While the move to a more dynamic, IP-based infrastructure holds great technological challenges, the investment and revenue at stake is nothing short of phenomenal, as Alun Lewis explains.
” height=”<% height %>” align=”right” alt=”4G LTE – Let battle commence” class=”articleimage” />
The radio spectrum, it has to be said, is becoming an increasingly crowded, complex and fought-over battleground. Just like any other finite – and fragile – natural resource, the stakes are high. What happens in the largely invisible world of wireless will decide the future of much broader issues in the communications and content sectors as companies battle to find new ways to reach out to their customers.
While core communications networks and their supporting IT systems are undergoing their own transformations, moving towards a far more open and dynamic IP-based infrastructure, dramatic changes are also underway in the radio access area. Known loosely by the convenient catch-all term of 4G, the acronym effectively covers a range of technologies that encompass both evolutions of existing cellular radio systems as well as those of the WiFi community such as WiMax and WiBRO.
The sums at stake are huge – both for the service providers themselves and for the vendors involved. Sprint – one of a growing breed of historically fixed service providers who are now looking to use WiMax to compete head on with both cellular and other fixed providers – is, according to reports, planning to spend $1.1billion this year in WiMax infrastructure, while cable operators in both the US and Europe see the technology as an essential fit into their triple play strategies. The world too is already familiar with the various plans of other fixed service providers – such as BT – using WiFi in both the home and in public areas to complement traditional cellular
According to Miguel Myhrer, Wireless Network Lead for Accenture in North America, “There’s a lot of important activity going on across a number of different parts of the radio spectrum around the world, especially in the US, Eastern Europe and Latin America. This includes bands previously assigned to satellite communications. Some existing service providers see an entry into wireless as a defensive move while other, often newer players have both aggressive business plans and available cash. Comcast, for example, currently has around $23 billion in cash to invest…”
For the cellular service provider, these initiatives present very real threats in the longer term. While voice and basic messaging remain the mainstays of the mobile sector, the world’s consumers are getting increasingly used to living, working and entertaining themselves and their friends in a DSL or WiFi-based broadband world. As the original technology agnostics, they couldn’t care less about auctions, spectral efficiencies and the usual realpolitik games of the standards and licensing bodies. They just want reliable and ever-cheaper wide area coverage – as has happened with the provision of both basic cellular and DSL services to their communities over the last few years.
For the cellular community – represented by the 3GPP – the wireless future that they’ve been considering for the last few years is known as LTE – Long Term Evolution. As such, it fits alongside the complementary developments going on in the core network with the introduction of an all IP network, itself known as System Architecture Evolution (SAE). But what exactly is LTE and what benefits will it bring to the competition-coexistence landscape that’s currently emerging between the different service provider communities ?
Eduardo Sanchez, LTE product marketing manager at Nortel sums up a few of the main points, “LTE is intended to increase both network capacity and throughput by delivering three to four times the spectral efficiency of HSPA Release 6. Latency will also be improved, with a round trip delay time of less than 20 milliseconds – necessary for applications like VoIP and gaming.
“LTE itself is based on a new modulation technology called OFDM (Orthogonal Frequency Division Multiplex),” he adds, “and a new antenna technology known as MIMO (Multiple Input/Multiple Output. OFDM uses a bundle of adjacent narrowband carriers that are transmitted in parallel at different frequencies from the same source. Because of the narrowband nature of the sub-carriers, OFDM is highly resistant to multipath interference that imposes overheads in other technologies.
“On top of this, the orthogonal characteristics of each sub-carrrier allow OFDM sub-carriers to use up the entire allotted spectrum without significant waste, particularly for guard-bands – in contrast to the single carrier techniques specified for TDMA or CDMA systems. Finally, as the OFDM receiver system deploys well understood digital signal processing technologies, this makes it less complex to develop and less expensive to produce.
“The MIMO aspect of the technology works by creating multiple parallel data streams between multiple transmit and receive antennas and by exploiting the multi-path phenomenon to differentiate among the parallel signal paths. Through this, MIMO technology can achieve a multi-fold increase in user throughput gain as well as an aggregated increase in network capacity compared with current generation networks.”
But what of the actual expected performance and the demands that it is going to make on available spectrum? Håkan Andersson, LTE product management at Ericsson, positions LTE as being available in 2007 and says that, “With two antennas per base station and terminal, top speeds will be 144 Mbps downlink and 50 Mbps uplink on a 20 MHz carrier – while round trip time in the radio network is down to about 10 milliseconds. LTE is going to be launched on a number of frequency bands – for example on 2.6 GHz which will be licensed in the coming years. When GSM operators’ licences start to expire around 2010-2015, there should be LTE products available that suit those frequencies.
“Overall,” he adds, “LTE will have support for bandwidths ranging from 1.25 MHz to 20Mhz. The bandwidth used depends on the amount of spectrum available. For example, on the 2.6 GHz band, it will be possible to deploy 20MHz carriers and thereby exploit the full performance of LTE. On frequency bands already used by, for example, GSM, the introduction should probably be done using a 1.25 MHz carrier in the first place.”
This issue of spectrum availability is a crucial one that was also accentuated by Alastair Brydon co-author of an Analysys report.
