• Head of Business Intelligence -...
• Project Manager
• Business Analyst - Melbourne CBD...
• Project Manager - Trading...
• Desktop Team Leader

The path to 3G

Steve Jones, Information Age

21/08/2001 17:12:15

The path to 3G

3G: a wireless marvel or an investment disaster? This question is weighing heavily on the minds of many telecommunications carriers around the globe, especially as the initial allure of 3G wireless technologies wears off. The technology promises high-speed data, mobile streaming video and anytime-anywhere access. But implementation costs have significantly dampened enthusiasm. Depending on whom you ask, 3G wireless technology will be the greatest innovation in the 21st century, or the biggest bunch of hype in the world.

As it stands, 3G is riddled with the deep craters indicative of an evolving wireless landscape. Naturally then the confusion surrounding this new and uncharted territory can be overwhelming, and the danger of losing one's way great. To help sort out the challenges and avoid potential pitfalls, we present this 3G map as a guide for the seemingly treacherous journey through the wireless world.

Definition

3G wireless technology is a global communication technology that makes possible packet-based transmission of digitised voice, data and video. The 3G umbrella encompasses a range of competing wireless technologies, such as Code Division Multiple Access (CDMA) 2000, Universal Mobile Telecommunications Service (UMTS) and wideband CDMA (WCDMA). The International Telecommunication Union last year spelled out the guidelines for 3G in the IMT-2000 framework as being capable of data transmission speeds of 144Kbit/sec inside a moving vehicle and 2Mbit/sec in a fixed location, using packet-based rather than circuit-switched technology and permitting global roaming.

Carriers

The race to 3G is about spectrum. Nowhere is this more apparent than in the US, where a shortage of frequencies may seriously undermine 3G implementation. In an effort to comply with an agreement made at the World Radio Conference (WRC) in the spring of 2000, the US Federal Communications Commission is examining ways to make more spectrum available for 3G. The range from 1710 to 1855 MHz, in addition to the 2520 to 2670 MHz frequency band, was determined to be the global spectrum assignment for 3G applications at the WRC.

Currently, the US Department of Defense uses this 1.7GHz band that is being eyed by prospective 3G wireless carriers for services such as wireless high-speed Internet access. The Defense Department uses this spectrum for satellite control and military purposes. According to the Defense Department, relocating to another portion of the spectrum would cost billions of dollars and take a decade or two to achieve. A proposed alternative has been for wireless providers to share the spectrum with the Defense Department, but that would impose many restrictions on the Defense Department and the carrier's operations. It remains to be seen how the FCC handles this sensitive issue.

Agreements made at the WRC in the spring of 2000 were designed to facilitate worldwide roaming capabilities. The establishment of specific frequencies for 3G wireless technology requires that the US re-evaluate its spectrum allocation to embark on 3G at a global level. However, citing issues of national security, the military is leery of sharing this band, especially with the commercial sector. In an attempt to compromise, the National Telecommunications and Information Administration, the government agency that issues government-operated spectrum to the commercial sector, may pay the Defense Department to move to another spectrum.

Despite a potential deadlock in the US, the 3G frenzy in North America continues as carriers jockey for position. Although no 3G licensing has yet occurred in the US or Canada, several of the major carriers have outlined their 3G plans. Unlike Europe, which has a uniform wireless standard based on the Global System for Mobile Communication, North America (and the US in particular), currently uses Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), GSM and CDMA. With four platforms in existence, the road to 3G is not guaranteed to be straight and narrow.

The existing networks are determining what path the carriers are choosing to get to 3G. In North America, as in the rest of the world, there are two seemingly related 3G options: WCDMA and CDMA 2000. While WCDMA is a natural progression from GSM networks, CDMA 2000 builds on platforms currently using CDMA technology. Because Europe already has GSM in place, these carriers will use WCDMA technology to provide 3G services such as mobile streaming video. This also means that North America and the US may be using both WCDMA and CDMA 2000 to move to the next generation of mobile applications.

As if the challenges presented by spectrum allocation were not enough, it's also not possible for carriers to implement WCDMA and CDMA 2000 within their networks. So in addition to finite resources, North American wireless service providers must roll the dice on which 3G platform will be most advantageous to implement. Currently, CDMA 2000 and WCDMA have strong potential, according to preliminary 3G plans.

