P2T Opens its Options

The hardest part may be over for those whose last 18 months were spent building the consensus that resulted in what is now known as the Open Mobile Alliance's Push-to-talk Over Cellular Candidate Enabler 1.0, but the work is far from over, and the boost it will provide to the marketplace could still be a year away.

POC 1.0 — the open specification developed by working groups within the OMA — holds the promise of interoperability for push to talk (P2T), widely considered the primary impediment to widespread adoption.

P2T is a service that to date is far from open. Since Nextel first introduced the service in September 1996, it has dominated the market. Nextel has enjoyed all the benefits of being first-to-market, holding strong patents and being proprietary, running its Direct Connect service over its Motorola-built, packet-based iDEN network for most of the last decade. However, time has run out on its enviable position, as within the last two years, carriers such as Verizon Wireless, Cingular Wireless, Sprint, Alltel, U.K.-based Orange and others have begun to launch P2T services on a wider range of network types.

Still, there has been no cross-network interoperability to allow users of different carriers' P2T services to speak to one another, the way they can with traditional voice services. That kind of closed network environment likely has limited the ability of competitive P2T offerings to gain a foothold in some market niches, especially the corporate enterprise market, in which companies have shown a preference to purchases services and devices from multiple carriers. Service providers and vendors alike have been waiting for an industry wide consensus on interoperability to provide the much-needed boost. Even Nextel, now part of Sprint after the recent completion of their merger, should be looking forward to interoperability as much as the next operator.

Members of the OMA that participated in the development of POC 1.0, and analyst Iain Gillott, president of iGillottResearch, characterize the OMA's effort to reach that consensus on interoperability as a difficult but focused process that ultimately accomplished exactly what it set out to do.

The standards, or specification development, process is not always a smooth one. Companies come to the table with their technology biases and their own expectations for what the standards should include. However, the process itself tends to generalize these predilections to provide better interoperability across all technologies.

“The OMA process has worked well to accomplish a good consensus and has industry wide [acceptance] as an enabler that will work over many technologies,” said Craig Rhoades, POC chair for the OMA.

The process has resulted in a specification that supports basic one-to-one POC sessions, one-to-many POC sessions, instant personal alerts and group advertisement. Personal alerts allow callers to inform others of their wish to communicate and request a call back. Group advertisement informs other users about existing POC groups.

The POC Candidate Enabler 1.0 will work with other OMA-defined enablers to expand its capabilities. The OMA XML Document Management Enabler and the OMA Presence Enabler allow for group list management and the integration of presence and availability information.

Most of all, by implementing solutions and services built according to the spec, operators bring P2T to a level of flexibility required for any solution to succeed in the mainstream market.

“[Flexibility] is extremely important to us because [the service] will be peer-to-peer, and success in the marketplace comes only when consumers using different networks can talk to each other,” said Ileana Leuca, director of service optimization at Cingular Wireless.

As great and necessary as these features and capabilities are, they are good, so far, only on paper. They are, however, being solidified every month through various test events designed to put the specification into practice and identify any shortcomings, inefficiencies or incompatibilities.

Two OMA-sponsored events have already taken place since POC 1.0 was released, in an effort to test the spec and work out any kinks. The first was a panel discussion the day before the release in San Diego that featured experts from Cingular Wireless, Ericsson, Nokia, Siemens, Sonim Technologies and Vodafone. The second event took place in late July. It was a test fest sponsored by Vodafone in Milan, Italy.

Now, the next three months will be consumed with vendor testing, tweaking and refining of the spec, while operators huddle to dream of what they would like to see in POC Version 2. There is a test fest scheduled for the middle of this month in Seoul, Korea, hosted by SK Telecom, and one in Mid-November in Düsseldorf, Germany, sponsored by Vodafone.

While the early events focused on testing interfaces and connectivity, these later events will be testing functionality, refining the spec as it progresses through at least five different stages of usage. They validate the work put into the spec in the early going, when the OMA working groups spent much time hashing out their differences to produce what was then a workable spec, if not actually a tested standard. At that point, working group members conducted a consistency review, after which they presented the spec to all OMA members for review and approval. It is then given candidate status, which is where POC 1.0 is today. Despite the fact that testing of the standard only recently began, some companies may be well down the path of developing their own products based on the standard.

The OMA typically subjects its candidate specs to a minimum of three test events such as those listed above. It is here where companies — primarily vendors prove — interoperability between different client/server implementations and make sure people are interpreting the spec properly. Provided all goes well, the spec will again be brought in front of the OMA membership for — as the U.S. Senate would say — its up or down vote.

“We hope that by year end, POC 1.0 will have achieved that level of maturity,” OMA's Rhoades said.

Continuing on the OMA's theme of building consensus, there appears to be consensus that the testing so far has gone well.

“In general, the test was successful in that a lot of test cases were run, some incompatibilities were found — though no major ones — and it showed we managed to achieve interoperability with all the servers and with the clients, even if there are a few issues we have to sort out,” said Joakim Wiklund, chief technology officer of Sonim Technologies.

Cingular's Leuca saw success in meeting the main objectives of the test fest, which to her were developing a way to debug the protocols and to consolidate the end-to-end functionality of the services.

