Latest Comments

  • PREMCHANDRA J LOKHAN... More
    i support ericsson as rcom can dupe anyone , they had... Sunday, 14 October 2018
  • Bud Biswas More
    Our company, Polaris Networks, has helped other smaller... Friday, 12 October 2018
  • Developing Telecoms More
    That is correct - it is the coastline of Equatorial Guinea,... Friday, 12 October 2018
  • Xavier Muñoz More
    This photo is not from São Tomé e Príncipe Thursday, 04 October 2018
  • Hello,
    My name is Adewale. I am a Healthcare Manager in... Friday, 21 September 2018

Signal mapping charts path for affordable, effective rural coverage

Signal mapping charts path for affordable, effective rural coverage

For operators in many emerging markets, delivering remote coverage in a cost-effective manner is hard enough – but working out how effective that coverage actually is can be even more difficult.

Varied topographies, scattered population centres and challenging environmental features can all be factors that cause discrepancies between the coverage that telecom sites are capable of delivering and the connectivity that customers actually receive. Deploying infrastructure to remote areas is costly, so it’s unpalatable for operators to discover that their coverage falls short of their investment.

While it’s possible to use small cells to plug the gaps in the coverage delivered by larger sites, doing so requires knowledge of where the gaps are – and this requires accurate coverage mapping. In a continent as vast as Africa, sparsely-populated regions are a given, and delivering coverage to these regions cost-effectively is a holy grail for operators. Now, Vanu, Inc., in cooperation with the international non-profit FHI 360, has created VanuMapsÔ, a new tool that it believes will facilitate this mission, mapping coverage using a method that the company believes is superior to current coverage estimates. Developing Telecoms Editor James Barton spoke to Vanu’s CEO, Andy Beard, to discover more.

----

Vanu’s objective is to provide coverage for the 1.2 billion people in the world who don’t have connectivity today. This involves figuring out where these people are; you need to know where the coverage is and, conversely, where there is not coverage. We spent a lot of time thinking about different ways of approaching this, and have identified a way of doing this using a series of data sets that include tower locations, terrain, and population using propagation calculations, along with algorithms for joining together numerous slices of what this propagation looks like from a given site or collection of sites. It’s then possible to aggregate a very compelling vision of connectivity.

For anyone who is looking to invest in connectivity within these markets, whether that’s Vanu or a mobile network operator we’re partnering with, this mapping process has to be not only comprehensive, but also high resolution – the last thing you want to do is assume an area is covered when it isn’t or vice versa, because either one of those can undermine the viability of your business case. You don’t want a narrow but deep view of one square mile in, say, Uganda – that’s not sufficient data to build a network; you have to have a global view. You have to be able to say “in this market, there are 15 million people without coverage and by building 200 sites we can cover 4 million of them.” You need this kind of granularity to quickly assess the market, decide if you want to operate there, and then work out how much of an investment is required to address the market in a meaningful way. That’s what VanuMaps is intended to support.

What VanuMaps Can Tell Us & Why It Matters

Green, yellow, orange and red denote varying levels of estimated coverage.  Uncolored background space (white space in this picture) denotes no coverage.

In Africa, the coast is moderately well covered, and some inland areas are covered with comprehensiveness and viability, but there are also vast swathes of the continent’s interior with no coverage. This high-level view of the whole continent took a lot of data processing to generate the underlying information and stitch it together into an overarching map, but it’s doable – Vanu’s also made a coverage map of Southeast Asia and we are currently producing maps for Central and South America.

vanu coverage estimate

The comprehensive view of VanuMaps is highlighted when contrasted with other available datasets, including one released by a major industry trade group that shows an area of south-western Uganda with circles of coverage. While this could in theory be accurate if the terrain was as flat as a pancake, there were no obstructions, and you were getting ideal propagation, that’s not reality.

vanu coverage accurate

By contrast, Vanu’s map of the same area looks very different, but it’s based on the actual tower sites. It’s also an aggregation of all the operators – there are a lot of holes in the aggregate coverage. A lot of towers will paint the tops of hills, as while a tower on top of a hill may cover one valley, it will go straight across and hit the hilltops on the other side of an adjacent valley. The coverage on our map is almost overstated; even though there appears to be coverage in certain areas, if users actually want to receive it, they might have to climb a hill. While that’s not actually unheard of in Africa – or even in North America – it’s also sub-optimal; you’re not going to generate much traffic on that network and it won’t be a very satisfactory experience for users. Plus it will exclude certain people from being able to use connectivity devices effectively.

vanu rwakahinda

From Vanu’s Uganda map, as an example, in the Rwakahinda region it’s clear that while there’s coverage atop a nearby hill, the densely populated valley receives no coverage. This is a substantial area of 10km x 5km, and with a population density of at least 250 per square kilometre across at least half of this area, that’s a significant number of people left out of the equation in this segment. This tells Vanu or its partners where they need to go to build a site - any area with 25-1,000 people per square kilometre represents a noteworthy financial opportunity, as even if all 50 square kilometres have the lowest population density of around 25 per site, that’s still 1,250 people receiving coverage.

