re: GPRS: Ian Channing's "Not so Fast!"
Ian Channing is Technology Editor for Mobile Communications International and over recent years has written some of the better articles on GSM migration, WAP, 3G, and the technologies.
His articles are, literate, and generally technically accurate. His observations are keen, and timely.
This article was published early April. Some key points from the article:
* Amid the arguments about data speeds the fact that GPRS is simply a bearer technology is often overlooked.
* 130 GPRS infrastructure contracts have been let worldwide in countries as far-flung as Australia, Hong Kong, the Philippines, Bolivia and South Africa, as well as the dozens within Europe.
* Handset vendors always get the blame. Lack of hand-sets as the reason for launch delays is a long-running mantra of the cellular industry going back to the introduction of GSM.
* One of the problems dogging GPRS is the time lapse between the much-hyped launch of a GPRS phone and the moment when a customer can go into a store and buy one.
* It is now generally accepted that the industry over-hyped the data speed capabilities of GPRS and, as a result, it has been working hard to reduce user expectation. Nonetheless, at this stage of technology development, the data speed reality is even less than the lowered expectation levels ... although there is general support for the instant connectivity GPRS offers
* The initial equipment has not given the gains of bit rate that have been talked about in the hype associated with GPRS. The actual gains over traditional circuit-switched data transmission so far have been fairly minimal. With GPRS you need to grab a number of timeslots to achieve the higher bit rates and you need to use the higher coding schemes to achieve those higher bit rates, so there is a matrix between bit rates and the coding schemes that you can employ. Initial handsets have operated in the lower quarter in that they have only been able to grab one or two timeslots and use the lower coding schemes. So there is a time lag in terms of the higher bit rates becoming available. However, this is righting itself and there will be vast improvements in achievable bit rates within six months or a year.
* Historically, the network operator has been the sole proprietor, owning the network, the customer and all the services and applications. In the new content and application-driven scenario, this historical situation is no longer sustainable, but many operators have been slow to recognise this paradigm shift.
* It is all about creating the business case for data services ... and beyond GPRS lies the mobile multimedia world of 3G and getting the billing model right at this stage is imperative.
The full article:
>> Waiting for GPRS
Ian Channing
Mobile Communications International
23 April 2001
mobilelatestnews.com
In 1997, a speaker at the GSM World Congress got a little carried away and predicted that GPRS trials would begin in 1998 with commercial services being launched in 1999. Four years on, with almost the entire global mobile industry assembled in Cannes for its annual get-together at the 3GSM World Congress, one would have expected the mood concerning GPRS and the evolution to 3G to be upbeat and positive. After all, we’re finally on the verge of widespread services. Aren’t we?
In fact, the messages coming from key players were very mixed and frequently gloomy. Intel vice president and general manager, Hans Geyer, even went so far as to predict that the world’s telecommunications industry was heading for bankruptcy even before a 3G call had been made. On the subject of GPRS, Alcatel’s Jacques Combet told a press conference that he did not believe that the market would explode until 2002 due to a backlog of unanswered and very serious technical questions. "The technology is not yet bringing what people expected, and operators want to avoid the disappointment they saw with WAP," said Combet. Adding to the general air of discouragement, Orange executive vice president, Richard Brennon, told a Cannes press conference that the company did not see a mass consumer market for GPRS in France and Britain before next year.
It seems as if GPRS, the saviour of WAP and the key stepping-stone to 3G - after lengthy delays during the standardisation phase - is going to suffer further delays during implementation and commercialisation.
The question must be why? Is it handset delays? Functional issues with the technology? Interoperability issues? Billing? Or are we trying to go too far too fast?
Handset vendors always get the blame. Lack of hand-sets as the reason for launch delays is a long-running mantra of the cellular industry going back to the introduction of GSM. Certainly, GPRS handsets are not thick on the ground but whether this shortage is the main reason behind the ongoing delay in GPRS is questionable. In any event, the handset vendors seem to be getting themselves revved up to begin volume deliveries, although the timescales are lengthy.
During the early phases of GPRS roll-out the Motorola Timeport 260 has been virtually the only phone available in large quantities and the mainstay of the few networks which have launched commercial GPRS services. Motorola is consistently vocal about GPRS and at Cannes promised four more GPRS handsets by mid-year adding that its entire range would be GPRS-enabled by year-end. Fred Kuznik, president of Motorola EMEA, was one of the few at Cannes to have a positive spin on GPRS, saying, "we are saying GPRS is coming to life. It’s no more babies with nappies."
