Cambridge Wireless Small Cell SIG ran a half-day event on rural and remote small cells, with expert speakers from EE, Nokia, Microsoft and Real Wireless. There is clearly both need and opportunity. Technical progress continues but the overall theme was that a change in operator mindset is needed on how best to deploy and support them using different processes from the traditional macrocell methods.
The title and theme of the event was “Is Small Cell thinking changing the face of remote and rural coverage”. As one of the SIG champions, Thinksmallcell helped organise the event and chaired the panel Q&A session. Our thanks to other champions, host Microsoft Research, presenters and the CW administrators who all contributed to its success.
Microsoft enable public Cloud cellular core network service
Microsoft Research Centre in Cambridge hosted the event, which like other CW seminars is not run along strongly commercial lines. Attendance is free or at minimal cost, speakers don’t pay to speak and you find this attracts more of the “hands-on” planners and integrators from across the UK.
One of Microsoft’s projects has been to run a core network EPC on their Azure cloud service, connected via an S1 interface to 4G small cells. The issues is that Cloud services have a much lower availability than telco-grade (say 99.9% rather than 99.999%). Bozidar Radunovic described how they have distributed instances of the EPC across multiple geographic locations and modified the S1 interface to handle any outage. The additional signalling latency introduced is still within tolerance, and the additional code involved was less than 2000 lines.
The relevance is that this might allow new organisations to operate their own small cells and core networks running on very low cost public Cloud services.
Nokia’s KUHA initiative enables local communities to install and maintain small cells
Enrico Nigra Gattinotta from Nokia UK spoke about the practical challenges of remote deployment and maintenance. Nokia’s KUHA initiative uses standard outdoor small cells (or in the future perhaps even Opensource ones) connected by public Internet broadband. The difference is a cloud OSS running on Amazon which can be accessed by local people, who would then see the status and possible remedies for any issues.
By using local expertise, such as an electrician or satellite TV installer, and simplifying installation to a “plug and play” basis, there is no need to send expertly trained staff to commission or repair each site. Installation involves connecting to power and backhaul, then waiting 10-15 minutes for it to self-configure. Anything other than power or backhaul failure might be more simply dealt with by replacing the unit. He gave an example of an LTE site installed in the Isle of Lewis in Scotland which was operational in a few tens of minutes.
Enrico felt there needs to be a change in mindset by operators. While these sites may not be as profitable as busy urban ones, nonetheless there is a positive business case for universal service and more widespread coverage especially where cheaper methods and local expertise are used.
ip.access reach Darkest Peru
For the most remote areas, ip.access gave an extreme example from Peru where it took two days just to reach the site. Again the financial figures look small but can still be profitable. Where a macrocell might serve 1000-2000 users, a remote small cell might serve 100. In this example, a village with 300 people attracted 91 users with an ARPU of $11 generating $12,000 per annum.
This was an EU funded research project called TUCAN3G. There are more details about it on the project website.
Real Wireless expand on the rural small cell business case
Julie Bradford from Real Wireless clarified the wider business case. They had researched a 50 page whitepaper for the Small Cell Forum which explained why rural small cells are cheaper than macrocells, detailing which countries and states would benefit most. The population distribution, even in sparsely populated areas, is for us to live in clusters – hamlets and villages – for social interaction. Targeting these isolated pockets with a small cell is much more efficient than widespread macrocell coverage.
The paper is quite extensive and includes many charts, covering both developed and developing countries. The slides and white paper are worth reading.
Small Cells take to the air
The final speaker was Ayan Ghosh of EE who spoke mainly about their rapid response research projects using both helikites and drones which I reported on earlier. It has applications including short term coverage for major incidents and outages, such as the Helikite solution architecture shown below. Their most recent initiative has been to experiment with backhaul using millimetre wave multi-sector product from Phazr, allowing them to add the 5G label to their project scope.
A consensus for the industry to take action
The session closed with a panel Q&A that I chaired. The general consensus seems to be that a mindset change is needed to encourage the industry to address rural and remote coverage. This should make use of local expertise and simpler methods than often used for macrocell deployments today. It may involve accepting a lower profit margin but not necessarily a loss. Some encouragement by regulators may be required, and this could take the form of opening up spectrum for use by others in locations underserved by incumbents.
Audience participation was good. Zahid Ghadialy of Parallel Wireless noted that where service is provided in locations without any wired or wireless internet, he’s seen subscribers buy up 100Gbyte monthly mobile plans and use it for everything from Netflix to being able to work from home for the first time. Nick Johnson of ip.access picked up on the spectrum discussion, indicating that the attitude towards spectrum sharing continues to thaw – he’s seen a growing willingness to consider using this method to deliver multi-operator service in remote areas using a single infrastructure. The panel agreed that while 5G virtual architectures might hold out a promise of doing this in a different way, this is still quite a few years away.
So in summary, there has been progress in addressing rural and remote coverage over recent years. This isn’t a technology issue, although aspects of self-configuration and open access to service status/faults will help. Instead, there needs to be a larger appetite for operators to address the problem. This could either be done in-house by setting up a separate division with different targets/KPIs to meet, or outsourced to a third party. Where insufficient progress is being made, regulators may need to step in to enable others to fill the gap.