Time on the side of WiMAX

WiMAX has a clear time-to-market advantage over 4G rival LTE

WiMAX has a clear time-to-market advantage over 4G rival LTE

WiMAX has a clear time-to-market advantage over 4G rival LTE.

A US senator once famously remarked that the first casualty of war is truth. From the perspective of the WiMAX camp, its ‘battle’ with LTE to win the hearts and minds of mobile operators has illustrated all too clearly the validity of that statement.

“There is a huge amount of misinformation coming from the LTE side, not least in the name itself,” says Siavash Alamouti, CTO of Intel’s Mobile Wireless Group. “LTE is not an evolution.”

Supporters of LTE, or Long Term Evolution, typically claim that it is a natural progression from 3GPP (GSM, EDGE, W-CDMA, HSPA) and 3GPP2 (CDMA 1x, EV-DO) networks. LTE, they say, is part of the 3GPP ‘family of standards’. As such, call handover is less challenging to achieve between LTE and ‘legacy’ cellular networks than it is with WiMAX, whose origins are to be found in the IEEE standards body.

In turn-so this argument goes-multimode LTE/HSPA and LTE/EV-DO devices are less dif-ficult (and cheaper) to develop than WiMAX/HSPA devices. And since multimode devices are important for operators who place a premium on providing wide area coverage via single ‘4G’ devices from the word go-LTE or WiMAX will inevitably be rolled out in stages, so the first devices will need recourse to 3G if they are to retain connectivity beyond the initial rollout areas-this could be a persuasive argument to take the 3GPP path to 4G.

Th is type of thinking has certainly struck a chord with AT&T. A spokesperson for the US wireless giant, on explaining why it opted for LTE as its preferred 4G option, talks about a ‘graceful transition’ from 3G technologies.

But Intel’s Alamouti argues the LTE camp is being disingenuous about its evolutionary credentials. Like WiMAX, LTE is an overlay network based on OFDM (orthogonal frequency division multiplexing) technology. Both WiMAX and LTE, he says, represent a step change from a legacy circuit-switched 3G network to an all-IP ‘fl at’ network; and both will require spectrum not in active use by legacy base stations. Mobile operators who have access or potential access (via licence awards) to spectrum in any of the standardised WiMAX frequency bands (2.3GHz, 2.5GHz and 3.5GHz) are the main addressable markets for WIMAX among the cellular community.

“From the network core to the base stations, along with the need to deploy femto and pico cells, LTE and WiMAX both represent a complete revolution in the network,” continues Alamouti. “LTE proponents are trying to portray something that is not true.”

Alamouti dismisses the view that call handover between LTE and cellular networks is somehow technologically easier to establish than it is between WiMAX and 3GPP standards. “If it is easier, then a lot of that is down to the resistance of the 3GPP community to fully embrace non-3GPP interworking,” he says. “From a technology perspective there’s no barrier and not much work needs to be done in order to provide tighter handover between WiMAX and legacy systems.”

Richard Keith, senior director of broadband access solutions at Motorola-which backs WiMAX and LTE-confi rms that much progress has already been made on interworking between 3GPP and non-3GPP standards, which would appear to auger well for WiMAX interworking with 3GPP networks.

“Motorola and others have demonstrated handover between LTE and non-3GPP technologies,” says Keith. “And although we believe this type of handover to be even more challenging than handover within the 3GPP family of standards, once the handover task is implemented it becomes irrelevant to say how hard or easy it is.” And WiMAX supporters can point to numerous multimode devices (WiMAX/wifi /Bluetooth, WiMAX/EV-DO) that are already commercially available. In Korea, KT’s WiBro service off ers triple-mode PDAs (WiBro, EV-DO, T-DMB) from both Samsung and LG. At the time of going to press, Sprint was planning to unveil multimode EV-DO/WiMAX devices to coincide with the Xohm launch.

“The crucial thing missing for the LTE community is the handset and this is where I see WiMAX as having a window of opportunity,” adds Rob Westwick, a consultant in the wireless technology division at the PA Consulting Group. “WiMAX handsets are now coming onto the market.”

Westwick does add, though, that there will need to be a large number of WiMAX devices, at diff erent price points, for consumers to choose from. Otherwise, he says, WiMAX will still struggle to take off . “Th e lack of success for BT’s Fusion service [wifi /cellular] has oft en been attributed to its small range of handsets,” he warns.

