Adrian Scrase, CTO of ETSI, presented a clear and coherent status report of 5G standards progress at Broadband World Forum. It cut through a lot of the hype and uncertainty, providing a useful status update on what it is and isn’t. Based on this and related research, I’ve summarised the status of 5G and considered the implications for Small Cells.
Not just another technology cycle
When I heard Hans Vestburg, ex-CEO of Ericsson, talk about 5G at Mobile World Congress last February, it was presented as just another 10 year repeat cycle of 3G or 4G. The big RAN vendors would invest 1,000’s of man years in R&D to create a wonderful new radio interface, operators would spend billions on new spectrum and everybody would make a good return on investment.
This just isn’t the case and 5G is quite different. Wireless communication is a huge business globally, and so there is strong political, economic and technical motivation to win market share. Several regions want to be first to market and establish leadership, both as suppliers and end users.
Requirements remain conflicting
There are three conflicting sets of requirements
- - Faster broadband data rates
- - Massive connectivity, supporting huge numbers of devices
- - Ultra reliable, always-on service with low latency
There has been a proliferation of “spider charts” showing the conflicting requirements (we reported one of the first of these shown by Moray Rumney of Keysight technologies in 2014).
Faster mobile broadband is considered to be the commercially easiest to monetise, and so has become the first priority. Expect the short term 5G claims to be all about peak data rates.
To improve requirements analysis, 3GPP has been engaging with industry stakeholders other than the traditional network operators and suppliers. They’ve opened up discussions with many organisations representing many different vertical market segments – everything from emergency services to agriculture. It takes time to explain and align the value of standards to those unused it them, but progress is being made.
There will be a completely new subsystem alongside the existing equipment:
- - A new radio interface (called NR), which supports very fast data rates.
- - A new Core Network (Next Gen)
- - An evolved LTE core network (based on the current EPC)
- - An evolved LTE radio
The New Radio will operate at very high RF frequency, above 6GHz, leading to short range and poor in-building penetration. The 5G radio component below 6GHz will be LTE with minor enhancements.
So 5G won’t replace LTE but instead will augment it, initially offering much faster data rates in targeted locations.
There is some talk of re-engineering IP (Internet Protocol), specifically TCP/IP, which has become a major bottleneck. It’s a provocative proposal but is gathering momentum. [Ed Note: There has recently been a considerable improvement in HTTP protocol, with HTTP/2 considerably improving responsiveness of websites, but this sits on top of TCP/IP and remains constrained by it].
When will 5G be available?
Several countries are very keen to be first to market, which has encouraged 3GPP to issue an early 5G Phase 1 specification that will be followed by a more comprehensive package in the subsequent release. Their current release schedule is:
- - September 2018: 3GPP Release 15 will include enough for an early 5G New Radio launch
- - 2020: 3GPP Release 16 with the more complete 5G
It’s reasonable to expect a period of around 18 months from formal publication to commercial service, although pioneers might launch trials more quickly.
Meanwhile 4G continues to evolve
We can expect the vast bulk of users and new applications to continue to rely on LTE, which itself will continue to evolve and expand. Shared and unlicensed spectrum use, through MulteFire and CBRS are good examples of what’s to come. But there’s also proximity/location services, vehicle-to-vehicle direct connection (even when outside network coverage), five-channel aggregation to combine up to 100MHz of bandwidth, public safety, public transport (eg railways) and much more.
Who will be the primary suppliers?
Clearly the major RAN vendors are making the most visible progress today, each with strong marketing and many live demonstrations.
There is a lot of political pressure to drive 5G differently from previous generations, augmenting rather than replacing the current technology. This includes introducing new vendors and eco-systems into the mix.
Adrian thought that the short range/high frequency 5G NR had Small Cells written all over it. These would require different installation and operation from macrocell towers, opening up opportunities for new entrants to become involved.
What impact will this have on today’s 3G and LTE rollouts?
I expect high speed 5G NR would be limited to a few dense urban areas and premier venues with high traffic, such as stadiums, concert venues, transport hubs. The short range and lack of in-building penetration will make it impractical to bring very high speeds throughout wide areas. As with any new technology, it will require new terminals/devices to support it and lots of infrastructure. It won’t be directly compatible with existing in-building systems, such as DAS.
This isn’t a quick fix to the more urgent issue of poor inbuilding service. It won’t be available in sufficient quantity or maturity by 2020 to combat forecast data consumption. For those simply wanting super-fast speeds, but prepared to stand still, Wi-Fi’s 802.11ad Gigabit datarates will already be widely available.
So instead I foresee a healthy and substantial future in LTE technology for at least the next 10 years, as it becomes more widespread particularly in-building.
5G may capture some headlines but won’t handle the bulk of wireless traffic for quite some time to come.
All the major RAN vendors have published white papers with their views of 5G
Several industry organisations have published their views on 5G, including: