WiMAX establishes a clear foothold in Africa with over 100 installations across the continent

Top Story

The number of potential wireless data applications is bewildering but the big battle appears to be between WiMAX and a technology that doesn’t yet exist, LTE. But in Africa, there are over 100 installed WiMAX applications and all of the big players have either got it working for them or will do over the next year. Russell Southwood investigates what it is WiMAX can and can’t do.

One of the largest WiMAX equipment vendors in Africa, Alvarion, estimates that there are just over 100 installed WiMAX systems in Africa; over 45 of these are operating in unlicensed spectrum and over 55 in 2.X or 3.X spectrum. Increasingly new systems are being licensed in the 2.X spectrum. Alvarion claims to have more 50% of the market and out of the 40 key major operators (many with Pan-African operations), it claims to have sold to 26 of them. Other vendors who have sold in the market include: Telsima, ZTE, Huawei, Soma Networks, Smart Link Communication, Galaxia Telecom, Redline Communications and Airspan Networks.

The 2.X spectrum has increasingly been opened by regulators in part under pressure from satellite operators who want WiMAX to move from that spectrum band to avoid claimed interference. However, the move is causing uncertainty in some countries. For example, Nigeria’s NCC has announced it will make the transition but there is currently no clear message on how it will do it or how it will to compensate existing investors. Potential users see a dividend in 2.X as it will offer better coverage.

A factor holding back all wireless technologies in some countries is the lack of a defined and enforced frequency allocation policy. Even some of the best of African regulators have little capacity to police how frequency is used as the chaos over wi-fi in some cities illustrates. In some countries like the DRC several operators have been given the same frequencies. So there is a direct relationship between transparency on this issue and the desire for companies to invest in technologies like WiMAX.

So who is making the transition to WiMAX and why? Africa’s fixed operators in particular seem to have been persuaded to sell the low speed versions of CDMA 2000 for both fixed voice and data. And as we described in issue 411, this seems to have had a good run for its money as a fixed line substitute product.

Many are planning to offer CDMA 2000 Rev A which has achievable speeds on data of 300-700 kbps per second and can be operated in the 800 mghz spectrum in SADC countries. But one operator in a large market is offering Rev A to household customers and WiMAX to corporates, particularly offering hot zones in business centres. But logically if household customers want faster speeds in future then will it be WiMAX that’s used to deliver it? Possibly not, for as this operators told us:”The limitation of WiMAX is that it’s just not 4G.” This comment captures the almost theological struggle over what performance the different technologies will deliver, both now and in the future.

What no-one is telling their CDMA 2000 customers is that when everyone else gets cheaper and faster household data connections, as they will do when the new international fibre arrives mid-year 2009, they will have to throw away the widely sold CDMA 2000 phones and buy another piece of kit if they want to upgrade.

For Africa’s mobile operators, the installation of WiMAX has had a double motivation. As mobile data prices have been lowered, the use of the service has shot up with tens of thousands of customers even in mid-scale markets. If someone in a cell is downloading a large file, it quickly impacts on voice customers. Likewise, if there are a lot of data customers in the cell. So the mobile operators need to have two parallel networks if they are to preserve the integrity and quality of their voice services. For as David Levy, General Manager, Africa and the Middle East at Alvarion says:” You need to use something like WiMAX because it’s easier to add services on top. HSDPA is not a long-term usage, mobile-based IP technology.”

However, as the larger mobile operators like MTN go through a process of becoming vertically integrated operators, they are keen to build their knowledge internally about how to operate fixed wireless services. In the case of MTN this education process extends to having different vendors in different territories to get a clear picture of actual performance.

There’s been a long symbiosis between WiMAX in its formative years and Wi-Fi. However, the reality in many African cities is that there are high levels of interference in services delivered in the unlicensed spectrum. Also like the mobile operators’ data services, if you have one heavy download customer or too many customers on a base station, it gets to be relatively slow.

Wi-Fi has been the boon for travellers in Africa that has replaced the previously excruciatingly slow dial-up services. When it works well, you can almost believe that it’s just what you’ve always wanted. This seeming bandwidth paradise tempts those from developed countries to treat it as if they could make heavy downloads like they do back home. The result is often gridlock in speed terms.

In some ways these issues are as much to do with how the technology is provisioned as the intrinsic limitations of the technology. Africa’s chronically high user contention rates are at least half the answer to the question of why it gets so slow. But as Alvarion’s Levy points out:” Wi-fi’s a fantastic way to extend data services into rural areas. There’s practically no CPE cost and little interference”.

