Last year during reputed events like the NFV World Congress and the Open Stack Summit, a powerful insight that was gathered is that Multi-access Edge Computing is the hot topic of every conversation. We have heard Google talk about the next generation Google Assistant, which promises to evolve search in new and incredible ways. You would be able to stand in front of a public statue and simply say to your phone “Who made this?”, your phone will be able to capture the contextual data of your location and not only will it tell you the name of the statue but also offer information about the artist and other interesting facts. Extend that use case out and at some point you can actually interact with a digitally enhanced version of that statue. These are some of the promises of the near future.
But how does the promise actually become a reality and what happens when these practices become commonplace? When we see the drivers of this technology in the current infrastructure, what we come across is enormous inefficiency due to the current architecture and traffic flow. If a command has to be sent to the core of the network to be analyzed and responded to, the response time that itwill take would be considered unacceptable. Imagine the resource requirements for these sorts of applications when one user is multiplied to become many thousands or even millions. The result is a massive strain on the network.
Enter Mobile Edge Computing
The way to make these sorts of resource heavy applications viable is to migrate compute power to the edge of the network. This approach is known as Mobile Edge Computing (MEC.) By placing cloud computing capabilities at the edge you effectively create a “fog” which enables mobile operators, service and content providers to capture value from the proximity, context, agility and speed that an edge compute solution can provide.
Service providers around the world have put their focus on this technology when they consider growth opportunities. Till last year, MEC stood for Mobile Edge Computing; however a MEC group discovered that apart from mobile access this technology has other benefits as well. Hence the term Multi-access Edge computing was coined.
Service providers are actively going after MEC projects; this is because of both the operational cost reductions that are promised by the application and the opportunities to generate revenue from new customers. By bringing content and applications to data centres present in Radio Access Network, MEC enables service providers to introduce new services which were unachievable via cloud hosted architectures due to latency or bandwidth constraints.
Service providers consistently identify four types of applications that they see around MEC:
- Small cell services in locations with high density, like stadiums, where stadium owners and service providers can offer a host of personalized content to the fans real time by deploying an application at the network edge. This will also reduce backhaul loading, as new content would be created and delivered locally.
- Augmented Reality, Virtual Reality and other internet applications require a super-fast response time, local image analytics and low latency communication. Apart from this remote medical diagnostics and tele surgery also demand quick response times. In all these cases round trip communication with a remote cloud data center is not sufficient.
- Vehicle to Everything communication requires high bandwidth, low latency, availability and robust security. Whether it is Vehicle to Vehicle or Vehicle to Infrastructure, a cloud hosted model would not be able to meet these requirements.
- Content providers have learned that video quality is critical for retaining subscribers. Reports have shown that a one second rebuffer in a 10 minute clip can cause a 43% drop in user engagement. This has led to mobile HD video and Premium TV come with end to end Quality of Experience which enables smarter utilization of resources. MEC enables service providers to optimize video content, ensure a fast start and smooth delivery.
No single company has expertise in all the product areas required for a complete end-to-end MEC solution. It’s a classic opportunity for effective ecosystem-based collaboration. Delivering MEC services that constitute a great user experience while minimizing operational costs is not easy. It requires a combination of application software, platform software and hardware that have been optimized for this application and integrated to work together seamlessly.
Wind River, an Intel company, is a world leader in delivering software for the Internet of Things. Wind River offers the industry’s most comprehensive embedded software portfolio, supported by world-class global professional services and support and a broad partner ecosystem. For more information, visit www.windriver.com