Comments on report “Options for terrestrial connectivity in sub-Saharan Africa” Notes: 1- See report under the Infrastructure and Connectivity section of the Resources section of the website. 2- These comments have been shared with Bengt at SIDA, Bjorn at KTH, Mike Trucano at Info Dev I have read the reports “Internet for Everyone in African GSM Networks” and “Options for terrestrial connectivity in sub-Saharan Africa”. Let me point out at the beginning that my analysis of the reports separates very distinctly two parts of the telecommunications network- the “access” network that connects directly to the user and the “backbone” or national network that aggregates traffic to and from the users. The usage of these terms in my comments is in line with the description I have given above. Let me say that overall, the reports do present an interesting and feasible scenario for the use of internet over mobile networks. The first report in particular highlights the possibilities that the wide coverage of GSM networks in Africa afford to start connecting many people to the Internet. This is laudable because few people ever consider that they can get connected by GSM. However, connection over GSM should only be a stop-gap, initial (catalytic if you like) measure because it is a narrowband access system. We should not believe that we have solved Africa's connectivity problems because people now have the potential to connect to the Internet with this narrowband technology. Today's and future applications in education, science, technology, commerce, entertainment, medicine, research etc require are increasingly becoming bandwidth intensive. Narrowband communications will not do. What Africa needs is a broadband access system. As I will point out in my detailed comments, broadband access for Africa is increasingly becoming a possibility that can be achieved without spending huge amounts of money with the new and emerging wireless networks coupled with a strategic deployment of fiber networks. Let me get into more details with my comments now. My comments will mainly touch on the later report (Options for terrestrial connectivity) as it relates to the deployment of fiber in Africa. It is a fact that mobile phone or GSM networks with extensive coverage offer a possibility to provide internet access to anybody who has a phone or other enabled device. GSM through GPRS/EDGE enhancements offer a feasible data “access” technology. For Africa, this mobile technology offers a stepping stone for many Africans to get connected. However, we can not and should not get complacent- this technology and related mobile phone technologies can not and should be not be an access technology of first choice for data communications because a) they were designed for voice and voice traffic always takes a precedence over data, b) of their limited data transmission capacity and c) the emergence and increasing availability of equally affordable, easy-todeploy, broadband capacity technologies such as mesh WIFI and WIMAX. In fact, the African mobile phone providers know this- that is why they are all currently studying, trialing or rolling out WIMAX systems. I will come back to this in exploring the various observations and conclusions the report makes. On the national backbone network end, the report on options for terrestrial connectivity explores the existing backbone networks in Africa and makes interesting observations about the appearance of wide spread availability of GSM/ mobile backbone infrastructure. This backbone infrastructure has been built by the mobile operators and is based on microwave radio relay transmission infrastructure. The report then makes a number of observations and conclusions, some of which I find disturbing because they can be misleading and distract people from thinking about strategic choices for African broadband networks. I will explore some of these observations and conclusions by way of making some general comments in the sections that follows. Yes, GPRS and EDGE provide a feasible access technology that can be widely deployed in very short order. However, this technology can not be a backbone technology nor can it be used for anything beyond simple communications services. This is because the data rates for GPRS and EDGE (as the authors acknowledge realistically about 40 kbps and 120 kbps) is very low for anything but basic email and web browsing and these rates don’t even take into account sharing/ contention of the links. Even then, the data will always play second fiddle to voice in a GPRS/EDGE network because the network design prioritizes voice over data. In short, the existing mobile telephony technologies do not have the capabilities to deliver the type of high capacity access at low cost required to stimulate the development of a significant and innovative ICT industry or indeed the kind of capacity demanded of today’s consumers and applications. The mobile phone companies know that Internet over GPRS/EDGE is inadequate for today's needs or applications- that’s why they have grudgingly taken up and are deploying WIMAX (also possibly because they don’t have enough capacity on their GSM networks to provide any meaningful internet access services). WIMAX with mesh WIFI is fast becoming a feasible technology that can be deployed as fast and at a lower cost than GPRS/EDGE while providing true broadband access. Broadband access for Africa should be the aim and norm and not the exception. Why is it that while every home in the developed world moves to broadband access, Africans should be relegated to narrowband communications only? Mobile providers, with their existing investments in towers and power systems, have a competitive advantage when it comes to deploying WiMAX and mesh WIFI. They should and are taking advantage of this competitive edge to provide broadband access. They should be encouraged to do so by African governments. Yes, GPRS and EDGE provide a feasible access option for areas where no other infrastructure exists. However, this can only be a short term solution and one that can not be considered “good enough”. Just like VSAT, these mobile data technologies should be technologies of last resort. BTW, in Namibia, we have explored and proposed the use of the mobile network (using GPRS and EDGE) to provide access to schools that do not have access to other wireless (primary WLAN and CDMA) technologies or have access to ADSL. The report acknowledges that “low-cost and widely available international bandwidth is a prerequisite for ICT services to take off in sub-Saharan Africa”. What it does not acknowledge is that low cost, widely available, highly reliable, nation-wide, high capacity access and backbone connectivity is probably more important for developing ICT services for sub-Saharan Africa. There are numerous sectors and applications that would benefit from nation-wide, low cost, high capacity access and backbone