Thursday, 22 December 2016

5G implies small cells

The UK’s National Infrastructure Commission (NIC) has blasted the level of coverage achieved with 4G and urged early action to deploy 5G more effectively. The organisation’s report particularly highlighted the role small cells will play in providing good services in urban areas, and on roads and railways, where the NIC says cellular coverage is “frankly appalling”. It has also looked further ahead and suggested new approaches to 5G spectrum allocation and usage, in order to open the market to hundreds of new service providers, supporting localized or specialized services. If adopted, these proposals would create a far more open landscape for 5G – but these big ideas should not be allowed to obscure the fact that 5G is not necessary to provide mobile broadband connectivity to the whole population. It is clearly essential that the “digital deserts” identified by the study are addressed and that the UK moves up the rankings in terms of 4G availability (it currently lies below far poorer economies like Albania, and countries with far more challenging size and terrain, like Peru). But the emphasis on 5G, in the report but also in the media and public discussion after its publication, is misplaced.

This is a coverage issue, and excellent coverage is very achievable with 4G, as other countries in the report’s international league table showed – for consumers at least, though deep penetration for some Internet of Things applications may require new spectrum and access point choices. Small cells will be essential to remove these deserts, bringing coverage cost-effectively to remote areas, as well as roads and railways, while adding capacity in areas of high usage. The NIC predicts that tens of thousands of small cells will be needed in urban areas to support 5G services, and calls for these networks, as well as roadside and trackside connectivity, to be in place by 2025. Coverage and 5G are two separate issues with different solutions But this is not a 5G issue, even if that label is necessary to generate headlines, and perhaps government interest. Better coverage for passengers and rural communities, and better quality of service in areas of high usage, are achievable now, using current technologies. Entering an international race to be the "first to 5G" is a red herring. The UK government does not need to emulate South Korea or Japan – its priority should be to deliver universal, predictable and good quality mobile coverage to all its citizens and visitors. That means creating the environment in which the mobile operators can achieve this and still make a profit. That is eminently possible using currently available small cell technologies for 4G and, in some cases, 3G. Examples round the world show operators improving coverage dramatically, and affordably, using 3G or LTE small cells. Some of these examples are in the countries which are highlighted as being ahead of the UK in mobile broadband availability – not just the 5G trendsetters like Japan, but nations like Peru or Colombia. The barrier to ubiquitous coverage is not technology – it is logistical.

To deploy cells in large numbers in cities, or in remote areas, simplified processes are needed to acquire sites (often municipal locations like lamp posts); to secure equipment approvals; and to deploy and manage the sites physically. National and local government can facilitate this with a stringent review of the regulations for deploying small cells, to accelerate build-out and reduce cost of ownership. The NIC report points to this, calling on local authorities to work with operators to enable small cell networks, and also to amend regulations to lower barriers to entry for new service providers, which might have a different business model for challenging commercial environments like remote communities. A streamlined framework to deploy small cells rapidly and cheaply would result in a dramatic improvement in coverage without having to wait for 5G. Indeed, there is a risk that, by holding out for 5G, the UK will fall even further behind. 5G networks will certainly be based on small cells, probably at a new level of density, but those should be planned to support new services, not as the solution to current coverage challenges which can be addressed today. The importance of shared spectrum In another recent proposal to the UK government, the British Infrastructure Group – a cross-party group of members of parliament - proposed compulsory roaming for all MNOs in rural areas, a suggestion which the operators have opposed vociferously. Small cell vendor ip.access is leading the push behind and alternative approach focused around spectrum sharing. In a statement in response to the BIG report "Mobile Coverage: A good call for Britain?", ip.access CEO Malcolm Gordon said: “The BIG report identifies 17m UK customers who experience poor reception at home and 525 areas with non-existent mobile coverage. We recognise the importance of this issue, and strongly believe that regulators and operators should commit to supporting shared spectrum as the most effective approach to connecting the unconnected in rural communities.” The company argues that a shared spectrum system, like a shared macro RAN, would present fewer technical challenges than localized roaming, because standards exist; would reduce capex and opex and so improve a difficult business case for the MNO; would still offer a choice of services to consumers and allow each MNO to manage and monetize its own subscribers; and preserve the first mover advantage for any operator which chooses to invest in the network (an MNO or potentially a neutral host third party), because they can reserve some capacity for their exclusive use. Radical proposals for spectrum Another important enabler of better coverage will be spectrum sharing and neutral host or shared networks – especially to improve the economics of rural deployment, and the logistics of rolling out small cells in dense areas where more than one physical network would be hard to achieve. Other recommendations were that any future Ofcom spectrum decisions should: support community or small provider solutions for underserved areas allow niche or localized providers to access new 5G spectrum rather than sticking to national licences alone open access to spectrum for enterprises, universities and others to use within their own buildings – even licensed spectrum where there are no interference risks

The third point would enable multiple wireless service provider approaches, including self-provision in the Internet of Things and robotics. More generally, the recommendations would open up a wholly new approach to the wireless market, which could set interesting precedents for other countries. Moving away from a model based on operator-owned spectrum and long, exclusive and national licences is already happening with developments like LTE-LAA and MuLTEfire, as well as other shared or dynamic spectrum ideas. But regulatory change will have to be radical, in the UK and elsewhere, to fulfil the real potential of 5G, to support large numbers of service providers, many of them industry-specific or regionally localized, which can go beyond a traditional MVNO and control their own spectrum and "sub-nets". This aspect of the NIC report was largely missed in the media debate, but is far more radical than its approach to the more urgent, but far simpler, issue of coverage for humans and smartphones, as opposed to all those unpredictable "things". It remains to be seen whether Ofcom and the UK government choose to be trailblazers in the move towards flexible, virtualized 5G networks and slicing. In the meantime, they should not use 5G as an excuse to delay bringing decent 4G availability to the whole UK population.