Stuck at the starting gate

The global mobile broadband market is currently a battlefield between two competing technologies: LTE and WiMax. But in SA, the question is moot.

2 December 2009
MWeb Business' Roman Hogh believes there is minimal difference between LTE and WiMax at a technological level.

The nature of the ICT industry is such that vendors have often pitted their technologies and/or marketing strategies against one another in a bid to sway customers toward their offerings and away from those of their competitors. Such appears to be the case with the battle for dominance of the mobile broadband market.

The two protagonists are LTE and WiMax. LTE, or Long Term Evolution, is a wireless broadband technology developed by means of a collaboration between groups of telecommunications associations called the Third Generation Partnership Project (3GPP). The 3GPP is essentially a group that represents the interests of mobile operators from around the planet and, importantly, the equipment vendors they depend upon.

WiMax, on the other hand, is a wireless digital communications system intended for wireless ‘metropolitan area networks’ and is also known by the informative but rather clumsy moniker: IEEE 802.16. This is informative because it tells us it has evolved from WiFi (802.11) and is, therefore, a networking standard backed by the IEEE.


Implicitly, this means that WiMax is an ideal mechanism for the transmission and reception of TCP/IP packets. Ironically, so is LTE; buried in the detail of its definition on the 3GPP website is: “... optimising it for packet mode and in particular for the IPMultimedia Subsystem...”.

Translation: TCP/IP.

Roman Hogh, head of technology and product strategy at MWeb Business, confirms this: “WiMax is about pure IP. There is nothing else that WiMax does but transport IP.” And later, in the same interview, he adds: “LTE is also pure IP.”

Noel Kirkaldy, director: solutions marketing at Motorola, agrees: “Both technologies are based around the next generation technology known as OFDMA [Orthogonal Frequency- Division Multiple Access] … that will allow higher data rates, better latency and, critically, quality of service.”

Hogh adds that the differences are not really technical. “If you look at a technological level, the technology used on 802.16e mobile WiMax and what LTE will become is not the same but it’s [very] close.”

Richard Hurst, telecoms analyst at Pasco Risk, is more succinct when asked if there are clear advantages in one technology over the other: “No,” he says. However, he does add that existing GSM operators would be more likely to opt for the LTE path to a 4G solution.

Seamless hand-over

So, what’s the difference? According to Aingharan Kanagaratnam, VP of network solutions at Ericsson, many suspect claims are made about WiMax, like that it’s easier to deploy and offers greater coverage. But Ericsson disputes this, he says, and believes increased network load would reveal the error of those claims.

Another difference between LTE and WiMax, he adds, is that LTE supports seamless handover of a connection to 3G and GSM and back again. This allows existing operators to provide a seamless service to subscribers using or reusing the network, benefiting from the economies of scale but also following the technology track.

Angus Hay, CTO at Neotel, confirms Kanagaratnam’s assertion but notes that the lack of seamless hand-over is not, strictly speaking, a technical limitation of WiMax.

And he has a point because, as Hogh points out, GSM operators typically run multiple networks already. “If you’re a mobile operator who has to provide backward compatibility, you are going to run three separate networks ... because the GSM network is a whole separate bunch of radios and [antennas]… using the same back-haul links...”

All of which takes us back to Hay’s comment. “It’s possible to build handsets that can accomplish the hand-over between a GSM network and a WiMax one but none of the big mobile handset manufacturers is doing it,” he says.

There is another important difference between LTE and WiMax: cost of deployment. Even Kanagaratnam concedes this when he says: “The WiMax spectrum is better regulated and a bit cheaper than typical 3G and GSM spectrum.”

Hogh is more specific: “... if you go to Ericsson and buy an HSPA base station, you’re paying in the R5-million region...; when WiMax came out it was about R700 000 [per base station] but already it’s much cheaper [at around] R100 000 because it is so commoditised and available from so many vendors.”

Hurst confirms this and raises another point: “WiMax can be cheaper but it’s more suited to ISPs [and] new entrants to the market.”

Kirkaldy agrees, describing a market split into two camps, broadly: existing cellular operators, and wireless internet service providers. The former, he maintains, are entrenched in the GSM family of technologies while the latter is more comfortable with IEEE standards and technology.

Hogh puts it differently: “The classic voice provider going into data and the classic data provider going into voice or a voice-type service. But anybody whose roots are in IP, something like WiMax, is such a natural fit, it’s just a kind of wireless ADSL technology.

So if you’re providing mostly data and a bit of voice, which is a typical ISP model, the technology just fits in so much nicer – there’s no legacy baggage.”

On the other hand, he says, there is nothing stopping existing cellular type providers from choosing WiMax over LTE to complement their existing GSM and UMTS networks. But the vendors such operators have historically dealt with – Ericsson, Nokia, etc – will design their LTE equipment to fit exactly with the existing kit. “So you’re pretty much caught in that upgrade path. They’re going to lock you in with OSF and GSF management systems,” he adds.

Still, WiMax currently has a lead in the race globally. As Kirkaldy points out, there are already some 500 WiMax implementations dotted around the world but, as yet, commercial LTE hasn’t left the starting block. Locally, the question is moot because Icasa has yet to allocate spectrum to players other than Telkom and Sentech.