Internet connectivity is the lifeblood of any organisation today, regardless of whether it is transacting on-line or using more traditional services such as e-mail, VoIP and remote access to back-end systems to keep its business processes ticking over.
Even though the situation with regards to telecoms services in South Africa has shown great improvement over the past couple of years, much still needs to be done before telcos can confidently claim to offer the same level of service as the rest of the world. Thankfully, recent developments with regards to telecoms liberalisation and infrastructure have the region well poised to catch up with the rest of the world over the next two to three years. Since communications minister, Ivy Matsepe Casaburri has decided not to appeal the high-court’s ruling in favour of Altech and rapid deregulation of the market, all VANs licensees will soon be allowed to self-provision infrastructure, in turn giving them the ability to compete with incumbent telcos.
Add this to the wealth of backhaul Internet connectivity coming into South Africa through the likes of the undersea cables landing on our shores over the next two years and it is clear the entire market will look vastly different to the way it looks today.
So what awaits us over the next couple of years?
Let us start with what the results from the recent high-court ruling will mean.
It has been clear for some time now that if some of the larger ISP players gained access to a license that granted them the same rights as the incumbent telcos, they would undoubtedly respond by trying to become telcos themselves, albeit on a much smaller, regional scale.
With no legacy infrastructure of which to speak, that means a greenfields approach to technology rollout – much like the scenario Neotel found itself in a couple of years ago.
As we saw with Neotel’s decision to roll a fibre backbone out across the country and fulfill on much of their last mile using wireless technologies, there’s plenty of room for doing things differently.
The only thing is, it is a couple of years on from when Neotel made a decision on which technologies to use in building out their network. So the chances are good that the ‘new kids on the telco block’ will choose a route that is different still to the one taken by the second network operator.
Plenty of options exist and even the route taken by Neotel could be viable. By and large however, it seems like most players are backing the WiMAX horse.
Going with WiMAX
Noel Kirkaldy, marketing director Wireless Broadband, Middle East, Africa and Pakistan at Motorola Home & Networks Mobility says that WiMAX (802.16) is an intriguing and relevant technology for new telco entrants.“The problem is however that WiMAX have not yet seen a true regional implementation as yet,” he says.
Kirkaldy says that WiMAX is however an important technology because of how it will differentiate the newer players in the market.
“The cellular players have driven their business models with a predominantly voice focus,” he says.
“But WiMAX targets data first and then through using VoIP can handle voice traffic. It is a smart play,” he adds.
Kirkaldy also believes that WiMAX is a good fit for ECNS players since the ISPs that are likely to get these kinds of licenses understand the marketing and selling of data services better than the cellular players do.
“We cannot however discount the cellular players however. They are already acquiring the expertise required to be strong contenders here,” he says.
Kirkaldy says that WiMAX also allows new entrants to start small and grow from there. “They do not necessarily have to build out a large nationwide network at the outset. By using WiMAX they could start by offering a limited, regional service, and expand that offering as and when cash flow and the business case allow.
“Importantly, it is also a fixed nomadic solution,” he says, “so it will allow for mobility in the future. We see it as a strong alternative to DSL and as we know, that is somewhat of a hot button today.”
Despite the conviction with which Kirkaldy puts forward the point about WiMAX’s viability, it remains to be seen whether it makes good sense in practise.
In fact, Dr. Alapan Arnab, innovation manager at T-Systems South Africa says that the high costs associated with building a WiMAX network coupled with the low average revenue per user (ARPU) in underserviced areas and the inherently exorbitant costs of providing Internet services in South Africa have raised questions around its viability.
A different perspective
“In particular,” says Arnab, “while WiMAX may make economic sense in areas with high population density, the economic fundamentals are more challenging in areas with low population density, such as rural areas.
“The business case for low population density locales only exists with cross subsidisation (either from the operator’s business case in other locales or from other sources such as the government).
“In T-Systems’ opinion then, cheaper, alternate technologies are more suitable for areas with low population density and WiMAX should be seen as one of a range of technologies for developing countries,” Arnab explains.
Tackling the first set of challenges for WiMAX, Arnab says telcos can conservatively expect to pay in the region of R75m for a 50 base station WiMAX Network.
“They will furthermore have to contend with maintenance and operational expenses of about R35m per annum, for this 50 base station network,” he says.
“These costs are in line with reality, since covering Gauteng for example, would take a minimum of 50 base stations spread out across the three major municipalities in the province – Ekurhuleni, the City of Johannesburg and the City of Tshwane.
When it comes to providing Internet services, Arnab says it is safe to assume the operator would use a 1:50 contention ratio for Internet access and provision roughly 1 Mbps (contended) to each user.
“Add in all of the overheads, such as call centres, marketing, management or the cost of running and maintaining Internet gateways, e-mail servers and the likes, and the cost to the consumer will end at the R999 per user mark,” he says.
Unfortunately, notes Arnab with the price of R1000 per user, it is probable that the maximum potential market for WiMAX will be much lower than many players originally anticipated – it will quite simply be too expensive to appeal to the mass market.