A question of standards

WiMAX supporters are generally disbelieving that the high performance levels claimed on behalf of LTE-in lab-like conditions-can be sustained in a real network environment. Ericsson announced earlier this year that a downlink rate of 154Mbps had been achieved up to 1km from the base station-with line of sight-on a 20MHz channel, in the 2.6MHz band, using a 2×2 MIMO confi guration. Th e Swedish vendor-which eschews WiMAX in favour of LTE-says it will be able to deliver up to 200Mbps to individual users via LTE next year.Speaking at the WiMAX Forum Global Congress in Amsterdam (June 2008), Barry West, president of Xohm, Sprint’s Mobile WiMAX business unit-and designated president of the new Clearwire joint venture-dismissed the claims made by LTE supporters as ‘PowerPoint propaganda’. “WiMAX or 3G LTE, in the absence of a large swathe of spectrum, cannot perform much better than current 3G technologies,” he said. “Th ere is no magic about LTE.”

What WiMAX has got that LTE doesn’t, added West, is a developed chipset ecosystem (over 20 suppliers), which includes the mighty Intel. By seeding the market with low cost chipsets, Intel (and others) have established a clear time-to-market advantage over LTE that could stretch to three or four years, argued West, if looked at from the two technologies’ respective eco-systems. “LTE could wither without multiple chipset vendors,” the Xohm president warned.

And while the 3GPP community is hoping to rubber stamp Release 8 by the end of this year, which defines the LTE air interface and would allow, theoretically, a downlink maximum speed of 300Mbps per 20MHz of spectrum, the standardisation deadline could be a tough one to meet.

“We have tracked the LTE standardisation process very closely over the last 12 months and we think the implementation of LTE equipment is still a long way to go,” says PA Consulting’s Westwick. “I think the LTE camp will reach a standardisation milestone by the end of this year of sorts, but whether it is a milestone that allows true interoperability is another matter. Change requests, in all likelihood, will still need to go into the standard aft er the end of this year. WiMAX is certainly ahead of the game compared to LTE on standardisation.”

Th e WiMAX Forum established a Mobile WiMAX standard as far back as December 2005 but it took over two years of trials and testing before mobile product certifi cation began. “And we are an aggressive industry in terms of getting things done,” says Dr. Mohammad Shakouri, WiMAX Forum board member and vice president of marketing. “WiMAX is happening now and billions of dollars have been invested in it. How can you compare it with something that doesn’t exist?”

Dr. Shakouri argues that many of the claims made on behalf of LTE are “scare tactics” deployed by people, with vested interests, who say that mobile operators should not stray from the 3GPP path. He adds: “WiMAX has done a ground-up design for mobile broadband data. Everything WiMAX has been working on for the last two years, such as OFDM, MIMO and wide channels [10MHz], has been adapted by the LTE community. That’s a major compliment to us.”

And the time-to-market advantage that WiMAX has over LTE could prove key if mobile operators start to run out of capacity soon and need more spectrum. It’s a scenario that Dr. Shakouri envisages and he unequivocally rejects the claim by Ericsson’s director of LTE products, Thomas Norén, that HSPA is a “viable alternative” to Mobile WiMAX until LTE arrives, particularly in markets where there has already been a high take-up of mobile data services. “I think HSPA is excellent for WiMAX because it shows the limitation of 3G bandwidth channels,” he says. “What we need today is much higher capacity, greater density, wider channels and multimode devices. WiMAX is not going after HSPA business.”

Motorola’s Keith concurs with Dr. Shakouri that many HSPA operators will soon face a capacity crunch due to a mixture of fl at-rate tariffs, easy-to-use USB HSPA dongles on laptops, and better multimedia handheld devices driving up mobile data traffic. “The yearly data growth on HSPA networks in 2007 was ranging between six and 15 times increase, and a few operators are starting to report predicted growth as high as 20-50 times for 2008,” he says. “As a consequence it is expected that many operators will reach capacity and exhaust their current HSPA spectrum holding between 2009 and 2012.”

Big name backers for LTE

Despite the strong arguments put forward by the WiMAX community, LTE looks certain to be the preferred choice for the majority of mobile network operators (MNOs) around the world. “LTE is expected to be the technology of choice for most existing 3GPP and 3GPP2 cellular operators looking to migrate to a next generation network,” says Keith. “WiMAX proves to be an attractive choice for new entrants, and also established firms seeking to differentiate a wireless broadband service.” Dr Shakouri appears similarly pragmatic.

“It’s most likely we will see two different industries forming,” he says. “One from the 3G community, with LTE plans, and the other from the WiMAX Forum.” One important question remains is how far WiMAX can invade cellular ‘territory’? From recent announcements it would seem unlikely that WiMAX will make much headway with most tier one mobile operators. US wireless heavyweights, AT&T and Verizon, nailed their 4G colours to the LTE mast last year. During the MWC (Mobile World Congress) event held in Barcelona (February 2008), China Mobile, Verizon and Vodafone also announced they would be coordinating their LTE trials. And NTT DoCoMo, Japan’s largest mobile operator in terms of subscribers, reinforced its 3GPP commitment by declaring at MWC it had selected Ericsson for its LTE base station development project.

