Andy Sutton, principal network architect for UK mobile network EE and a visiting professor at the University of Salford, says 5G will be built around small cells and a heterogeneous network.
According to the good prof, whose institution is working with EE on streaming 4K hi-def TV, there will be a level of meshing between small cells, which will become the workhorse for future mobile networks, with the macro level – the traditional cells that cover large areas – becoming the control plane.
This will lead to very high bandwidth and low latency, and will keep the subscriber on the most appropriate network in terms of cell size depending on the circumstances; for instance how fast the subscriber is moving.
Professor Sutton sees some scope for device-to-device mesh technology in 5G, particularly for emergency services which might want to operate in a disaster area and thus need to cope with the networks being overloaded or destroyed.
Emergency services currently use TETRA, which can mesh between devices, but that radio technology is coming to the end of its natural life. The emergency services are now looking to 4G where it’s hoped the Release 12 specification will include device-to-device options.
Generally EE has seen the need for bandwidth as being asymmetric, with subscribers using ten times as much download bandwidth as upload – although there are exceptions. The amount of video uploaded from (EE-sponsored) Glastonbury was huge and for some users, like broadcast TV companies, EE is working on custom solutions, some involving multiple SIMs.
Ed Ellis from EE’s strategy and forecasting division is expecting a wholesale shift from fixed to mobile data usage. In 2014 data consumption was twice that of 2012 and is predicted to double again in 2015. By 2018 he envisages data usage as twelve times that of 2012, and as Vodafone has recently said this is heavily driven by the adoption of 4G.
Professor Sutton sees a lot of the progress necessary for 5G as building on 4G, which moved to an IP world with better tools and techniques like carrier aggregation. He points out that EE has a particular advantage in this because it has the spectrum of both T-Mobile and Orange – less the bit Ofcom made them sell.
He sees small cells, typically with a coverage of 50m to 150m, as important in growing 4G coverage. These cells will be the size of a router with a separate box to link to the backhaul and will typically be installed in places like lampposts. With each cell supporting 16 to 32 users the backhaul requirement is significant – “hundreds of megabits per second” – and Professor Sutton expects to see devices which aggregate three carriers to give around 400Mb/s, so future small cells will need to support them.
This will involve a mix of running fibre – either by digging up the ground or through existing power ducts – and direct radio links. Again, Sutton points to EE’s spectrum advantage for doing this.
Most interestingly, Professor Sutton sees 5G as a fundamental platform. Once we are there future improvements can be built on that and there will be no need for the kind of wholesale changes we’ve seen before. Typically there has been a ten year gap between new cellular technologies. TACS was launched in 1985, GSM around 1995, 3G in 2005 and next year we should see real numbers in 4G.
Professor Sutton argues that there will be a natural shift to an ever-expanding 5G network by 2035, unless there is a marketing ploy in renaming whatever evolution of 5G we have by then as 6G.
Still, this is looking twenty years into the future for an industry that is only thirty years old. It may be too soon to say “never”. ®