We've been running packetized voice traffic on data networks for almost five years. In the beginning, proprietary systems digitized voice calls and placed them onto frame relay networks. VoATM (voice over ATM) was typically done through circuit emulation, without compressing the voice and without silence suppression. Voice-over technologies have been standardized, with VoIP (voice over IP) getting the bulk of the attention.
Voice over frame relay and VoATM have continued to expand the capabilities of the enterprise, allowing voice calls from small or branch offices to be easily placed onto a WAN or PSTN (public switched telephone network) via a wide variety of IADs (integrated access devices). In fact, in our Sept. 4 review of ATM IADs, Memotec Communications' CX950 ended the reign of Mariposa Technology, now Marconi Communications, as our favored voice and data transport device (see "Integrated and Ready To Go").
Yet even with the stream of new products and service providers, voice-over technologies have yet to take off. The biggest roadblock? Cost of implementation vs. return on investment. Voice-over technologies were meant to lower the cost of long-distance calls by shunting voice calls to their destinations over WANs instead of the PSTN. As the climate for carriers has grown more competitive, however, the price of long distance has been driven down to pennies per minute, far below what even 64 Kbps of frame relay or ATM bandwidth would cost for the occasional phone call.
In addition, IP phones may have great features, but they cost just as much as or more than digital handsets, making the replacement of an enterprise PBX with VoIP equipment a very expensive proposition. And even if cost were not an issue, there's no compelling reason as yet for businesses to switch to VoIP. The quality isn't any better, and IP phones typically don't offer features that can't be found in traditional digital handsets for PBXes. Even for those companies building out new infrastructures, there hasn't yet been any clear incentive to use IP phones or even regular handsets connected to an IAD or gateway for voice calls. Then again, stranger things have happened.
Take a SIP
On the protocol side, we have been looking forward to the day when SIP (Session Initiation Protocol) finally topples H.323 as the king of the voice-over hill. Expecting to enter the new century with a truckload of products supporting SIP, we are disappointed at the trickle that has actually emerged. SIP has more long-term functionality and growth potential than H.323, but its acceptance will be most likely tied to the overall acceptance of voice-over technologies. SIP will still have its day, we hope in the near future.
But all is not lost in Convergenceville. With MANs (metropolitan area networks) becoming more prevalent and extremely fat pipes (OC-12 or OC-48) finding their way into buildings, plenty of low-cost bandwidth is available for packetized voice. Once a MAN is completed, offices in the wired area can be connected to high-speed backbones for a lower bandwidth cost than when those offices were linked via WAN. In addition, most providers enable customers on MANs to provision their own bandwidth. Companies also no longer need to bring in additional DS-1 lines or change out to DS-3 or higher as their bandwidth requirements grow; instead, they can provision more bandwidth on the fiber to their buildings.
You Should Be In Video
This growing availability of bandwidth in metropolitan areas is also boosting use of video technologies. A few years ago, an Internet user who wanted video delivered to the desktop might have spent hours downloading the video before viewing. RealNetworks changed that, leading the pack of streaming-media-server vendors. Although a few months back we chose RealNetworks' offering as the best server (see "RealNetworks Skins Apple, Microsoft"), its competitors, Apple Computer and Microsoft, are now closing the gap by pushing the need for better quality at higher resolutions.
When it comes to streaming media in the workplace, enterprises are moving beyond providing video to customers via the Internet, instead going straight to the desktops of corporate users. Internal uses of streaming media include enabling employees to watch presentations and receive training sessions from the "comfort" of their offices. Streaming media will also have a big impact in the area of technical support; for example, video help files could be streamed to the desktop to help solve IT problems.
Driving the video rage and keeping bandwidth providers in business are CDNs (content delivery networks). CDNs were developed to rein in the abundance of duplicate data that was going to the four corners of the world from the same point of origin. This clogging of networks with duplicate information spurred ISPs to create separate networks, thus interconnecting dispersed servers and putting video, audio and other media closer to the end user by storing them at multiple points of origin (see "Feeding the Streaming Media Frenzy").
CDNs will continue to grow in size, number and importance as customers demand ever-faster access to rapidly increasing amounts of data. There also will be a greater variety of data types being provided by CDNs: Services providing audio and/or video on demand will mushroom as broadband services to homes and businesses increase. On-demand services will no longer be the domain of satellite providers, as service providers and CDNs begin to invade that space. The first generation of these devices is already in homes across the country. While boxes like the TiVo digital video recorder now store video information locally, this data will eventually be stored in one location, accessible to just about everyone.
Send your comments on this article to Darrin Woods at dwoods@nwc.com.
RSS
Digital Convergence