connectivity: education (including distance learning), research, banking, online transactions, intra and inter-company communications and transactions, general e-commerce, local government (e-government) services, agriculture (the backbone of most African economies), and entertainment, arts and culture industry. These sectors and applications for any country would most benefit from access and national backbone connectivity and not (necessarily) from international connectivity! I believe that the need for (more) international capacity is real but it is over hyped at the expense of the need for national/ regional capacity. Imagine the local content, services and applications industry that would be spawned if there was widely available, high speed nation-wide connectivity (backbone and access). In fact, national bandwidth with its attendant local content and applications industry would finally enable Africa and Africans to more equitably participate in the global “information society” or “knowledge economy.” It would enable Africans to become both exporters and importers of data, content and services rather being only importers as the case is today. The national infrastructure required for these developments to take off, grow, expand and to be sustained requires both high speed access technologies as well as super high speed national backbones. Narrowband access technologies like GPRS/EDGE might be a starting point because of their current potential wide availability (given the wide coverage of GSM networks) but will not lead to innovation, creativity and growth in ICT applications, content and services because of their limited capacity. Of even more critical importance to the development of a local, innovative and revolutionary ICT industry in African countries will be the backbone transmission networks- these must be capable of very high capacities and of cost effective upgrade to match the surge in demand once the ICT industry takes off. Only fiber optics can guarantee very high, almost limitless capacity and relatively lower cost upgrades. Microwave radio relays on their own will not do. I find the conclusion that “unless demand from the mobile sector is included, there is insufficient demand to financially justify fiber optic cables” extremely misleading and frankly disturbing. I believe that the ICT industry, services, applications and content (of which mobile telephony while being the dominant piece of the ICT pie at the moment is nevertheless actually a very small portion of the overall potential ICT pie) in Africa have not developed because of lack of infrastructure capable of delivering the required capacities at affordable cost and consistent reliability. It is impossible to measure this demand for ICT services, applications and content in Africa because the enabling/ basic environment- widely available, high capacity, low cost infrastructure coupled with innovative business models- just does not exist. It’s a little bit like the proverbial "chicken and egg" dilemma. It is the same reason that it was impossible to forecast the demand for mobile phone services in the early days! The original mobile telephony demand forecasts were not off the mark- they were totally wrong, several orders of magnitude wrong. This should be a telling and loud lesson to authors of such reports that undertake demand analysis studies for ICTs in Africa. What is worrying is that this conclusion (not financially justifiable because of demand) seems to portray the fact that fiber is not good enough for Africa because “there is no demand”! Why should the rest of the world, and notably Sweden, massively invest in fiber and not Africa? The developed world is rolling out high capacity fiber and replacing all the old microwave radio relay networks with fiber, not because there is a current demand (actually, there is a current fiber bandwidth glut in the developed world but this has not stopped the roll out of more fiber) but because they have their eye on future applications and services. The ICT applications and services that are being developed for tomorrow are extremely bandwidth intensive. What is good for Sweden ought to be good for Africa! Holding back African countries from investing in high capacity networks such as those afforded by fiber optics “because there is no demand” is not only unfair, it is almost negligence. It is this kind of thinking that will only widen the so called “digital divide” between the developed and less developed worlds. The report seems to draw conclusions from a comparison of mobile telephony providers’ and the incumbents’ operations. Yes, the difference in operations, coverage and efficiencies is stark and leaves the mobile phone companies shinning as stars. The report notes that the mobile providers have built their own extensive transmission networks because “their cost of building their own capacity is much lower than to lease transmission lines from the incumbents”. Further, the report notes that the cost of deploying and upgrading microwave transmission facilities is much less than the costs of building fixed line facilities thus mobile operators prefer this to laying fiber. The fact that incumbents have not managed to provide the mobile providers with a transmission backbone, or that the mobile providers have built their own transmission networks does not mean that mobile providers are the answer to Africa’s ICT “gap.” Just as mobile phone technology took the incumbents by surprise and has relegated the incumbents to obscurity, newer and lower cost technologies like WIMAX and mesh WIFI (helped with the convergence to IP) coupled with high capacity fiber backbones are already beginning to make mobile companies seem like the incumbents of old! Actually, governments must act to ensure that these “new incumbents” do not hinder the adoption of new, cheaper and more powerful technologies or the entry into the market of more innovative players. The conclusion that transmission networks built by mobile companies in countries where mobile coverage is around 80% (estimated at about 23% of the African countries from the report) are “geographically adequate” is extremely misleading whichever way it is interpreted. Firstly, the report does not mention anywhere the reliance of mobile phone companies on VSATs for backhaul links. VSATs with their expensive, inefficient relatively low capacity capabilities are completely unsuited as a backbone technology. Secondly, the report does not even consider (as it acknowledges) the capacity or sufficiency of capacity on these backbone networks. With these transmission networks using microwave radio relay technologies, the availability of capacity to handle both data and voice on a truly large scale (multi-gigabit) is questionable. Certainly, the ability of the mobile networks to upgrade cost effectively to handle very high data transfer capacities such as those required for