MEF Expands Reach to Cover Cloud Applications with IMTC Integration

28 Nov 2017

As some readers might know, I have been working in the industry on Software Defined Networks and how end-user applications can interact with these programmable networks. Today MEF announced the integration of IMTC into MEF with the mission of making networks more application aware. Allow me to explain.

The press release details that as more and more enterprises move their applications to the cloud, there is a need for cloud applications to deliver on a high-quality, secure and frictionless end-user experience. While the mission is accurate the details are much more encompassing. The strategy includes numerous convergences that are happening in the industry where Software Defined WANs (SD-WANs), virtualization and automation are driving the necessity for Communication Service Providers (CSPs) worldwide to move up the stack. Nothing explains this better than this video from Light Reading covering how leading North American tier 1 service providers see the trend of network transformation.

MEF in mid-November launched MEF 3.0 Transformational Global Services Framework, which helps operators worldwide automate their existing connectivity services, thereby driving down OPEX, while at the same time adding in new virtualized L4-L7 higher margin revenue services. If Information Technology (IT) is moving to the cloud, then CSPs worldwide are in the best opportunity to take advantage of this trend on numerous fronts:

1. SD-WANs – By unifying the underlay access transports (fiber, coax, copper, wireless, etc.) to an overlay networking model, common application and service assurance policies can be unified. This means CSPs are not operating each underlay transport separately but unified under a common overlay policy framework. MEF in July 2017 launched their initiative in Defining and Orchestrating SD-WAN Managed Services.

2. Evolution of SD-WANs – As more and more CSPs move into deploying SD-WAN as a managed service worldwide, the value of integrating Security-as-a-Service and Application Performance Management-as-a-Service becomes a natural evolution of SD-WANs than just a pure connectivity play. CSPs worldwide are in most instances in the middle of how information flows from the enterprise to the cloud. So CSPs are in the best position to secure and assure application performance. In October of 2017, MEF launched their Application Committee and placed SD-WAN in this committee to define not only SD-WAN managed services for connectivity but also to focus on application assurance. The diagram best describes this trend.

3. Network-as-a-Service (NaaS) – As more and more networks get virtualized in software, network software functions can be overlaid on top of existing physical networks so that new network services can be added and orchestrated in what the industry is defining as Network-as-a-Service. If you think about how hyperscale data centers evolved into utilizing network virtualization, from where SDN grew out of, this same paradigm shift is occurring in CSP networks worldwide. The virtualization of almost every network function is occurring at massive rates which equates to the velocity to deliver new services from years to months. Furthermore, SD-WANs are critical to the deployments of overlay NaaS deployments because it allows CSPs to get to off-net global access loops much like how easy it is to connect to the Internet. The diagram below best illustrates this trend. 

4. AI Analytics – Finally, the biggest trend that is now occurring in almost every industry, including networking, is to allow machines to operate the network in an autonomous fashion. Just like self-driving cars, Self-Organizing Networks (SONs) are being worked on in many R&D projects. So how do end-user applications interact with networks? Big data lakes are constructed where all kinds of unrelated and unstructured telemetry data are massively streamed into the data lake from networks, applications and even devices. Specially built AI applications look for events from the networks (ex: queue overflow), applications (ex: poor call quality) or devices (wireless interference) and then perform automated self-optimizing and self-organizing functions to correct the issue. There are many use cases when it comes to AI Analytics like zero touch automation, automated diagnostics and closed loop service assurance. The diagram below depicts the current trends in how applications, networks and devices all work together to deliver on increasing end user quality of experience by machines.

In conclusion, there is a lot going on in networking that is aiming to make end user applications like Real Time Media (RTM) work flawlessly from device-to-cloud.

Stay tuned as next month I talk about how APIs are being generated using a model drive approach, including future articles on Intent and policy frameworks.


There are currently no comments on this article.

You must be a registered user to make comments