Hoping to be the first to implement 3G in the US, Sprint PCS Group and Verizon Wireless are conducting field trials of CDMA 2000 1XRTT. The 1X denotes a CDMA 2000 standard that is intended to double the voice capacity of current CDMA systems and improve data rates to 144Kbit/sec. Sprint PCS expects to complete its network upgrade to CDMA 2000 by the end of the first quarter 2002 with Verizon following closely on its heels. On the other hand, VoiceStream Wireless and AT&T Wireless Group are taking the WCDMA route. Another major player, Cingular Wireless, had not officially committed to either platform by press time.

The 3G path is also a guessing game from a global perspective. While Europe is moving toward WCDMA, countries such as China and South Korea seem poised to use CDMA 2000. This is in contrast to Japan where all eyes are on NTT DoCoMo, which upon completion of a nationwide WCDMA platform for 3G will have $US8 billion in committed resources, as estimated by The Goldman Sachs Group. Currently, DoCoMo is conducting a beta test of approximately 4000 users with a large-scale implementation scheduled for October.

However, until spectrum allocation is resolved worldwide, the technology used for 3G network implementation is not the issue companies are most worried about. With 3G spectrum auctions in Germany and the UK setting a dangerous precedent, some companies are finding that the prices they paid for licensing are affecting their ability to invest in 3G infrastructure. This has surfaced recently as a major pothole on the road to 3G, one that is causing many wireless carriers to need new wheel alignment.

Vendor talk

In addition to 3G's already hefty price tag, the end-user devices to be used in conjunction with the next generation of wireless infrastructure are also expected to be expensive. It's estimated that 3G-enabled handsets will cost $300, in addition to monthly service fees that could be as high as $90.

Further aggravating the delay of 3G implementation is the lack of availability of these pricey handsets. For example, of the 10 vendors expected to make phones for DoCoMo's 3G implementation this October, only two companies have products available. Moreover, the lack of applications for 3G handsets is another hurdle. Not only do equipment manufacturers need to speed up 3G development, but so do application developers if 3G is to reach its potential. The catch is that both players are motivated by demand, and the jury is still out on whether there will be widespread consumer demand for 3G products and services.

Despite these hurdles, major equipment vendors have mostly laid out their blueprints for migrating toward 3G technology. In general, companies are making devices for CDMA 2000 and WCDMA although many corporations have aligned themselves with specific platforms through strategic partnerships. For example, Motorola, Lucent, Nortel, Samsung and Qualcomm are providing equipment for beta tests of CDMA 2000 rollouts in the US Meanwhile, Ericsson and Japan Telecom recently partnered to test the world's first field trial of voice over IP using WCDMA. The following summaries give a brief overview of the direction some of these major players are moving.n Nokia: Nokia's solution for 3G implementation revolves around an all-IP network consisting of an IP radio access network (IP-RAN) and IP Mobility Core network. These networks can support major interface technologies, including GSM, wireless LAN Wi-Fi/802.11b and WCDMA, and are backward compatible with existing mobile networks. The IP Mobile Core network uses IPv6, providing easy interface for multivendor technologies and has quality of service and security management features incorporated in it.n Ericsson: Ericsson recently announced the completion of a 3G CDMA 2000 Mobile Switching Center (MSC) that is set to be deployed into its mobile networks. Actively involved in defining standards for CDMA 2001*EV DO (Evolution Data only), Ericsson's solution needs only a software upgrade from existing CDMA One networks and will support various technologies, such as Wireless Application Protocol, Bluetooth and Java on its 3G platform, focusing on offering real-time wireless voice and data capabilities to customers.n Motorola: Motorola's Aspira technology is based on all-IP core network in collaboration with Cisco. It devises separate paths for migration from CDMA and GSM platforms and includes IP billing, network management and operational support features. The company recently completed lab tests of delivering live video over its 3G productApplicationsWith all the grey areas surrounding 3G, one might wonder what the rush is to use a technology that has different standards, costs a lot and evicts other residents from their spectrum assignments. The answer is in the potential applications that range from streaming video, integrated voice and data services, to full-scale multimedia.