research, education and telemedicine and even more important future innovative and bandwidth-intensive applications is highly questionable. "Future-proof", high capacity backbone networks can only be guaranteed with the deployment of fiber. Thirdly, such conclusions pay, if any, scant attention to the licensing service requirements that sometimes drive such wide coverage say in places like Uganda. The mobile phone company must show that it is providing “national” coverage- it is questionable that in these circumstances, that the transmission capacity is anything to write home about. The mobile phone company is only doing what it is required to do by lawprovide national coverage and charge and will charge high costs to recoup investments made. The issues of “not enough demand” or lack of financial justification for fiber are just a distraction from the strategic considerations that African governments must make in stimulating the development of innovative and creative ICT industries for overall economic development. I have already tackled the issue of demand above. The issue of financial justification is also linked to demand. However, I should add that the deployment of fiber is becoming a much cheaper and affordable option even by any standards. In the first instance, costs of deployment can be lowered substantially (estimated at about 30% and lower of the cost of laying underground fiber) by deploying fiber over existing telephone poles, street lights, electricity poles, gas pipelines. African countries do not have to copy developed countries by deploying underground fiber- they can more easily and cheaply deploy overhead fiber. The development of optical Giga bit Ethernet technologies that do not require as many electronics (repeaters/ amplifiers) over long distance fiber links has also made the deployment and use of fiber relatively cheaper than 5 or say 10 year ago. The report criticizes national governments in Africa that are trying to build their own national fiber backbone networks because they are “often forgetting that communications are in the hands of private companies”. This is shocking. I would have actually thought the report would applaud African governments for finally taking the initiative to undertake some strategic investments. African governments at last understand that unless a widely available, nation-wide, low cost, high capacity network is in place, the development of a local ICT industry and the leverage of technology to benefit education, health, governance and private enterprise will not happen. Fiber offers the only futureproof route. Mobile companies or indeed the private sector can not be relied upon to provide these networks because in their narrow minded push for profit maximization, such networks are not “financially feasible” given the existing “demand.” And even if they did, such networks would be run as monopoly assets which would not help! The report notes that regulatory constraints do not allow mobile service providers to resale their transmission backbone capacity. The report however fails to mention or acknowledge or consider the evidence in countries where the mobile phone companies also hold Second National Operator (SNO) and other “public network” licenses such as in Uganda which allow them to resale capacity on their transmission links. It is not hard to imagine that even if mobile providers were to develop such networks, they would be “closed” to competition- the African mobile provider has become another incumbent. The report also neglects to mention or explore a corresponding “government interference syndrome” (my words!) that is afflicting the developed world (including Sweden) the emergence of publicly funded networks such as municipal wireless and fiber networks in many parts of the developed world. Of course, these networks have drawn the curse of existing mobile companies, incumbents and other ISPs all claiming that governments have no business running telecommunication networks or as the reports puts it, governments should not forget “that communications are in the hands of private companies”. Why should governments in Europe and the rest of the developed world be looking at financing their own telecommunications networks? The authors of the report should be tasked to explain this. In the same vein, why exactly shouldn’t African governments look to finance their own networks? Perhaps the authors might also care to shed more light on this. These distractions aside, African governments can (and in my view should) indeed finance the development of national fiber backbones which would be availed on an open access basis providing a level playing field for all existing and potential service providers and stimulating the emergency of new and innovative services and applications. Financing these networks does not mean that the government should run these networks! Governments can subcontract the running of these entities to the private sector. Even if they did run the networks themselves, government owned and run networks can still be run on a commercial, efficient and profitable basis. As Sweden has ably demonstrated in its utilities and ICT (fiber networks) sectors, governments can successful own and run businesses through a number of approaches including outsourcing, joint ventures, and commercialization. As an aside, actually privatization (leaving things completely in the hand of private companies as the report suggests), without complete deregulation, does not seem to have provided any noticeable benefits for Africa but that is another (long) debate. African governments can also develop new licensing regimes that enable the development of national infrastructure and a level playing field. Most likely, African governments are going to have to do bothundertake targeted investments in infrastructure while revising their licensing regimes- because the development of a reliable high speed national telecommunication backbone is a strategic consideration for overall economic development. In conclusion, what is clearly lacking from these conclusions is strategic long term considerations for Africa. If we believe that this is a knowledge economy or an information society, then surely ICT infrastructure should be treated as a national asset. Forward thinking national governments should be doing everything possible to develop their national ICT infrastructures to a level that stimulates the take off of their national ICT industries. It is a realization that some African countries appear to have made. The development of widely available high capacity fiber optic national backbone networks coupled with high capacity broadband access technologies such as mesh WIFI and WIMAX offer a promising future. The notion that fiber is not good enough (due to low demand) for Africa is the kind of thinking that continuously relegates Africa to the doldrums of underdevelopment. My two cents. Alex