ETSI Gets Edgy About Mobile
ETSI's solution to those problems is to propose that resources be deployed at secure aggregation points, rather than individual basestations, thereby reducing the number of physical locations and lowering costs. But because those aggregation points are not inherent to the architecture, their rollout could pose fresh technical challenges, according to Brown. Operators might have to introduce a so-called "bump-in-the-wire" -- a break in the system -- to apply certain protocols, for instance.
Nevertheless, ETSI regards the aggregation point scenario as one of two major spurs to mobile edge computing that have recently swung into view.
The other is 5G. Expected to hit the market in 2020, 5G technology will depend on much lower "latency" (the delay that occurs as data traffic moves from one place to another) to support critical services such as remote surgery. To minimize latency, 5G will have to bring service and application logic much closer to the end user. That means edge computing platforms will be an integral feature of 5G network design from the outset.
Yet the appearance of new communications networking trends such as 5G is also a complicating factor for mobile edge computing. Within ETSI itself, a separate ISG looking at network functions virtualization (NFV) -- which would allow network functions to run as software programs on off-the-shelf hardware -- is working on a distributed NFV infrastructure that seems to have considerable overlap with the specifications for mobile edge computing, according to Brown. "Why have two different management and orchestration systems?" he says.
ETSI says it is trying to address this duplication risk by ensuring its ISGs are in close alignment. "The intent is to re-use as much as possible from the ETSI management and orchestration [specifications]," says Neal. "We are at the early stages of this but the requirements on the MEC [mobile edge computing] side have been stable for some time."
Outside the organization, ETSI will also need to coordinate its efforts with those of the 3rd Generation Partnership Project (3GPP) standardization body to ensure the mobile edge computing specifications find their way into 5G. "When we have MEC standards available -- ideally at the end of this year or in March next year -- we can ask the 3GPP to accept this aggregation point scenario as part of its architecture," says Neal.
As he points out, however, the 3GPP has come up with its own proposals for aggregating the baseband units (BBUs) at basestations into a so-called "BBU hotel" serving many cells. Neal believes these are "complementary" to the work ETSI is doing, expressing optimism that "all these things [will] come together." But the emergence of different technology visions and associations is raising doubts about the role for mobile edge computing in a distributed cloud future.
Next page: Broader horizons