Potential 3G-enabled wireless devices (mobile phones, PDAs and others) could include the functions of a video camera, computer, stereo and radio combined. The biggest attraction for mobile consumers would be "Mobile Internet" services, creation of personalised portals that would integrate unified messaging and mobile-commerce (m-commerce) features. Applications such as e-mail, Web-browsing, stock-quote tracking and airline reservations are only the beginning.

Videoconferencing and online "window" shopping are set to be the next generation of applications, with roaming capabilities throughout Europe, Japan and North America being the next big step. In particular, one school of thought believes that the push for 3G will be driven by m-commerce, the idea that financial transactions can be carried out over mobile devices as easily as the wired network.

Studies supporting this predict that Europe will benefit the most from m-commerce in the next five years, consistently acquiring 40 to 50 per cent of the m-commerce market. This is in contrast to North America, where the outlook is at 15 to 25 per cent. But other reports cite spectrum issues and 3G infrastructure implementation delays as hurdles to the potential of m-commerce. These studies conjecture that m-commerce will not become a major market player until 2007.

Technical framework

From a technical standpoint, the IMT-2000 framework also supports both of the signal transmission technologies of WCDMA and CDMA 2000. In particular, European countries and Japan have chosen the UMTS terrestrial radio access (UTRA) as a solution to move their networks into the next generation of mobile networks.

This is the third generation of mobile developed by the European Telecommunications Standard Institute (ETSI) based on the ITU's IMT-2000 framework. On the US side, the new generation, based on the same framework, is represented by the standard CDMA 2000.

UMTS is a packet-based transmission of text, digital voice, video and multimedia at data rates up to (and possibly higher than) 2Mbit/sec. In January 1998, the European Institute settled a specification of the UTRA using a combination of two 3G-radio wideband multiple access technologies: WCDMA and Time-division/CDMA (TD-CDMA).

The first system, FRAMES Radio Wideband Multiple Access mode 1 (FMA1), operates on unpaired frequency bands and is used with time-division duplex (TDD) operations or TD-CDMA. Each carrier has a frequency spacing of 1.6MHz, on which user channels are separated into time slots. Within each slot, spreading codes create supplemental separations. Its frequency range is between 1885 and 1920MHz for the downstream bandwidth and 2010 to 2025 MHz for the upstream bandwidth.

The second system, FMA2, is used for wide-area service and refers to WCDMA. WCDMA is based on frequency division duplex (FDD) operations that operate in paired frequency bands.

The FDD bands use the 1920 to 1980MHz downstream bandwidth and the 2110 to 2170MHz band for upstream purposes. FMA2 allows 5MHz frequency spaces per carrier slots to implement FDD. Frequency bands 1980 to 2010MHz and 2170 to 2200 MHz are used for Earth-to-satellite communications.

The maximum throughput of UTRA is related to the environment. In rural areas, it will be at least 144Kbit/sec; in cities, at least 384Kbit/sec; and indoors, a minimum of 2Mbit/sec.

They will provide high service flexibility with support of multiple parallel variable-rate services on each connection. To define what can be considered as rural, city or indoor places, a cell structure is implemented.

Cell principles

The 3G network architecture is based on two main principles: Cellular networks should be structured to maximise network capacity and offer multimedia services independently of the place of the end users. The UMTS network will be planned to use a hierarchical cell structure using pico, micro, macro and satellite cells.

The pico cells would be deployed for private, indoor services in the unpaired frequency band. Their locations would be in areas where there is a demand for high data rate services, such as laptop networking or multimedia conferencing.

The micro cells would be used for outdoor coverage, where high capacity is required and larger cells cannot cover, such as city streets hidden by large buildings. The shape of the micro cells would be canyon-like, in order to fit the topography of a street.

The macro cells have a hexagonal shape and provide wide-area coverage. They constitute the core of the network coverage and are based on the existing second-generation mobile network.

Finally, satellite cells allow a global wireless communication with high capacity. In each cell, the UMTS standard lets the operator choose between using the FDD or TDD methods, depending on the traffic.