“The big mobile operators today have said they are going to look at LTE for the long run but I still think there is a window of opportunity for WiMAX to be part of the network,” adds Dr Shakouri. “We don’t see that one will overtake the other. Both will have a market opportunity.” There are a number of reasons why many cellular operators will support LTE. For a start, Ericsson itself, as the world’s largest mobile infrastructure supplier, will put forcibly the 3GPP case for 4G to its numerous MNO customers around the world. (As of March 2008, Ericsson had supplied 90 out of the 185 launched HSPA networks worldwide.)

LTE can also be deployed in both new and existing FDD spectrum, and it is FDD spectrum that most cellular operators have got. By contrast, the strength of WiMAX lies in TDD, although the WiMAX Forum is currently developing an FDD profile. In June 2008 the WiMAX Forum submitted a proposal to an ITU Working Party to enhance the performance of the 802.16e interface, which includes an FDD option.

But in the on-going PR battle between WiMAX and LTE, the LTE camp arguably claimed its biggest victory in June 2008 when the Next Generation Mobile Networks (NGMN) Alliance, whose members include the world’s biggest mobile operators, announced it had selected LTE/SAE as its preferred choice for next-generation mobile broadband technology.

After nearly a year evaluating Mobile WiMAX, LTE and Qualcomm’s UMB system, only LTE was deemed to have measured up to the alliance’s selection criteria, which were based on four parameters: performance, functionality, interworking and time-to-market.

The range of performance criteria included maximum bandwidth, downlink and uplink rates, latency, support for VoIP and spectral efficiency. Candidate technologies were required to be able to meet the criteria set by the NGMN Alliance in time for a 2010 target launch date.

WiMAX supporters, not surprisingly, felt aggrieved by the decision-Sprint and Intel both left the NGNM Alliance shortly after the announcement. While the US mobile operator expressed disappointment that the mobile operator-led industry body had ditched its technology-neutral stance, Intel’s grievances appeared to run much deeper.

“[The NGMN Alliance] used a biased methodology to evaluate the different mobile broadband technologies,” says Intel’s Alamouti. The crux of Alamouti’s argument is that by adopting a 3GPP-developed methodology for evaluating technology performance, the NGMN Alliance disadvantaged many WiMAX companies as they did not have simulation results to hand based on the 3GPP approach. “WiMAX companies had to reproduce simulation results for a highly complex system according to a new methodology in an unreasonable amount of time,” he says. “It was no surprise that many WiMAX companies weren’t able to produce optimised results [for their equipment].”

Although Alamouti says Intel had the resources to achieve ‘proper optimisation” using 3GPP methodologies, the NGMN Alliance averaged out the sub-optimal results generated by a number of WiMAX companies. This, says Alamouti, gave the NGMN Alliance a false impression about WiMAX capabilities.

“Companies were also providing results for WiMAX when they didn’t even have any WiMAX products, which creates a conflict of interest,” adds Alamouti. “Although Intel had many LTE results, we didn’t submit them because we felt we didn’t do proper optimisation on it.”

From Alamouti’s point of view there is “no appreciable difference” between the respective performance levels of 802.16e and LTE kit as they are both based on similar OFDM and MIMO techniques. And so it comes back again to a question of timing. “We are comparing something that is available today with something that is a few years away,” says Alamouti. And Dr Shakouri believes the NGMN Alliance may well re-consider its 4G verdict. “The door hasn’t been closed on WiMAX,” he says.

Show me the spectrum

The headline 100Mbps plus speeds for individual users, claimed by LTE supporters, is based on the assumption that there will be a 20MHz chunk of contiguous spectrum available. It is a dangerous assumption to make, says Dr. Shakouri. “The reason why WiMAX started off with a 10MHz channel base line is because we wanted to get equipment out into the market as soon as possible, ” he says. “We didn’t start with 20MHz because there are not enough places where it is available.”

LTE supporters might counter that mobile operators could re-farm existing spectrum assets if new spectrum is not available. It’s an argument that PA Consulting’s Westwick believes has some mileage. “Access to spectrum currently favours the UMTS/LTE camp,” he says. “Many mobile operators have got lower frequency bands [GSM at 900MHz], which could be re-farmed for UMTS/LTE and provide national coverage at relatively low cost due to the longer reach of base stations at the lower frequencies. By contrast, the lowest standardised frequency bands available for WiMAX is 2.3GHz and 2.5GHz.”

Dr. Shakouri is not convinced by this argument, as only small chunks of spectrum are available at 900MHz and so would not be suitable for delivering mobile broadband services. “I hear a lot [from the LTE community] about re-farming this and re-farming that, but where are they going to fit 20MHz? This is the biggest challenge that they have. There is no spectrum strategy behind LTE.”