“And this will naturally have a significant impact on the economics involved with providing such a service,” he says.
Not the ultimate solution
While WiMAX can be more affordable than current broadband offerings in the South African market (nobody besides Neotel currently offers an uncapped broadband service for less that R1000 per month), the price is still not significantly low enough to allow for large-scale broadband uptake.
“And we’ve based this example on Gauteng – the largest and most prosperous metropolitan area in the country,” he says.
“When this financial model is applied to areas with significantly lower population density (such as currently underserviced areas), the financial business case for WiMax just does not make sense,” he says.
“We therefore do not foresee a financial success for regional network operators in deploying WiMax as the technology for primary network infrastructure.
“Alternative technologies will be required to provide more universal, affordable services – specifically in less dense, rural areas of South Africa,” Arnab says.
This is possibly one of the reasons none of the players with WiMAX licenses have been prepared to bet the farm on this single technology.
Even Motorola’s Kirkaldy concedes that we have not yet seen an operator use WiMAX as its sole technology yet.
“While a number of parties have acquired WiMAX spectrum, all of them have other interests,” he says.
“That said though, I do believe that that new entrants stemming from the ECNS process, will make use of WiMAX for their nationwide and regional networks.
“And since Motorola already supports 802.16e (the fixed-wireless deployment of WiMAX), the ecosystem to roll networks out and support them is very much in place at this stage,” he says.
So, while opinions continue to vary, one cannot however deny that it will in some way mean or form be a part of the networks we are likely to see in the market over the coming years.
Activities in the communications market will certainly not be limited to the moves of the new telco players, however.
As time marches on, so demands change and the technologies being used to deliver communications services begin to age – ask Telkom, the one company that it seems will be playing catch-up over the next couple of years.
In a nutshell, the natural course of things means that telcos, especially our mobile telcos, need to stay on the crest of what their customers are demanding – and more and more it looks like demands centre on faster, cheaper, more ubiquitous data.
And those demands looks like a specification straight out of the LTE (long-term evolution) playbook.
The promise of LTE
LTE is the next logical step for networks premised on the 3GPP (GPRS to HSDPA) or 3GPP2 (EVDO to UMTS) networks being used both locally and abroad; and the reasons for its viability are far reaching.
Kutty Kanagaratnam, director of Radio Access Sales for Ericsson in South Africa says for starters, the skills required to rollout an LTE network can easily be built on top of existing skill in the market.
A second benefit is its ability to converge the different technologies being used by operators in various regions around the world, namely GSM and CDMA – here in South Africa, the cellular operators use GSM, while Neotel is delivering its last mile over CDMA.
Kanagaratnam says that GSM is by far the bigger standard, with about 88% of the market share. CDMA on the other hand is much smaller with 10% of the worldwide market share of wireless networks.
“It makes great sense because it dovetails with the worldwide move towards IP as the defacto protocol of choice – after all, voice and other services can easily be carried by IP networks,” he says.
Another reason it makes sense, he says it because it only entails an access layer or last mile change in the network – the core backbone operators have built will stay intact.
By comparison to 3G Kanagaratnam says LTE has the ability to cater for a greater number of users, while delivering better performance – it delivers more bandwidth in a similar spectrum range, for a much lower price.
“In the initial stages, LTE will be capable of data rates in the 100 Mbps downlink and 50 Mbps uplink realm. Over time however, that will evolve into services capable of 300 Mbps downlink and 100 Mbps uplink,” he says.
And unlike the claims for other wireless technologies, which soon peter out into insignificance the moment the user base scales, Kanagaratnam says it is plausible that all of the users on a single cell could get that level of performance.
Too soon for South Africa?
To date, he says however that there have a been few discussions with local mobile telcos, but that it does not look like LTE will roll through in South Africa for a good couple of years yet.
“There is way too much potential left in HSPA,” he says.
A little known fact is that by next year, HSPA networks will be capable of delivering downstream data rates of 42 Mbps and upstream data rates of between 11 and 12 Mbps.
“This will be a quick win for the operators and therefore, something I see preceding the move to LTE in the local market,” he says.
The international markets could be an entirely different story however.
Because of the pressure being placed by WiMAX on incumbent players with GSM and EVDO/UMTS networks, Motorola’s Kirkaldy says timelines for LTE have been shifted from 2012/2015 to 2009/2010.
“The first commercial LTE solutions will be commercially available during the course of 2009,” he says. Kanagaratnam too says that target is realistic.
While Ericsson and other vendors are still in the development phase of their products based on this technology, he confirms that his company’s products should be ready during the second or third quarter of next year.
One has to ask whether the kind of data rates offered by LTE are needed today – sure, it is great to have an excess of capability in reserve, but telcos prefer to cater for market demands, rather than creating a glut. There is no doubt that LTE has a role to play, but whether or not that role is as close as two to three years away remains to be seen. WiMAX however is a different story – the demand is there and the technology is capable of catering for it. Whether or not an entirely WiMAX network is plausible is questionable. All things considered, it will be interesting to see how the technology being used pans out over the coming years.