Google and Microsoft have recently show a lot of interest in providing rural access in Africa. The recent “Balloon craze” aside, TVWS experiments are on the way in several countries. The most well known, may be the experiment run by in Tygerberg, Cape Town, South Africa,by the CSIR Meraka Institute(picture) Why would TV White Space be a good idea in Africa, if it seems to be a bad idea in Western Europe? For this purpose let’s get a few things straight. It’s not about the “shortage” of spectrum, but a matter of cost. We have managed to achieve pretty good coverage for mobile broadband services in Scandinavia – a region with population density variations very similar to Southern Africa – basically without the UHF TV spectrum (or only with tiny pieces at 450 MHz). In the short term perspective, there is not very much spectrum needed either in the rural areas, since the population density is low. Spectrum shortage is mainly a dense urban problem.
So why does everyone want to use the UHF spectrum? One reason that has been argued is that the propagation conditions are significantly better at 450 MHz than on 900 MHz or 2GHz. This would mean larger range and fewer base stations to deploy. This is, to some extent true, although somewhat exaggerated by those pointing at the free space formula that seems to indicate that the path loss increases with the square of the frequency. Anyone with elementary insight in propagation, will of course now that that this dependence in fact comes from the size of the isotropic) antennas. So in fact (efficient) antennas get larger inversely proportional to the frequency (making their area grow with the inverse square of the frequency … voila!). This is very important for mobile communication since we need reasonably omnidirectional antennas since we do now know the relative location of the base-station. (This is of course if you use the proper antenna – the current smartphone designs with small built in antennas, would likely ruin most of potential gain.) In rural fixed broadband access, this is of less important since we can build large antennas with increasing directivity since it can be mounted in a fixed position, pointing at the base-station. What, on the other hand matters and provides some advantage, is that diffraction is more prominent and shadowing from smaller obstacles is less pronounced at lower frequencies.
So there is still a cost advantage using UHF spectrum, but why using TV white space – or to be more precise as secondary user on a non-interference basis among TV-broadcasters? In Europe, this is not an attractive solution – there are three reasons.
- There is not enough “White Space” around – where it is actually needed (in dense population areas) – to make it commercially interesting to use a complex, new technology. Yes, it works in Rural areas, but there ordinary cellular systems in the new (licensed) 800MHz band have enough spectrum to do fine.
- Why would I (as operator) invest billions in infrastructure and only have a temporary solution for spectrum. If my infrastructure has a 25 year life-time, I would want spectrum to go with it during the same lifespan.
- The demand for spectrum for Digital broadcasting in the UHF band is rapidly shrinking as more fiber &cable is installed in homes and the viewing habits are changing. There are more and more channels that are viewed by less and less people. In a decade or so, there will be a few channels distributed by DVB-T2, mainly in rural areas, and the rest of the spectrum is repurposed for mobile internet access system (that can handle the remaining channels)
This means that in European perspective, the operators can wait for a second (and third, etc) “digital dividend” that provides exclusive, licensed spectrum for a flexible internet access platform that provides all kinds of services, not just only one, e.g. TV-broadcasting. You might say, this way platform service flexibility is achieved on the network layer, not on the physical layer by allocating spectrum to different services. From a regulator perspective, its hard to see that you can utilize the spectrum more efficiently than that.
So what are the advantages for TV-White Space in the African setting? There are obvious competitors, if you want to connect rural communities, schools, hospitals etc. You could deploy fixed microwave links or a cellular broadband infrastructure, Both the latter are likely to provide far better performance (microwave link much higher data rate, cellular better mobility) than current “WhiteFi” systems tested in South Africa, but are more expensive. On the other hand, you get what you pay for. Will the rural school in the long run be satisfied with a few Mbit/s to share for all? Another issues is market “traction” for end-user equipment. The latter is tremendous for LTE devices, were volumes are in the billions. Will the end user equipment cost for “special” solutions be able to match the mass market for LTE devices? Interesting questions …
You can actually do the math: The well-known Okumura-Hata model for city environments suggests a 6.16dB/decade frequency dependence in field strength (i.e. when leaving the antennas out of the equation), This means for instance that at 450MHz, you have only about a 2 dB advantage compared to 900 MHz in city environments, Looking further at Hata’s corrections for suburban and open terrain, the gain is even less, eventually going to zero in open areas(as you may expect)
Pingback: Money & Power Cost savings Utilizing Cellular Sensor Systems | greenhope.net
As I got it, a main driver for the RSA TVWS initiative was regulative/political. The cellular providers are not extending service in to low economy areas, and they do not allow any third parties to utilize their spectrum. Thus the GSM/3G/LTE spectrum bands are all blocked for these servcies. Without being aware of all details I see another obvious solution to provide local Internet and telephony. Intead of going for an advanced (and interesting) TVWS scheme the following ought to be considered:
– Step 1: Low cost point-to-point backhaul, 2-6 GHz, with local Ethernet/WiFi. (Products are on the market at attractive prices since a few years.)
– Option to extend with local GSM at end point provided that frequencies are made available
– Step 2: LTE for new rural providers in TV-bands, first phase might be ch 60-68.
@Mats: Political or not… if its not profitable to provide these services (say by cellular or microwave backhaul), why should a different frequency band change the game ?
I think you’re all thinking about this technology through telco and service provider eyes. It has a huge potential value by being (almost) licence-exempt. It’s community-sized, simple to install, potentially cheap and requires no permission to operate. Think of the value wrung out of the tiny 898MHz/2.4GHz ISM bands by end-users. A telco mindset will expect to sit on spectrum for years. Maybe it’s just not suitable for _their_ business models.