The Long View

While 1,250 people or 25 per square kilometre is the lower end of what Vanu can cover at the moment, we’d like to go lower to deliver more extensive coverage, and this is economically feasible. Getting up to the higher end, with around 1,000 people per square kilometre, would be a bonanza – we don’t expect to find 50,000 people in most of the places we’re building, because if there were that many people living there then MNOs would already have built there. The key is to look at areas where there are probably fewer than 10,000 people in the aggregate coverage but more than 1,000 – that’s the sweet spot for Vanu to look into today, although in the long run we think we can go further than that.

VanuMaps very quickly identify these ideal locations, and it can do it across the scope of an entire continent, drawing us to the places that are most appealing today so we can go country by country and pick off each of these opportunities. The tool is well-suited to identifying where there are coverage gaps, and where we need to go to fill these gaps based on the populations in these areas.

It also has relevance for the flipside of the equation; if a business providing – for example - off-grid power or clean water wants to use mobile payments to realise its business model, then having access to this data provides a road map in terms of showing where and where not to invest in extending services. Vanu is keen to work with companies to help them realise their business objectives using VanuMaps.

We can also help mobile operators substantiate claims of covering a specific percentage of the population, as well as understand where there might be opportunities for them to go. Additionally, we can help them draw lines with a good deal of precision about where they could put sites in order to cover these populations. In addition to mapping the presence of coverage, the tool then allows Vanu to say “let’s pick this site, because we think the propagation will be good from here and we can model it.”

If it does indeed look like it will give us the propagation we hoped for, we can, potentially, tune the site slightly differently or modify its location, but Vanu does have a process for optimising the site locations not only with respect to individual sites but also to a collection of sites. If you want to cover a given area and estimate that it requires five sites, we can designate the desired area and figure out the minimum number of sites to cover the maximum population in the area. By playing with the variables we can automate that whole planning process to provide a handful of optimal sites for achieving the desired coverage. This allows operators to reduce the time spent identifying the sites and minimise the capital and operating expenses associated with providing that coverage by making sure that they’re getting the sites in the right locations based on the population distribution, the terrain, and therefore the propagation to be able to address that population effectively.

New, Critical Information

Another major strength of the tool is its ability to quantify signal strength. In general, the intuition about Africa lagging behind other regions is borne out by our maps – Southeast Asia is definitely better covered. There are certainly opportunities to improve coverage – for example in the north-east of India there are a number of places where coverage is poor, but in the country as a whole – and indeed in Bangladesh – there are wide areas that are well-covered.

Topography is a major factor in the effectiveness of a cell site; one of the key value propositions for Vanu is using small cells to provide coverage because the traditional architecture employed for covering rural areas was a high site blasting out a signal - this approach would lower your cost of coverage per area. However, our key observation is that area doesn’t generate revenue – people do. The industry needs to focus coverage on where the people are and to lower the cost of addressing the coverage you need to be as small as possible.

In an area such as a desert around a city, a high site is not a bad solution as it’s so flat that you’re covering a lot of area and a lot of people, and the terrain is so flat that there’s nothing to block the signal. However, in areas with more mountainous or hilly terrain – or other coverage challenges such as dense foliage or a lot of buildings – then the propagation associated with the high site really doesn’t address the coverage challenge and that’s where moving to a small cell architecture can help. You can put spotlight of coverage where it’s required, and have much lower CAP- and OPEX to boot; however, to do this you really have to know where the coverage gaps are and where the people are, all with a high degree of resolution to be able to implement that model. Otherwise you’ll be putting up small cells in places that don’t need it and there’ll be no economic advantage.

That was the genesis of VanuMaps – the hilly terrain in particular is a key driver in enabling the network architecture to be more effective than the traditional approach of a high site for remote area coverage. Once you get that observation, you’ve got to then have the high-resolution data to be able to be smart about your building plans.

Making the Business Case

In terms of attracting investment into this market, this is a really important tool. Vanu can now quantify – in a way that it couldn’t before – what the market is, where it is, and how we can address it. It’s a key missing element in creating the business model for delivering rural coverage, and we hope this will attract investment into this space, which will hopefully push forward the delivery of coverage in these areas. It eliminates some of the uncertainty over the economic viability of providing connectivity in rural regions.

Given the amount of time it takes to generate these maps – it can be weeks of crunching data – it’s reasonable to ask what happens next time we want to update the map to take into account newly deployed infrastructure. Fortunately, VanuMaps is able to look at changes to the dataset and model just the changes – future updates are significantly more convenient. Since it’s built from the bottom up, we can figure out the changes on that bottom layer, model these changes, and add them to the rest of the pre-existing models to accommodate the evolution of the tool.

Comments powered by CComment