At this year's CW Future of Wireless conference, Simon Saunders of Real Wireless characterised the cost of servicing remote and rural areas. This is quite topical, since the Small Cell Forum is currently developing its next release on that subject. It seems that people who live in rural regions aren't quite as anti-social or equally dispersed as we city-dwellers might think. They often live in groups within villages or hamlets, providing adequate return on investment for a low cost deployment. This favours infill using pockets of coverage rather than increasing capacity through a heavy-duty macrocell approach.
Serving the rural population – Social conscience or business opportunity?
Their recent studies looked at the population density spread across these remote regions, and in particular the economic value generated (ie revenue). Smaller pockets should be more valuable than wide areas and justify investment. It will depend to some extent on how interested and motivated the network operators are – after all, it may be more lucrative to focus on higher value metropolitan areas.
This theme can be found elsewhere. Alphonzo Samuels, CTO of Telekom South Africa, pointed out at BBWF recently that urban populations take up just 2% of landmass, but provide 50% of revenue and 80% of telecom operators' profit. He feels strongly that those in more rural areas shouldn't lose out and wants to find ways to serve them cost effectively.
Calculating economic viability for rural coverage with macrocells
Simon's team have assessed the potential revenue from geographical areas. They found that mobile takeup isn't related to population density, but rather to economic prosperity.
Real Wireless used a cost of $25k to $50k per square kilometre for macrocell coverage, depending on the "ruggedness" of the terrain. Reaching deep into the valleys is much harder than covering an expansive flat plain.
Assuming the population spend up to 10% of their income on mobile service, and half of that is spent on the macrocells, this leads to the population averaging some $5 to $10million of total income per square kilometre.
A study of Wales, which has a lot of high terrain and areas of low population, estimated the cost to increase population coverage from 90% to 99% using macrocells would be £300m (approx. $450 million). Each additional percent of coverage is substantial. The cost per premise covered increases from $15k (at 96%) to $75k (at 99%).
Population density isn't uniform
Where populations are uniformly spread out, then macrocells offer the best option. However, a large proportion of the rural population is clustered into settlements of hamlets, villages and small towns. This is where small cells can be most beneficial and cost effective.
The study calculated that the same investment of $150 million would achieve 93% population coverage rather than 75% using macrocells. Achieving 99% could be reached with just over half the investment for larger cells.
Backhaul was the most sensitive aspect of the cost estimate, with a view that availability of lower cost backhaul could reduce the costs from 50% to 10%.
The study points out that the reducing cost of providing coverage benefits enormous numbers of people. Halving the cost of coverage would bring in 500 million people. It would need a 10x cost reduction to benefit 1 billion. Small cells clearly have a role to play.
The uneven population distribution in sparsely populated areas makes it commercially viable to provide service for even the most remote areas. Backhaul is the critical factor.
Macrocells are good for providing that initial coverage in wider areas, but struggle in rougher terrain and in areas of very low population because the economic returns don't justify it.
Slides for the presentation can be found on the CW website
A video recording of the live presentation can be viewed on IET TV (registration required)