Lucky enough to be at Wimbledon's world-famous tennis venue in July, you are left slack-jawed in disbelief after Britain's Andy Murray, playing Canadian marksman Milos Raonic in the men's singles final, successfully returns a 147mph serve -- the second fastest in the tournament's history -- and then wins the point. Whipping out your smartphone, you watch an instant replay that confirms the athletic feat, in high-definition glory, before the next rally has even begun.
Such zippy localized service offerings have long been a talking point in telecom. Thanks to emerging technologies, they might soon become a lot easier to provide. By taking the compute-and-storage capabilities at the core of a mobile network and deploying them much closer to the end user, at the network's "edge," operators could boost performance to a new level. Instead of bouncing traffic on a costly, time-consuming journey to the core and back again, they could do all their processing in a flash at the venue itself. It would be like turning a looping tennis serve into a Raonic rocket.
Since March last year, the European Telecommunications Standards Institute (ETSI) has been working on the technical specifications for mobile edge computing (MEC), as it has become known. The stakes are getting higher. In the fast-growing market for cloud services, operators with edge infrastructure could boast a decisive performance-related advantage over cloud providers that rely on centralized resources, according to Gabriel Brown, a senior analyst at the Heavy Reading market research unit. "Theoretically, they could start hosting other people's services and content," he says. "It will be hard for providers without edge infrastructure to compete."
Yet despite all the promise, mobile edge computing has had a number of false dawns. As a concept, it has been around since the heyday of 3G, when the idea of using edge infrastructure to support mobile video traffic first took hold. Since then, it has been pummeled by a series of challenges, obstinately refusing to be knocked down. "The principle is sound and solid," says Brown. "The practical application has been a bit elusive."
Economics may have wrong-footed the technology in the early days. The original view was that compute-and-storage resources would be deployed only at the mobile basestation. That would have meant installing servers at many different sites, driving up costs substantially. Usage patterns also meant the benefits of this approach may have been hard to realize. "How many people have wanted to watch the same YouTube video at the same time in the same cell?" says Brown. "That 'cache ratio' has been too low."
Security could be a further headache were servers to be rolled out this way and user information stored at some distance from a safe zone. "A basestation might be in a farmer's field, where there is a possibility it gets physically intercepted," says Adrian Neal, the vice chair of ETSI's industry specification group (ISG) for mobile edge computing and a standards team leader at Vodafone Group plc (NYSE: VOD), one of the world's biggest mobile operators. "The way around that is to increase the physical security of basestations, but that comes with cost."
Next page: Edging ahead