But this is where it gets interesting because Namibian incumbent Telecom Namibia (also a CDMA 2000 user) has flipped the proposition on its head and is using WiMAX to roll-out fixed wireless voice to a farming area with a dispersed population of between 3-5,000 people. But now it is running a campaign to persuade them to also use data. Another major African mobile operator has a household CPE product that will support fixed wireless VoIP and is offering that wherever VoIP is legal.

But although WiMAX’s 16e support’s mobile voice, it has been very coy about if and how this will be used. Alvarion’s Levy is clear that it currently makes no sense for existing operators:”If the operator has no access to a GSM or CDMA licence, then it might make sense. For others, the price of the CPE needs to be lower to be attractive.”

WiMAX is also not really geared up for multimedia streaming or very large downloads of the kind those wishing to deploy Triple Play might need. Again Levy is clear on where the boundaries lie:”You can achieve 10-30 mbps per sector on WiMAX. You’d need a minimum of 500 kbps per second or more per customer. (To achieve that speed), you’d need only 35 customers per sector. Streaming video is not there yet. The compression rates are not there yet. On straight ARPUs, it doesn’t make sense. But we are exploring business models using advertising.” However, there is a planned speed increase that will match LTE by the time that technology is available.

So what of the future and the big arm-wrestle with LTE? The supporters of LTE are spinning a confusing number of different stories. Firstly, the technology will be up and ready in eighteen months to two years. Secondly, all existing technologies will merge towards LTE. Thirdly, WiMAX will become last year’s technology very shortly so don’t make any down payments on it.

To an outsider like myself who is not committed to any particularly technology but only that which works in the African context, these lines or argument are about spreading FUD (Fear, Uncertainty and Doubt). Anyone who has watched the lengthy and very frustrating discussions between vendors and other parties on WiMAX knows that nothing springs straight from the research lab into the showroom. By the time LTE arrives, WiMAX will already have a 3-6 year track record as a working technology. All technologies may well merge towards LTE but you first need to know what LTE actually is as a set of standards.

LTE’s promoters can claim with some justification that they have learnt these lessons from the WiMAX experience but that does not help them overcome the considerable divergent interests: this is always the circle that has to be squared. And in the African context, the vertically integrated new incumbents will know from what happens operationally that their data networks will need to work seamlessly if they are to keep a hold on these new ARPU’s.

As Alvarion’s Levy puts it:”If you want to keep your ARPUs, you need to install now and not wait for tomorrow. It’s pure time to market. If you’re 99.9% in voice, don’t get into WiMAX. But Africa has a hunger for connectivity which WiMAX meets and with the arrival of the new international cables in 2009 that hunger will only increase.”

So WiMAX investment in Africa can be good for fixed data and voice. Urban areas like Douala can be covered with about 10-15 base stations. The average equipment cost per base station is US$50,000 plus or minus 10%. It can also deliver fixed data and voice in rural areas as the Telecom Namibia example demonstrates. And with a more open attitude to VoIP by more countries, these could all be low-cost IP calls thus helping to expand the rural voice market.

Current CPE costs are an obstacle that will either mean operators offer a partially or fully subsidised product. A PCMCIA card currently costs between US$80-100. An indoors unit costs between US$200-220 whereas an outdoor rooftop unit costs US$300. Nevertheless Alvarion’s Levy believes prices are set to fall:”In 2009, equipment will become standardised for 16e and with higher runs, costs will begin to come down”.

Further afield, all eyes are on Clearwire’s network in Baltimore. Sprint completed its takeover of the company at the beginning of this month. It represents the first real attempt to put a network together that operates seamlessly at all levels. Information Week’s J.Nicholas Hoover reports getting download speeds of 3.4 mbps and upload speeds of 1.2 mbps. Nokia has released an N810 Tablet WiMAX edition specially for Clearwire which sold through very quickly. Cost? A cool US$450 or thereabouts.

Wi-MAX and Wi-FI enabled laptops have started to go on sale and are being used by the network’s customers. Numbers will begin to build from next year. For the majority of us who will be without a Wi-MAX-enabled chip, there will a US$20-30 dongle. Not surprisingly, Clearwire’s CEO Ben Wolff has the same line on LTE as Alvarion’s Levy:”Clearwire chose WiMAX because it may be four years before LTE is ready for commercial use.”

So it will be interesting to see when the first LTE products are announced, when they start being sold into Africa and what bar they will set for other new wireless products.