By Peter Dykes

Very few industry professionals still doubt that Internet Protocol (IP) will become the industry standard for next-generation voice and unified telecommunications services. A recent report from UK-based independent telecoms consultants Schema is the latest to confirm that IP technology is enabling the rapid convergence of network intelligence, IP and mobile communications. According to report author Robin Duke-Wolley: "The trend towards IP telephony will present tremendous service opportunities for old and new service providers alike. It will help create an open environment for fast and easy service development, opening up the market to numerous horizontal and vertical, sector-targeted, value-added services."

There is no doubt either that IP will make a tremendous difference to the services offered and the ways in which they are presented, particularly with the impending introduction of data networks in 2G environments and, of course, third generation WCDMA and CDMA2000. At the same time, IP brings with it some very fundamental changes to an aspect of telecommunications which has altered little since Mr. Strowger first introduced his automatic, electro-mechanical switch-and that is the way in which calls are billed and paid for.

Even now, the vast majority of calls are billed by duration. A call is set up and either voice or data is sent and/or received. At the end of the call, the connection is broken and in most cases, the cost of the complete call, based on a time and distance model, is added to the caller's bill. While charging in this way will remain an option for the foreseeable future, the introduction of IP, particularly over mobile networks, will allow for a method of billing previously only available to customers on conventional mobile data networks such as Ram and Mobitex: charging by the packet. It could also mean the origin of a bill will not necessarily be the communications network or service provider. IP opens the way for permanent virtual connection to a network and, it would seem, cheaper communications bills. As with every new technology, however, there are challenges to be met before the vision becomes a reality.

Market segments
For instance, Henry Harrison, mobile consultant at Schema, reckons that charging by the packet will not be applicable to every sector of the market. In particular it will not suit the consumer segment, simply because for the average person the notion of a data packet is completely alien. "Paying by the packet is likely to differ between the business and consumer markets," he says. "In the business market, it is something which can be expected to be acceptable to customers-indeed, this is how the existing dedicated data operators such as RAM bill. A pay-for-packet approach is much more difficult to see in the consumer market. Firstly, consumers do not have much idea what a packet is. Secondly they do not have much control over how many packets they are sent-when you click on a link on the Web, you have no idea how much data is sitting behind it and thus how much you will be charged for it." With regard to Voice over IP, he adds, the opportunities for tariff arbitrage are likely to be much less than in the fixed world and it is quite likely that voice and video services will be carried separately over the air interface rather than within a generic data stream.

Fiona Fulton, network product strategist at billing specialist Kingston SCL, is a little more circumspect however. She says: "The whole issue of how to bill for packet data is an open question, although the most obvious answer is by data type and content. Realistically, the whole thing will become a lot easier with the availability of WAP (Wireless Application Protocol). Voice over IP is a different proposition and while it may be possible to introduce it by the back door and not tell anyone that is the way it is being carried, any attempt to impose a distance/time billing model would kill it stone dead. The idea that consumers could pay providers for the content, who in turn pay for the transmission of packets could become very popular as technology advances. "

Harrison agrees that one option is to move to a model where consumers pay for content, meaning that it is the content provider who pays for the packets required to deliver the service. This, he says, makes more sense since providers are the ones who control how much data is required to deliver the content. He warns however that this has fairly major implications for service and user interface design. Shlomo Baleli, president of R&D and corporate services division at the customer care and billing specialist Amdocs, reckons the split will be greater than that. He says: "In the consumer market, billing in a packet environment will take into consideration factors of volume, type of data, bandwidth, quality of services and content. There will be different billing charges based on value-added services, so that, for example, browsing, e-commerce and downloading videos will all be billed at different rates."

Interconnect market
He adds: "In the interconnect market, supporting invoicing between telecom carriers, we expect the more traditional parameters [billing by volume and type of data] to dominate." Indeed, according to XXACT, a US-based customer care and billing developer, since there is no 'single' application (like voice) in a packet network, the billing criterion has to be service- or application-specific. It could be time-on-line (minutes, peak/off-peak, time of day, day of week etc.), volume, distance, ToS (Type of Service) or QoS (Quality of Service). In some cases, says the company, it may even be all of these.

So how are providers going to get it right by the time 3G hits the streets? According to Schema and several others involved in the European market, it is GPRS on 2G networks, which will be the key test not only of successful billing approaches, but also of market demand for high-speed data services a point that could well apply to IS 95-B and 1xRTT in the cdmaOne domain. Harrison believes the move to packet data is likely to cause more headaches than anything else, although he qualifies this, saying: "What might still be an issue is the introduction of variable data rates for real-time multimedia services; many [GSM] operators will not implement HSCSD and even those which do may not offer variable rates, but use it simply to offer a single 'high-speed' data rate," he says. Fulton agrees, but claims that existing systems hold at least part of the answer. "The move to billing for 3G services has the potential to be hugely painful, although it can be made a lot easier by using current billing systems as a platform for future evolution. However, the move to GPRS will cause much more upheaval than the move from enhanced 2G to 3G. It's a much greater step." She also points out, however, that once GSM operators have gained experience from 2G systems, the basics won't change for 3G. She explains: "I've yet to see a 'killer' application for 3G that can't be offered over 2G technology, so operators will be able to test all manner of new services on 2G and migrate the most successful ones to 3G.