Mitsubishi, Philips, Samsung, Sagem and Siemens all have GPRS handsets in their portfolios, although none seem yet to be available in volume. Estimates of volume deliveries range from Q2 to Q4 2001. Ericsson has announced that its Bluetooth enabled R450 GPRS phone will be available in this quarter in quantity and this could have a significant impact on the GPRS market. Nokia says it will have GPRS phones in the third quarter and commercial volumes in the fourth quarter of this year. Alcatel launched its GPRS phone at Cannes but it will not be available before May.
Undoubtedly, one of the problems dogging GPRS is the time lapse between the much-hyped launch of a GPRS phone and the moment when a customer can go into a store and buy one. Only one operator with a commercial GPRS service - T-Mobil of Germany - has actually stated publicly that a customer can go into one of its retail outlets and buy a GPRS phone. According to John Hoffman, senior director GPRS for the GSM Association, "the handset issue is probably hurting the uptake of GPRS more than anything else. An operator is not going to put a lot of effort into publicly positioning and selling a product if there are no or only a limited number of devices that can access it."
However, to balance the debate, a senior engineer from one of the leading test and measurement outfits who has been benchmarking GPRS deployments, told MCI recently that even if there were plenty of handsets available, problems with the networks would often render them useless.
There are also issues surrounding the data speeds offered by the available GPRS handsets. At the current stage of development most are offering one timeslot upstream and two/three timeslot downstream operations - although the Ericsson R450 will provide four timeslots downstream. At 14.4kbit/s per timeslot, this gives potential data throughput speeds of 28.8kbit/s or 43.2kbit/s downstream - under ideal conditions and where there is sufficient network capacity. The reality, in the heavily loaded networks of Europe, is likely to be considerably less. Certainly, GPRS offers the always-on connectivity but in terms of data speeds it offers little more than circuit-switched GSM and less than was promised by HSCSD.
It is now generally accepted that the industry over-hyped the data speed capabilities of GPRS and, as a result, it has been working hard to reduce user expectation. Nonetheless, at this stage of technology development, the data speed reality is even less than the lowered expectation levels. Reports from operators in markets where GPRS is commercially available, such as Europolitan in Sweden, suggest that this shortfall in data capability is being picked up by users, although there is general support for the instant connectivity GPRS offers.
HSCSD was caught in the technology log jam caused by delays in standardisation and met with no real success, virtually its only proponent amongst the vendors being Nokia. However, in certain applications, such as file transfer, HSCSD does offer considerable benefits and, on present evidence, a faster data throughput experience. However, to exploit its full potential HSCSD needs access to four timeslots and in the mature networks of Europe that capacity simply does not exist. The same argument applies to a certain extent with GPRS which also needs access to multiple timeslots. But the packet technology does enable operators to maximise their available bandwidth, so in a mature network GPRS ought to offer customers higher data speeds than HSCSD.
Chris Harvey, 3G business development manager at Quotient, sums up the situation:
"The initial equipment has not given the gains of bit rate that have been talked about in the hype associated with GPRS. The actual gains over traditional circuit-switched data transmission so far have been fairly minimal. So there has not been a lot for the market to crow about initially although that is improving all the time".
"With GPRS you need to grab a number of timeslots to achieve the higher bit rates and you need to use the higher coding schemes to achieve those higher bit rates, so there is a matrix between bit rates and the coding schemes that you can employ. Initial handsets have operated in the lower quarter in that they have only been able to grab one or two timeslots and use the lower coding schemes. So there is a time lag in terms of the higher bit rates becoming available. However, this is righting itself and I think there will be vast improvements in achievable bit rates within six months or a year."
Amid the arguments about data speeds the fact that GPRS is simply a bearer technology is often overlooked. Without the availability of compelling applications operators will not be able to generate the service revenues that they need for return on their investments. The problem for operators is that the GPRS application picture is unclear and confused. As could be seen at Cannes, there is no shortage of new players who have created what they believe are compelling applications both for GPRS and ultimately for 3G. The issue is not only turning these applications from a booth demonstration into real services on live networks but also persuading the public to use them extensively.
There are also market issues involved. Historically, the network operator has been the sole proprietor, owning the network, the customer and all the services and applications. In the new content and application-driven scenario, this historical situation is no longer sustainable, but many operators have been slow to recognise this paradigm shift.
Keen as the new application and service developers are on popularising their products, they are not philanthropists and they will be looking to share in the revenues generated by the operators from the deployment of these applications. However, operators have, to date, been reluctant to admit these new players into the value chain - a situation which must change if the new world of GPRS and, more importantly, 3G is to be successful.
Ben Wood, a director of research consultancy Mobile Lifestreams, thinks the way forward is through partnerships.
