5Qs on 5G: With Cambridge Consultants
Founded in 1960 with the aim of putting ‘the brains of Cambridge University at the disposal of the problems of British industry and to provide solutions to real world problems’, Cambridge Consultants is now putting its mind to 5G technology.
Paul Beastall, director of telecommunications strategy at the firm, spoke to Inside5G about how the consultants view the next-generation network.
What is your focus in terms of 5G?
At present, our work is focused on strategic research and early stage technology research. We work to advise clients and develop specific technology for them. Our targets are driven by our clients’ requirements.
Client confidentiality is important to us but we are working on potential 5G technologies with a range of telecommunications operators, equipment vendors and technology providers. We are working with clients in the telecommunications and wireless industries in Europe, Asia and North America.
What are the biggest challenges for the industry as a whole in 5G development, standardisation and deployment?
There are a large number of challenges ahead, covering standardisation, development and deployment.
“some … requirements will either cause significant delay or drive compromises that will impact other areas of system performance”
In standardisation, the first significant challenge will be agreeing the key requirements in priority order and packaging those into appropriate releases that deliver key functionality in an appropriate timescale. Currently, there is a very broad range of use cases, requirements and expectations for 5G and some of those requirements will either cause significant delay or drive compromises that will impact other areas of system performance.
Only when the standardisation process is formally underway within 3GPP will those difficult decisions be addressed. The other key factor in standardisation is spectrum availability. New air interface technologies have been a key component of all “G” mobile technologies and each has required fresh, clean spectrum for early deployments. The World Radio Conference this year is too early to allocate spectrum to 5G, meaning the next opportunity will be in 2019.
“These issues are not insurmountable and there are many parallel applications that will drive innovation”
From a development perspective, the move to use millimetre wave technology will present new challenges, such as the ability to build microwave electronics in volume at a suitable price point. Although these spectrum bands have been used for communications for many years, they have focused on providing infrastructure services at different price points to consumer terminals.
These issues are not insurmountable and there are many parallel applications that will drive innovation, including automotive radar and Wi-Fi. The other key challenge will be delivering appropriate digital signal processing capability to support advanced modulation and coding schemes, and high channel bandwidths in an appropriate power consumption envelope and at a suitable price point. Whilst Moore’s Law has delivered, and will continue to deliver, an increase in performance, many of the algorithmic approaches being considered require a higher order of improvement in processing than simple doubling.
“Operators are likely to face two significant deployment challenges”
Operators are likely to face two significant deployment challenges. The first is the move to millimetre wave systems, where propagation and deployment is not well understood. Most operators understand radio propagation at bands below 3 GHz very well indeed and have spent years optimising the models they use for deployment planning. At present, research on how well millimetre wave signals propagate in urban areas, indoors and through walls and windows is nascent.
The other challenge will be the move from a relatively small number of large cells to a far larger number of small cells. This will present challenges in terms of planning and deployment, with future mobile networks looking far more like current fixed and Wi-Fi networks and requiring systems that can deliver high quality of service over a range of disparate technology platforms.
What do you see as the main use cases for 5G?
One of the key challenges faced by 5G in the short term is the refinement and prioritisation of use cases. We currently have a number of different applications which drive mutually exclusive technical decisions.
“5G may evolve to be a suite of different functionality and capability that can be optimised on a deployment-by-deployment or even application-by-application basis”
For example, Internet of Things networks need to be very low power and highly efficient, whereas potential applications to deliver true real-time services, such as those required for vehicle autonomy, need incredibly low latency.
This, coupled with the demand for just more of the same at a lower cost point when it comes to broadband services, suggests that 5G may evolve to be a suite of different functionality and capability that can be optimised on a deployment-by-deployment or even application-by-application basis.
This means that flexibility may well become the most important feature, although only if it can be delivered at a suitable price point. The key challenge with a flexible solution is that there is always a trade-off between flexibility, performance and price, which means there is a real danger of the industry developing a “jack of all trades, master of none” solution.
Have you identified any 5G technology of particular interest?
The first is the use of millimetre wave frequencies to deliver very high data capacity density and peak rates to users. This technology is starting to be commoditised in Wi-Fi 802.11ad and is essential to allow 5G to meet its aims of delivering multi-gigabit capacity. Such technology is reliant on new spectrum being made available for mobile services.
The second key area is in networking. The move to very high frequencies will mean that mobile networks will potentially contain millions of base stations, rather than the tens of thousands we see today. An increase of two orders of magnitude in base station count will require a very high degree of network autonomy and automation, rather than the current centralised planning approach.
“mobile infrastructure development to become more agile, with rapid iteration of functionality, rather than waiting for major releases”
Although these self-organising networks are already in early stages of development for 4G, a 5G network will need to be far more efficient at deploying, integrating and managing networks, as well as maintaining service continuity between different layers of a mobile network (from macrocell to femtocell).
Perhaps this massive uplift in scale and fundamental change in approach will finally allow mobile infrastructure development to become more agile, with rapid iteration of functionality, rather than waiting for major releases every couple of years.
This has already happened at the application layer where we don’t sit waiting for Google Release 12 or Facebook Release 8 and once the controls for heterogeneous networks have been agreed and messaging protocols standardised, vendors will be able to develop functionality independently of the standards process. The move to Network Function Virtualisation and mobile network functionality running on standard IT hardware will also drive the focus from hardware towards software innovation.
When do you think 5G can become a reality?
Although we expect early “pre-standard” 5G to be shown at both the Pyeongchang (South Korea) Winter Olympics 2018 and Tokyo Olympics in 2020, we don’t expect fully standardised and interworking technology to be available before 2022.
“unrealistic to expect fully standards compliant deployment much before 2022”
Although several technical elements that will form part of 5G are already being developed as extensions to LTE and Wi-Fi, the need to drive a fully specified standard before global deployment will extend timescales. Since 5G standardisation has yet to begin, and suitable spectrum will not be finalised before the World Radio Conference in 2019, it is unrealistic to expect fully standards compliant deployment much before 2022 .
Follow us on Twitter @Inside5G