By then they should have a good idea of how to bill for them. The services may be slightly different and hopefully faster, but the basics won't change. Billing systems sit on top of these services and relatively minor changes won't need to be reflected in the billing software."

In the UK, Vodafone holds a very advantageous position in terms of experience in billing for packet data, as it owns the UK-wide Paknet service, which has been running for several years. It has gained valuable experience in charging by volume of data in this hybrid circuit-switched/packet switched context and through Vodafone Packet Radio Service. Even so, it is still talking in terms of using HSCSD as a test bed for services offered under GPRS and UMTS.

With the possibility of billing so many different services, in so many different ways and at different rates, the next question has to be whether it will be more advantageous to implement billing platforms for separate services or to try to combine them in a standard format. Amdocs' Baleli suggests that the World Wide Web might hold the key to unified customer care and billing systems. "In a multi-service environment, Internet-based customer service makes billing easier, but this will only occur if telcos ensure that the various lines of business and value-added content services are standardized to support uniform usage data records and a uniform interface for customers, " he argues. "Migration to a standard e-commerce environment and widespread use of XML technology will enable billing systems to handle multiple services and to adhere to uniform customer service standards." This is a position held by others, including XXACT, which says that ideally every multi-service carrier wants one and only one billing platform in order to bill for all services with maximum operational efficiency.

Of course, having all the right services and billing systems in place will be of little use if there is little interest from customers and, at present, there is no guarantee of take-up. With this sort of pressure, the only way for 3G operators to respond, according to the experts, is to build a business infrastructure that gives the flexibility for designing and executing different pricing models, in near real-time. They need, as XXACT says, to "design for change". Again, Baleli is in agreement: "In order for operators to be competitive and profitable, they must be able to roll out trials of new services, products and price plans quickly and at low cost. That requires flexibility in acquisition and formatting, pricing, provisioning and billing. This flexibility can only be achieved by an advanced, integrated billing system that can fully support a convergent environment." He continues: "The trend is toward the convergence of data, voice and value-added services. Operators must however keep costs down on individual transactions; there will be a dramatic rise in transactions in which telcos deliver value-added services that carry extremely low profit margins for them. Micro-payments for the value-added services put pressure on the operator to reduce costs to achieve profitability. Scalability is a key issue; operators need to verify that their customer care and billing system can cope with huge, rapidly expanding volumes. In addition to scalability, the other key issue they face is complexity. While the volume of data on the network will increase exponentially, the number of ratable events won't increase at the same pace. The real challenge will be the complexity of billing, due to convergence issues and the emergence of new billing parameters."

XXACT, however, sees the differences between billing for voice and data in a different light. The company believes that, given the numerous studies concluding there is no longer a question of whether or not data will overtake the voice traffic, the only debate is when.

Challenge
The scale of the challenge is best exemplified by some sobering figures from the UMTS Forum. It says the world market for terrestrial mobile services (including multimedia) will be 426 million users by 2000, rising to 940 million by 2005 and to more than 1.7 billion users by 2010. Of these, 190 million will be in North America by 2005, rising to 220 million by 2010. Asia-Pacific will account for 400 million mobile users by 2005, rising to 850 million by 2010. In Western Europe, the Forum predicts there will be 200 million users by 2005, rising to 260 million by 2010. Some 32 million of these will be mobile multimedia users in 2005, rising to 90 million by 2010. It adds that the total Western European mobile market will be worth ECU104 billion ($107.8 billion) per year in 2005, representing total traffic levels of 6,300 million Mb/month. The mobile multimedia segment of this Western European market will be worth ECU24 billion ($24.9 billion) per year in 2005, representing total traffic levels of 3,800 million Mb/month and terminal revenues in 2005 will add a further ECU10 billion per year to this European market value.

However, a report on Unified Telecoms Services to be published shortly by Schema predicts that the data explosion will not take place in quite the same way with mobile as with fixed. "Firstly," says Harrison, "the need to manage capacity and margins means that massive volume discounting for data, as seen in the fixed world, is unlikely to be supportable. Secondly, fixed will continue to offer higher capabilities than mobile as cable modems and ADSL are rolled out. That said, there is a massive potential opportunity which could be realized, and we believe that one of the key enablers to making this opportunity happen is to get the right billing mechanisms in place."