Where new spectrum has been made available, such as India and Malaysia, Dr, Shakouri points out that mobile operators are already pushing ahead with Mobile WiMAX.

It is true, however, that 20MHz channels are more likely to be found in the IMT 2.6GHz (2.5GHz-2.69GHz) extension band. Yet the recent decision by the European Commission (EC) to give national regulators more flexibility in allocating TDD and FDD spectrum might diminish 20MHz availability, at least among EU member states.

Up until the new EC decision, the Electronic Communications Committee (ECC) appeared to mandate that 2x70MHz of FDD spectrum should be awarded in 2.6GHz with 50MHz set aside for TDD. Until the WiMAX Forum develops an FDD profile, such a spectrum allocation would favour the LTE camp. The EC now says that regulators should allocate TDD and FDD spectrum according to market demand, which would reflect a fairer and technology-neutral stance-and perhaps make 20MHz chunks more difficult to find in future European 2.6GHz auctions if TDD spectrum is much sought after.

The telecom industry is no stranger to standards wars but it is unlikely, in the battle for 4G, there will be one outright winner. But with its time-to-market advantage, WiMAX does at least have the chance to steal a march over rival LTE.

WiMAX and LTE not living in perfect harmony

In an ideal world there would be one 4G standard. This would maximise economies of scale; avoid the need for operators to undertake the costly and time-consuming exercise of weighing up the pros and cons of competing 4G variants; and customers would be able to roam more easily abroad.

Arun Sarin, speaking as Vodafone CEO at this year’s MWC event in Barcelona, expressed his desire for LTE and WiMAX to be “integrated” into one standard. And Sean Maloney, Intel’s head of marketing and sales, said in an interview with the BBC earlier this year that WiMAX and LTE “ought to be harmonised” and that Intel itself was “actively looking” at harmonisation.

Yet it is unlikely that LTE/WiMAX harmonisation is going to happen anytime soon. An Intel spokesperson, speaking to analysts at Informa Telecoms & Media, claimed that Maloney’s remarks were taken out of context and that Intel has no plans to develop multimode chipsets that support LTE.

“A single standard is desirable but we don’t see it as practical at this stage,” says Siavash Alamouti, CTO of Intel’s Mobile Wireless Group. “With the upcoming success of WiMAX, however, I hope the 3GPP community will be more eager to work with us on a path towards harmonisation. At this point, though, there are no activities in that direction from either parts of the [4G] community.”

The battle lines between LTE and WiMAX are also not as clearly drawn as they once were. While the initial LTE focus was FDD, Ericsson is now developing an LTE platform that supports both TDD and FDD. The WiMAX Forum, while initially focusing on TDD, is currently working with the IEEE and the ITU to develop a FDD profile. Rob Westwick, a consultant in the wireless technology division at the PA Consulting Group, nevertheless believes that closer interworking with LTE would benefit WiMAX in the long term. “WiMAX vendors have to ask themselves whether they want to be a big fish in a small pond or be a slightly smaller player in, potentially, an absolutely massive pond,” he says.

WiMAX Forum predicts user surge
The WiMAX Forum predicts there will be 133 million WiMAX users by 2012 with Asia expected to grab the lion’s share (45 percent).
WiMAX and LTE look to IMT-Advanced

Although the battle between WiMAX and LTE is invariably described as a ‘4G’ standards war, this is technically not true-at least not yet. While the 802.16e Mobile WiMAX standard has been accepted by the ITU into its IMT-2000 family, this is a family of ‘3G’ standards. Mobile WiMAX and LTE will only become ‘4G’, from an ITU perspective, when they meet the criteria set out by the international standards body for so called ‘IMT-Advanced’ technology.

The ITU’s IMT-Advanced criteria include a minimum downlink speed of 100Mbps and a minimum uplink speed of 50Mbps in the wide-area network. Speeds of up to 1Gbps are specified for local access or very-low-mobility environments. A maximum round-trip traffic delay (latency) of 10ms and a call-setup time of between 80ms and 100ms are also specified. IMT-Advanced also calls for wide channels of between 40MHz and 100MHz each.

Both the WiMAX and LTE communities are already preparing their respective candidate technologies for IMT-Advanced. The IEEE is developing 802.16m, and the 3GPP camp plans to put forward LTE Advanced. The submission of candidate technologies to the ITU is expected to begin in early 2009 and end in October. The evaluation process is set to run until mid-2010.

“We believe we can have 802.16m products ready by 2011-12,” says Dr. Mohammad Shakouri, WiMAX Forum board member and vice president of marketing. “But the biggest challenge facing 16m and other advanced technologies will not be technical. The key question is where will the spectrum come from?”

Leave a comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.