“We are bringing in protocols that have been defined outside of OMA, and this is the only way to verify that end-to-end systems work properly. That we have so many companies involved and that testing is every other month shows that vendors are dedicated to this feature,” Leuca said.

So far, the operators themselves — other than Nextel — haven't seemed all that interested in pushing P2T. After a lot of fanfare at their service launches, Verizon Wireless and Sprint have gone rather quiet on the subject.

“They haven't done that well with their services,” said Iain Gillott, president of iGillottResearch. “One of the problems they have had is that Nextel's patents are pretty strong, and they have had to be creative in getting around them.”

Sprint has found a way to get around that pretty handily by buying the company, but the others will be counting on the OMA to bring interoperability to the marketplace.

“Operators have all developed their own homegrown solutions in order to get to market quickly, but the OMA is right that the industry needs some standardization and interoperability,” Gillott said. “Then everyone benefits,”.

As for POC 1.0 being the answer, Gillott said, it's a compromise. “Will [version] 1.1 be better? Yes. Will 1.2 be better? Yes. But you have to start somewhere, and any spec is better than no spec if it gets people deploying it, using it and marketing it,” he said.

The industry is currently “blazing the trail” for P2T services, Gillott said. In his analogy, “Someone walks into a jungle and works bloody hard using a machete to clear a trail. Someone comes behind the first person and works fairly hard chopping down the trees and branches the first person missed. The third person down the trail skips through and remarks on the nice path and the sunny day.”

“Any new service is like that. Someone has to blaze the trail to develop the market. And that's what Nextel did,” Gillott said. Now, the OMA effort is clearing that path better, so that the service experience to come can be a smooth one.

For those still sitting on the fence waiting for something beyond POC 1.0, they should remember that the fourth guy down the path gets eaten by the lions.

While it appears as if the industry as a whole has been sitting back waiting for an open, interoperable spec such as the OMA's POC, some vendors have been far from idle. Sonim has been working toward this day of interoperable P2T since 1999. It is on its third generation of product and, through the latest OMA test fests, has shown that it has working solutions for OMA POC.

“At least at the protocol level,” Wiklund said. “We had our client as well as our presence and group management servers in Milan and tested all the functionality. Now we are moving that into a server that is hardened.”

He said that the protocols and functionality have been proved, and they won't change, meaning vendors can begin building hardened products.

One question that arises during open, standards-based implementations is whether enough room is left for product differentiation. Wiklund said that in the case of the OMA POC, there is.

“The standard defines the protocol and the functionality, but it doesn't define characteristics. You can always build a server that works better, faster or has better voice quality. You can add additional services like we did with our voice mail box,” he said. “And speed to market is an important factor, too.”

Wiklund said Sonim will have OMA product ready for market rollout in October.

The OMA's Rhoades said that a number of vendors have publicly stated they will have product ready by the end of 2005 or early 2006. However, from a forum perspective, he can't say when that product will begin to impact the customer. Rhoades did say that from the level of participation in the OMA, and in particular on the development of this spec, it is clear a lot of people believe in the technology.

Gillott, however, is in the business of predicting when technology will impact the market. He projects that — coinciding with the launch of OMA products in 2006 — there will be a sharp rise in the number of U.S. P2T subscribers and a marked difference in the number of new subscribers who do business with operators offering interconnection with subscribers of other operators.

The number of P2T subscribers for operators offering interconnection will triple by 2009 and will generate slightly over $4 billion.

The market won't reach these numbers on interoperability alone, Gillott said. In addition to interoperability, P2T will need a consistent marketing message.

“Right now, it is seen as a pretty blue collar service, but there are lots of markets for this, including white collar work groups and the family,” Gillott said. “If there is one standard that [operators] can put enough consistent marketing around, it will help the market a lot.”

Cingular's Leuca thinks the enterprise market for P2T will continue to be the early market driver and agrees with Gillott on the marketing aspect. “[We] need both education and good performance to make sure it becomes popular. But it will still be some time as it was with SMS. That took quite a few years to take off,” Leuca said.

While POC 1.0 provides for the absolutely necessary magic of interoperability and works with other enablers such as presence and document management to enable an end-to-end service, some operators are already looking forward to POC 2.0.

In fact, working groups already have met and are in the process of defining the second set of requirements. With basic interoperability complete, operators will be working with an eye on new functionality and enhanced interoperability.

Leuca would like to see some interoperability with other services like SMS and the ability for voice chatting. She would like to see a POC Box — a device to store conversations for future reference as well as the storage of associated information such as the duration of a session, the type of media used, the number of callers, etc.

Another feature, with an eye on the enterprise market, would be the dispatch function, also called a POC fleet service. This will be similar to a one-to-many session configuration, except that individuals can respond separately.

“Based on input we are getting from operators and vendors, Version 2 is getting pretty creative in the number of things you will be able to do in a push-to-talk environment,” Rhoades said.

Fred Harrison, board chair of the interoperability group at the OMA, added that, “Vendors are looking forward to push-to-talk being like SMS where you can exchange messages between various access technologies.”

By this time next year, we should know if the public has been looking forward to the same thing.