From this page, you can estimate the spending needed to make public access limited capacity broadband service available to all populated areas within a country, based on parameters of your own choosing. The model is based on a 10-year horizon: notionally 2006 to 2015.
This model uses a spatial approach to classify the areas of a country according to whether or not they are commercial viable and then estimate the cost of building infrastructure to serve these areas. The model assumes that broadband service will be provided using wireless technology: CDMA2000 1x EV-DO in the 450 megahertz (MHz) band for rural areas, and fixed worldwide interoperability for microwave access (WiMax) (rev 802.16-2004) in the 3.5 gigahertz (GHz) band for urban areas. While the two technologies do not provide identical service-feature sets in terms of speed or level of mobility, they both meet the minimum International Telecommunication Union (ITU) definition of broadband (at least 256 kilobytes per second, Kbps) and are commercially available. The model does not assess the extent of existing broadband coverage, because the overall level of access is still very low.
The model distinguishes between:
Market efficiency gap: the portion of the market that can be considered commercially viable, once any remaining regulatory barriers have been removed.
Sustainable coverage gap: the portion of the market that could be operated on a commercially viable basis, once the necessary investments had been funded by public subsidy.
Universal coverage gap: the portion of the market that could never be operated on a commercially viable basis, but would require public subsidy both for investment and operation.
Commercial viability is determined based on the balance of costs and revenues in each pixel of land. The revenues associated with any particular area are estimated based on population density, income distribution, and the assumed budget share going to broadband services. The costs of serving any particular area are estimated based on the steepness of terrain and the size of the wireless cell site needed to serve the estimated demand, which in turn determine the number of based stations that need to be built. Only the costs of providing ‘outdoor’ broadband connectivity are considered, excluding those associated with local retailing of broadband services (via phone shops and Internet cafés) and providing the computer terminals needed to access the internet.