"We have the packet technology but the reality is that we need some help to get the applications out there and that means bringing in new players, perhaps like AOL, who have an expertise in instant messaging. So maybe working in partnership will be the way that GPRS will be brought to the market successfully, rather than the operator controlling everything from start to finish."
As Wood points out it is all about creating the business case for data services:
"WAP has been a perfect example of a case where it is impossible for anyone to make revenue from the business except the network operators. Until there is a clear business case for wireless data applications, WAP service being one of many, the development community that needs to embrace these technologies is just going to sit back and wait".
How these partnerships are going to be created is another issue, says Wood:
"For example, Nokia and Motorola are creating wireless gaming servers and saying they will go to the operators and negotiate revenue-sharing deals on behalf of the smaller players. Perhaps there will be a situation where the operators broker deals direct or perhaps it will be done by going through third parties such as AOL. My view is that it is not going to be the network operators so it is going to be someone else who is going to lead the network operators to drive those applications which will use services like GPRS."
Even when the operators have done their partnership deals and integrated these new services and applications into their GPRS networks, there is still another major issue: how do they bill for them? Patently, regardless of how good the application or service may be in theory, if the operator cannot charge for ti then it becomes valueless.
Arguments about how to bill for data-based applications and services have been raging for some time and are little closer to resolution now than they were two years ago. There is a general consensus that the billing by airtime paradigm will not fly (which is a considerable disappointment to many operators) but on what basis can you charge? This is not an academic discussion; beyond GPRS lies the mobile multimedia world of 3G and getting the billing model right at this stage is imperative.
There is a number of possible solutions for billing in the GPRS/3G environment; billing by content or value, billing by packet sent or received, and billing by bundled bytes. No single solution enjoys universal support, mainly because this is a new world for both network operators and billing system vendors and both parties are feeling their way. Certainly the existing billing mechanisms used by network operators are inadequate to handle data services - which means that new mechanisms must be developed, validated and deployed. Whether these uncertainties on what is the right billing model are contributing to the delay in GPRS deployment is not yet clear.
Ben Wood sees the GPRS billing model developing over time:
"We believe that operators have difficulty billing GPRS packets or content, so it comes down, not to billing individual packets and showing that on the bill, but giving people these bundled bytes and policing it. So you give someone a notional bundle of 5Mbytes a month, you police it and roughly measure their data usage. If they start exceeding the allocation by an enormous amount you can go back and say you need to upgrade to the next tariff or cut them off. That may be the first route to market but we see those billing models becoming more sophisticated over time and expanding to content-based or packet-based billing."
The cellular industry has always enjoyed a high level of confidence in itself and its future. And why not, with subscriber numbers showing double digit growth each year and revenues following a similar pattern? This confidence was given a bad knock with the relative failure of WAP, the first time the industry has launched a major initiative which the public effectively rejected. The echoes of what may be described as The WAP Factor are still reverberating around the industry and are probably contributing to the current GPRS scenario.
The WAP Factor puts operators in an unenviable position; if they hype GPRS and it fails to deliver on its promise they will be condemned; if they are overly cautious the market will complain at, and be set on edge by, their hesitancy. Truly between the rock and the hard place.
Mobile Lifestreams’ Wood understands these concerns but warns that there are dangers in an overly cautious approach:
"From the point of view of the industry I am happy to see a more cautious approach but not at the expense of getting these technologies out there. What is going to happen is that 3G is going to rush up to delivery so quickly that unless they can start figuring out the GPRS market and making a real business case for things like entertainment and multimedia messaging using GPRS, then all of the business cases that have been laid down for 3G will be in jeopardy, and that will do even more damage to our industry as a whole. Expectations need to be managed but not over-managed, otherwise we will end up pedalling backwards".
The GPRS scene is not, however, all doom and gloom. According to EMC almost 130 GPRS infrastructure contracts have been let worldwide in countries as far-flung as Australia, Hong Kong, the Philippines, Bolivia and South Africa, as well as the dozens within Europe. The GSM Association says that as many as 40 operators have launched commercial or quasi-commercial GPRS services and the handset vendors are approaching the point of volume production.
In reality, GPRS must succeed. If the industry does not get GPRS right then 3G could be seriously, perhaps fatally, wounded. There are other factors.
The telecommunications sector has taken some heavy knocks from the world’s financial community in recent months and desperately needs to win back the confidence of the stock markets. Success for GPRS would provide a much-needed boost to the mobile sector; a WAP-like failure could have serious consequences.
GPRS has already been delayed, a further six months delay will not be significant if when the launch comes everything works perfectly.
With the arrival of 3G likely to be put back, GPRS is going to be around for a long time - so let’s make sure we get it right first time. <<
- Eric - |