Following the success of the Subscriber Identity Module (SIM) in GSM for its high security and customisation features, UMTS phones will keep this system with the UMTS Subscriber Identity Module. The high capacity of communication that UMTS offers will have to be in accordance with the capacities of the phones. The challenge of the next two years is for the "smart cards" constructors to create larger built-in memory SIM so to stock data such as pictures, personal files or signatures.

A simple vision of the UMTS network architecture is to divide the cellular network into parts where the UTRA system will be inserted. On one side of the UTRA system is the user equipment domain while on the other side is the infrastructure domain.

The user equipment consists of the cellular phone or another end device, and uses a radio interface to communicate with the UTRA system. UTRA consists of a set of Radio Network Sub-System (RNS) that can be assimilated to the Base Station System in the GSM world.

The RNS manages resource allocations of the radio link to the end user. Each RNS is composed of Radio Network Controllers: RNC (same main function as the Base Station Controller in GSM) and Node Bs (comparable to the Base Transceiver System in GSM). RNC control resources allocations of the Node Bs and manage hand-over features. RNC have the possibility to communicate between them via direct links.

The infrastructure domain or the core of the network provides the features of management of user location information, control of network features and services and the transfer (switching and transmission) mechanisms for signalling. It is subdivided into the Serving Network Domain, the Home Network Domain and the Transit (Jones is the associate professor at the Ball State University Center for Information & Communication Sciences.)Local 3G players spend up, slow downBy Pete YoungThe hangover from the 3G auction extravaganza has left Australia's telco sector treading a path of careful sobriety.

It may not appear that way to shareholders and lenders still reeling from the 3G spectrum auction blowout. Some of the carriers who splurged $2 billion on 3G wireless bandwidth have already begun signing deals worth at least as much again to start turning their spectrum into revenue-generating properties.

Over the past few months, for example, Hutchison 3G Australia and CW Optus have committed a total of $1.7 billion to build out their 3G infrastructure. Hutchison Telecommunications Australia and Telecom Corp of NZ say they will spend $830 million to build out their 3G mobile phone infrastructure in Australia. Their joint venture, Hutchison 3G Australia, is basing its network on Wideband CDMA technology with Ericsson as prime contractor.

The deal was unveiled two months after Cable and Wireless Optus announced a $900 million pact with Nokia to achieve the same result: an operational 3G network leveraging GSM technology. But underneath the formidable numbers, the telcos are pursuing a two-layer strategy characterised by prudence and caution.

The massive infrastructure investments of Hutchison and Optus are being stretched over periods of four years and seven years respectively.

The other two major holders of Australian 3G spectrum licences are being even more circumspect. Telstra and Vodafone are willing to flag 2004 as the rollout year for their third generation wireless offerings. But to date both have maintained a deep silence on when they plan to turn the first sod on their 3G buildouts.

Despite the 3G tag they've hung on their infrastructure announcements, neither Hutchison nor Optus is about to take a leap into the undeliverable deep blue of true 3G. For one thing, they must mark time until next year before the current tenants of their new spectrum are evicted by Canberra. For another, customer attitudes and device manufacturers have yet to catch up with 3G technologies.

So carriers are focusing their attentions in the short term on the more immediate money-spinner of 2.5G in the form of 40Kbps General Packet Radio Service.

"Our view is the battleground for 3G will be fought in the 2.5G world of GPRS," says Optus director of mobile business Peter Acheson, whose job is to work with corporate customers on implementing mobile data solutions.

"Corporate customers using WAP services that trundled along at 10Kpbs had new possibilities opened to them when Optus became the first carrier to introduce 40Kbps GPRS."

That's only a quarter of the speed that GPRS in theory can support but it fits the carriers' post-auction, step-at-a-time strategy to minimise risks while maximising returns. Over time, the implementation of full CDMA-2000 or W-CDMA networks promises to boost 40Kbps speeds about eight-fold.

As part of the cost of 3G enhancements, the migration to higher speeds will require carriers to increase the density of the base stations on their networks. However wireless applications developed for the slower environment won't have to be scrapped when 3G arrives, according to Acheson. They'll just run faster and be able to handle large enclosures like video segments, he says.

