Unlike its 2G, 3G and 4G predecessors, 5G was not focused on a specific radio technology or spectrum band(s), but rather, on creating a new user experience. By the end of 2018, the 3GPP declared that any network using 5G New Radio (5G NR) was, in fact, 5G. The NR technology, together with a set of spectrum bands and the use of millimeter wave technology, massive multiple input, multiple output (MIMO) and new beamforming technology, forms the foundation of 5G's technical portfolio. Lost in the sands of time, however, is that 5G was defined around the user experience – whether that demanded one or more of the following:
- Enhanced mobile broadband (eMBB)
- Ultra-reliable low latency communications (URLLC)
- Massive machine-type communications (mMTC)
These three service attributes benefit from edge computing. URLLC and mMTC, however, demand it as latency requirements become more exacting and implementations scale. To gauge how carriers are planning and implementing edge computing, and in collaboration with Schneider Electric, F5 and Wind River, Heavy Reading asked over 90 global telco service providers about their plans for edge computing. In our report, "Edge Computing in Telco Networks: Gaining the Competitive Edge," Heavy Reading analyzes the choices these service providers are making with their implementation – and how these implementations are evolving.
Distributed virtual RAN poised to take off
One of the study's key findings, shown below, is that carrier use of disaggregated virtual RAN (vRAN) solutions is expected to triple in the next two years. This year, in fact, over half of respondents will have 10% or more of their 5G network running on a vRAN solution. Two years from now, over half will use vRAN solutions in 30% or more of their 5G network.
There are an estimated 70 million basestations in the world today. 5G is expected to increase this number by a factor of between 3 and 7. The move to vRAN and a software-driven architecture for 5G will both enable and simplify the move to edge computing.
Low latency applications need to bring the cloud closer
Emerging artificial intelligence (AI), virtual reality (VR) and machine learning applications demand round-trip latency, which is difficult to achieve if the origin and destination points for the application are distant from each other. Thirteen percent of service provider respondents believe they will be satisfied by a sub-100ms response time. However, the majority, 87%, are pushing for sub-10ms, with 38% pushing for sub-1ms in the next two years.
Sub-10ms one-way latency will meet most needs. This is the design attribute most vendors are building into their solutions – for now. It is clear, however, that some low latency applications at the edge may have even more stringent requirements of sub-1ms on the local short-haul connection.
As the network expands, so does demand for security… and the ability to manage the expansion
Security – both data en route and security tied to end-user safety – topped out Heavy Reading's list of required intelligent edge features, with 89% and 81%, respectively, of "very important" or "important" responses. Low latency, the best publicized aspect of 5G, came in third with a great deal of support behind it as an "important" feature. Reliability, resiliency and remote management all elicited enthusiastic responses. Carriers cannot afford anything other than a high percentage of zero-touch, lights-out data centers if they are looking at an edge footprint that is 5x–10x that of their current 4G network.
Read the full report for more telco edge insights
Heavy Reading's findings are a good indication that telco carriers are poised and ready for the business influx that edge computing will bring. For more in-depth details about the telco service providers' perspective on the burgeoning edge market, download and read the full "Edge Computing in Telco Networks" report now.
— Jennifer P. Clark, Principal Analyst, Cloud Infrastructure & Edge Computing, Heavy Reading
This blog is sponsored by Wind River.