“The full benefits of 3G LTE will only be achieved if mobile operators are able to use it in 20 MHz of bandwidth – which will require new spectrum allocations. This means that mobile operators will need to secure additional spectrum, such as the GSM extension band, and ensure that other new entrants or new technologies do not get access to this valuable spectrum. 3G LTE is being designed to minimise the cost of upgrades for mobile networks, but there are still great uncertainties over the investments that operators will have to make in terms of network infrastructure and spectrum. The 3G community will need to address these uncertainties quickly if it is to build momentum in 3G LTE,” he concludes.
CDMA community
While the 3GPP naturally represents the GSM community and its descendants, there are also the activities of the CDMA community to take into account as Paul Nerger of mobile experience management specialists, ArgoGroup, explains.
“Although the underlying technologies have been getting closer since the advent of UMTS – which the CDMA community always calls W-CDMA – the CDMA 200 world looks to 3GPP2 rather than 3GPP for future inspiration. As such, their LTE looks to the same future as IMS and UMA – but based instead on EVDO technology evolution, rather than HSXPA. 3GPP2’s LTE is identical – it is just that the faster data and higher throughput will be provided by a competing technology and this has lots of implications throughout the stack.”
But where exactly does this longer term technology fit alongside the current path for HSXPA introductions and, perhaps more importantly in the race to market, against WiMAX?
Bob Brace, director of mobile messaging at the Mobile Data Association, explores some of the implications.
“One of the disadvantages of HSXPA technology is that it is asymmetrical and only HSDPA is available right now. HSUPA provides the uplink piece, but that’s still just over the horizon. One downside will be battery life and 3.5G won’t be any better. It’s also just an interim before HSOPA – High Speed OFDM Packet Access – gets ratified and this could offer peak data rates of 100 Mbps for the uplink and 50 Mbps for the downlink.
“High speed technologies like HSXPA and WiMAX offer the carrot of ubiquitous wireless coverage and incredibly high speeds – but unfortunately neither is true,” he adds. “ The common misconception about WiMAX is 70 Mbps and 70 miles – but the truth is not both at the same time! WiMAX is however available before HSOPA so therefore it’s going to be about coverage, devices and pricing. The problem with 3G was no flat-rate pricing, but I think operators are starting to see the light in coming up with tariffs like these and allowing users outside the ‘walled garden’. While people may start talking about operators becoming that hated and despised term – the ‘bit pipe carrier’ – I think that the operators themselves are starting to see the reality facing them: they can’t control the user experience or charge huge premiums if they are going to compete and control rival technologies like WiMAX. They really need to establish usage patterns and lock in a user base before other technologies can get a foothold.”
Not seen in isolation
As always however – with anything to do with telecommunications – no part of a network can be seen in isolation to the rest and concepts like 4G and LTE are no different. For Jeremy Barnes, director of strategy solutions at Tellabs, “In a market where operator margins are declining, LTE needs to be approached in the light of how much it costs to achieve. There are, after all, no revenue guarantees – however impressive the promises may be…
“That said, LTE should have a number of important impacts on today’s networks and operators have to start planning for this now and ensure that today’s capital expenditure investments result in transport networks capable of supporting LTE. Firstly, LTE will require a significant increase in backhaul capacity in order to deliver on the capabilities of its enhanced air interface and 100 Mbps down and 50 Mbps up speeds.
“Secondly,” Barnes continues, “With RNC functionality distributed to the NodeB, LTE creates a requirement to fully meshing the NodeBs – some 10,000 to 20,000 for a mobile operator running a network in a ‘mature’ market such as a country in Western Europe. The cause is call hand-off. A call originating on the ‘anchor’ NodeB that subsequently travels across the country – as during a car or train journey – moving from NodeB to NodeB requires that each NodeB utilised will need to communicate with both the anchor NodeB (for billing etc.) and the previous and next NodeB for call hand-off. IPVPNs are likely to be the best solution for this type of meshing and, to support a mesh of this size, some degree of hierarchy is likely to be required. A hierarchical IPVPN provides an ideal migration path for LTE and the all-IP RAN.”
These implications are also highlighted by cellular backhaul specialist Celtro as CTO Shahar Gorodeisky explains.
“Many assume that 4G is going to open the door to bandwidth intensive multimedia applications – but what’s the point of spending vast sums on 4G technologies if the backhaul can’t handle it? Traditional wireline and microwave backhaul technologies cannot meet the capacity requirements of widespread HSDPA, let alone HSUPA or 4G – so making the backhaul network is one of the primary challenges facing any operator looking to deploy a next generation network.”
While the R&D engineers continue their work and the investors and finance departments decide which parts of the 4G jigsaw look like giving the maximum buck for byte, it’s time as always to signal a déjà vu-tinged note of caution. Having the best technology is no guarantee of success – as anyone with a Betamax VCR in the attic can testify. Keith Miller of switching specialist Appium sounds a final caveat.
“With all mobile operators still squeezing as much traffic as possible onto their current and very efficient circuit-switched-over-RF delivery at the cellular end, it is hard to see what will be the business case for any telco to embrace 4G soon. Packet delivery of voice is notoriously inefficient so the only real driver must be lower services and lower cost of integration.
“In some ways, IMS is helping prepare the ground for this. As yet however, there have not been many successful 3G services that need 3.5G speeds – let alone 4G. There are, however, signs that convergence is starting to take hold as a useful differentiator with triple play and quad play offerings showing good success rates in some countries. Mobile operators will be slower to grab hold of the 4G nettle – having been stung badly with their 3G rollouts that are still very much work in progress…”