The second strand of the telco's strategies is to hedge their longer-range risks by setting up research facilities to smooth the path to full 3G applications even while concentrating short term on GPRS and 2.5G.

NEC Australia is proposing a next generation wireless communications research centre in Melbourne to be funded from $176 million in taxpayer funds allocated for Centres of Excellence in information and communications technology.

NEC already has 70 engineers in its existing Melbourne facilities where it is working with R&D partners on 3G mobile systems based on the same Wideband CDMA technology Hutchison and Optus are deploying. NEC would like to see the proposed new centre grow to 100 researchers over five years. Its areas of interest would include commercially viable enhanced air interface protocols, smart algorithms and design of wireless radio chip sets.

NEC's partner in the proposal is Advanced Communications Technologies, a US company with an Australian-based R&D team. ACT is working on software modules that promise to cut through the expensive muddle of different 3G protocols.

Called software defined radio, the concept would allow carriers to avoid the expense of building multiple networks to support multiple 3G protocols.

Much as different operating systems can be installed on a generic personal computer, CDMA 2000 or W-CDMA modules could be installed on generic base stations, depending on which standard a carrier has designed its network around.

"Particularly in the US market, which is a mishmash of conflicting 3G standards, you can buy a base box and then use software modules to meet whatever combination of protocols you need," says ACT corporate communications manager Jason Webster.

Already up and running as a platform for helping corporates tailor their existing applications for mobile use is FutureLab, a facility jointly supported by Nokia and Optus. Conceived in the middle of last year, FutureLab became operational in January.

"It is a place where corporates can come with an application they are interested in mobilising," says Optus' Acheson. "We can deal with systems integrations issues, scope it and test it once it's done."

Lotus Smartsuite is among the applications which have gone through FutureLab en route to Optus' mobile environment. The net result has been wireless access for corporate customers to their Lotus e-mail, calendar and diary functions.

Optus has also partnered with Wireless Knowledge to deliver applications that give corporate customers mobile access to their Microsoft Exchange e-mail.

With Microsoft and Lotus accounting for a claimed 95 per cent of the corporate e-mail market, "we feel we have that area covered", says Acheson. In addition, Optus is signing up what he typified as "a host of application development partners" to populate its stable of 2.5G offerings.

Gartner Group telecomms specialist Geoff Johnson says the Asian region, especially Hong Kong and China, is being thickly seeded with 3G mobile wireless R&D facilities. Backing them is "the heavy end of town - IBM, Nortel, Sybase", he says.

"They tend to be staffed by non-Chinese, particularly Europeans - because they want to take the results back to the European mobile market."

An Australian player carving out a reputation on Asia's mobile applications development scene is Sydney's Soprano Design. Vice-president Rob Hellstrom says carriers and corporates are looking for solutions that straddle the key areas of information, subscriber location and mobile commerce.

"My feeling is that China, Hong Kong, Japan, Korea and Taiwan are the most interesting countries today in terms of (3G) experimentation and successes.

"The good news for Australian developers in general is that they are in the same time zones as those countries. So there is a very great opportunity to take any solution they have for the corporate market and take it to Japan."

The downside is that no carrier in Australia appears to have a rounded migration strategy infused with a full understanding of the costs of rolling out 3G complete with a good appreciation of the end-to-end user experience, Hellstrom says.

On the plus side, Australia has relatively low-cost 3G spectrum ($US50-60 per head versus $500 per head in Europe according to Gartner). It also has Japan breaking trail when it comes to proving up 3G-enabled consumer applications.

But even when successful 3G applications start flowing out of the development labs, Hellstrom doesn't believe the changeover from 2.5G in Australia will take place in a short, sweet timeframe. What will evolve is an underlying 2.5G platform supporting islands of 3G which will be co-terminous with the urban population centres, he believes.

Gartner's Johnson argues that Australia is still two to three years from 3G and agrees it is "not unrealistic" to see GPRS as the battleground on which the eventual 3G victors will be decided.

"It is a question of business adoption, not technology, and the habit-forming patterns that dictate business adoption will come from today's 2G phones and PDAs."