HDVoice Summit: A Beginning

May 25, 2009
Michael Graves

polycom_ip650_256 This past week I had the pleasure of attending The HD Communication Summit hosted by Jeff Pulver in New York City. The summit was a one day event intended to rally a number of players in the telecom world around moving beyond the PSTN with the rollout of wideband telephony services.

The event was the first of its kind, and drew a group of over 100 attendees most of whom were telecom industry insiders. For me it was exciting and enlightening to be in the presence of such a group of knowledgeable people.

The day was marked by a number of presentations from various parties and a handful of panels. There’s simply too much to convey in a single post so I’ll start with some of my more general impressions of the issues discussed.

There are a number of different market segments to be considered; CableCo/MSO, Mobile and Enterprise. Each faces its own set of technical challenges the result of which is that each prescribes a different formula in its solution.

For example, the mobile/wireless industry is extremely bandwidth constrained, so they focus on spectral efficiency in order to wring the maximum performance out of their network. In contrast, larger enterprises have security concerns but may not care so much about the details of the codecs involved.

The matter of codec selection is in itself multi-faceted, and often devolves into a debate with religious overtones. Some people have very strongly held views. There are concerns about royalty cost, CPU requirement, latency, bandwidth efficiency and protection from legal issues around intellectual property concerns. It can be something of a quagmire, yet there was a consensus in two areas:

Manufacturers cannot practically build a vast suite of codecs into phones, only a limited subset of the available codecs
While less than ideal, some form of transcoding is going to be required

It seems very possible that the mobile folks will have a different set of codec standards than the wireline and enterprise folks, with transcoding common as calls traverse the gap between wired and wireless carriers.

The CableCos may be a better position than anyone else to rollout wideband in a big way to consumers. Cable Labs has been working on specs for a set-top box (eMTA) that includes a DECT CATiq base radio. When such a box is in the home the customer only needs a DECT CATiq compatible cordless handset to make calls via the CableCo. Since G.722 is part of the CATiq spec that customer would be able to have wideband calls to other customers of said CableCo.

This presumes that certain back-end network upgrades have been implemented since wideband calling is only possible when the calls transit IP end-to-end. Still, representatives from Time-Warner indicated that they could be in a position to roll this out very quickly when there was a competitive reason to make the move.

The problem is that they don’t see a business case for making that move at present. They don’t see customer demand. This was a widely held opinion, but it was also agreed that people would want wideband if they knew it was possible. Like the transition from back & white to color TV, when one carrier makes the move the others will be forced to act quickly to stay competitive.

From an enterprise perspective wideband is already in place as result of the natural migration from older PBXs to IP-PBXs. This transition has been ongoing for some time, and rapidly accelerating. Most IP-PBXs are widband capable, if not so configured by default. There was a general call for starting to connect up these islands of wideband capability, and the assertion that the growing popularity of SIP trunking was a means to that end.

The mobile industry has established some technology standards, for example 3GPP defined AMR-WB as their codec of choice, but lacks the handsets or back-end infrastructure to deploy wideband immediately. Like the other segments wireless networks need to migrate to an all-IP back-end infrastructure to make wideband possible.

Of all the market segments the mobile area is the one where the pace of innovation is quickest as the handsets turnover every few years. It’s also a growing segment, whereas wireline business is shrinking at an astonishing rate. These factors could mean that widespread implementation of wideband could happen first in the mobile space. Competitive pressure, or a move by an especially innovative carrier* is all that’s required.

There was some assertion that the availability of wideband might be able to revive the ailing land line business, but little evidence that this would actually be the case. It seems more likely that the mobile carriers will eventually leverage improved call quality to further dominate the residential market.

Then, as if from out of the blue we have Skype proposing their new SILK codec for broad use, and asking everyone to consider computer based telephony as an entirely new market segment, complete with its own issues and concerns. While both Skype-For-Asterisk and Skype-For-SIP were topics of discussion it’s clear that Skype sees itself initially as a provider of trunking services using narrowband technologies. Beyond that their vision is less clear, or perhaps it’s more accurate to characterise it as “less clearly stated.”

One area of genuine consensus was that wideband voice suffers a complete lack of public visibility. To remedy that issue Mr Pulver intends to see the formation of an HDVoice marketing group. This will be an organization allowing all the various players to establish HD Voice certification and an icon to indicate this on product packaging. This will be much like the long established and very successful USB certification program.

In conversation between sessions conference organizer Dan Behringer noted that the weekly VoIP Users Conference call, which uses the ZipDX wideband conference service, is at present the only regular, publicly accessible example of wideband telephony that he has found. ZipDX was one of the events sponsors.

From a single experimental conference to profile wideband telephony last November we have evolved our weekly conference to use a wideband bridge in parallel with the traditional Talkshoe bridge. More and more conference regulars join in wideband via a SIP URI. On a couple of recent calls we’ve actually had more attendees on the wideband bridge than the G.711 bridge! In a very grass-roots way we’re actively creating evangelists to spread the word about wideband voice.

As a result of the summit I’m also hopeful that we can get a few interesting guests for future VUC calls.

Clearly there were a lot of different opinions in the room. There was much discussion, and even some debate, but in the end there seems to be some genuine hope that the industry can rally itself around wideband call quality.

I for one certainly hope that this happens. Beyond that hope, I intend to carry on as I have been, actually using wideband telephony in my daily life. And helping anyone who cares to join me in that endevour.

I remain: sip:mgraves@mstvp.onsip.com