Top Three Trends at the Network Edge

By Casimer DeCusatis


Many leading computer industry analysts have been predicting a surge in edge computing and networking applications for the coming year despite the recent pandemic. The concept of bringing vital services closer to the network edge is being driven by a need to improve available application bandwidth, reduce latency, improve resiliency, and provide better security. There are many possible use cases including telecom networks that require service provisioning closer to end users—internet of things (IoT) systems, machine learning, augmented reality and more.  Since modern enterprises can use centralized resources for compute-intensive workloads and edge resources for real time applications, this approach compliments the hybrid cloud model. These topics will be discussed at the OFC Symposium on MEC-based network architectures in support of enterprise cloud. In preparation for this event, let’s take a look at some of the major industry trends for edge networking and cloud computing.

Machine Learning and Artificial Intelligence Algorithms

First, machine learning and artificial intelligence algorithms can benefit from edge computing, as Cisco will discuss during the OFC Symposium. Traditionally, applications that require a high volume of data traffic or high levels of complexity were challenging to deploy at the network edge. This has begun to change with the introduction of machine learning optimized hardware, containerized analytics applications and light weight implementations of machine learning frameworks such as TensorFlow Lite. Emerging open standards such as the Open Neural Network Exchange have also made it easier than ever before to deploy data analytics and machine learning solutions at the network edge. This is expected to enable faster automated decision making and allow for rapid customization of the user experience.

Cloud and Communication Providers Work Together

Second, a presentation by Microsoft will discuss cloud-native abstractions for network programmability. Partnerships between cloud and communication service providers are expected to drive big data analytics applications. Early phases of this work have already begun, as real time analytics expands beyond traditional enterprise locations into the network edge.  For example, the Linux Foundation Network includes several major carriers such as AT&T, Verizon, China Mobile, and Deutsche Telekom; this group recently added Google’s Cloud Platform to their membership. Not only is Google Cloud aligning with carrier interests, but Microsoft recently announced their Azure for Operators, a carrier-based cloud platform that includes significant edge capabilities. The major French telecom carrier, Orange, will also speak at the OFC Symposium about possible synergies between fixed access networks and mobile edge applications.

Edge Technologies

Third, leveraging the increasing adoption of 5G, edge device management will become a key differentiator for new types of service offerings. During the OFC symposium, leading optical networking company, ADVA, will discuss using 5G at the edge to provide latency-managed solutions for near-real time applications. Conventional approaches for cloud monitoring need to be modified so they work at the network edge, since edge technologies tend to be more volatile; operators need to monitor user experience, rather than simply doing health checks on their equipment. Cloud computing has already become a ubiquitous solution, and edge devices are poised for the same type of widespread adoption. This means there will be more demand for scalability and centralized control of the entire edge (such as edge-specific software defined networks).

Telecom and Enterprise Data Center Applications

At OFC this discussion will include both enterprise data center and telecom applications, both of which have different ways to leverage edge technologies. Telecom providers have been working on network function virtualization (NFV) for many years to replace specialized network appliances with lower cost, commodity servers and software. These solutions will likely be extended to the network edge. Enterprises, on the other hand, are likely to leverage Kubernetes for production-grade application container orchestration and other technology-driven solutions. In both cases, however, edge will become part of an overall distributed, hybrid networking strategy. It’s important to be aware of new developments in this area so that the edge doesn’t develop in isolation from the rest of the network. 

There are many other possible benefits, and this OFC symposium will bring together industry and academia to discuss a wide range of MEC technologies, pilot deployments, and use cases, as well as future roadmaps. This includes open RAN solutions, which are a huge potential business opportunity for enterprise users. As discussed in our prior blogs, low energy terabit applications have begun to emerge at the network edge. High efficiency, reliability, security, and privacy are all expected benefits from a comprehensive edge strategy. You’ll want to stay abreast of all these new developments, and the OFC symposia and special sessions are a great place to hear from leading developers about this exciting field.

What are your organization’s top three interests in MEC?  Drop me a line on Twitter (@Dr_Casimer), and maybe we’ll discuss them in a future blog. 

Attend the OFC Symposium on MEC-based network architectures in support of enterprise cloud 

Posted: 31 May 2021 by Casimer DeCusatis | with 0 comments

Comments
Blog post currently doesn't have any comments.
 Security code


The views expressed in this blog are those of the authors and do not necessarily reflect the views or policies of The Optical Fiber Communication Conference and Exposition (OFC)  or its sponsors.