Performance Monitoring, 3Q 2018)
The chart shows how operators' opinions fall across four different scenarios in the Heavy Reading survey. A sizeable 40% of operators believe the ability to host applications in their edge data centers would give them a competitive advantage over cloud players (that do not have access or edge network assets). A further 22% expect advantages even where the service provider “hosts the application in its central data center.” The 11% is also important, as it indicates that even if an app is hosted in the public cloud, the operator can still play a role in assuring application performance. A total of 73% of respondents therefore think ownership of access network and/or cloud assets enables operators to offer superior end-to-end performance guarantees in 5G compared with public cloud providers. This sounds about right; the challenge is to prove it.
By definition, if you distribute functions and applications, you also need to distribute monitoring of the network access and edge cloud assets. And because this edge environment is no longer static, monitoring must be implemented in a more dynamic way. Implementing distributed monitoring under the old model of probes and taps on network equipment interfaces is simply too expensive and restricts the rate at which the network can adapt to changing demand or new opportunities.
As I discuss in a new Heavy Reading white paper, "5G Performance Monitoring: Competitive Advantage at the Network Edge," this issue of static performance monitoring and service assurance is one of the reasons why telecom operators have been unable to address market opportunities with sufficient pace. It's also why they need a new approach in 5G.
High-value services that transform a customer's business processes and outcomes clearly require committed service-level agreements (SLAs) with accurate and timely reporting, as well as the ability to anticipate and respond to changing network conditions. The good news, as I write in the paper, is that 5G architecture provides operators with a framework to define service-level monitoring. 5G enables them to gain a cohesive view of what a service needs in terms of network capability and the associated SLAs and key performance indicators and apply this from edge to core.
This blog is sponsored by Accedian .
— Gabriel Brown, Principal Analyst, Mobile Networks & 5G, Heavy Reading