A Telephony Revolution

Quoting from the O'Riely  book "Asterisk - The Future of Telephony" 

 An incredible revolution is under way. It has been a long time in coming, but now that it has started, there will be no stopping it. It is taking place in an area of technology that has lapsed embarrassingly far behind every other industry that calls itself hightech. The industry is telecommunications, and the revolution is being fueled by an open source Private Branch eXchange (PBX) called Asterisk™.

 Telecommunications is arguably the last major electronics industry that has remained untouched by the open source revolution. Major telecommunications manufacturers still build ridiculously expensive, incompatible systems, running complicated, ancient code on impressively engineered yet obsolete hardware.*

   In selecting a phone system for my church several years ago, the most important features I sought were an open standard for the desk sets the system used, a system that ran on industry standard PC hardware, and a software interface for end users to interact with their voice mail and modify their call-handling settings. That search led me to recommend a product from Altigen Communications that used standard analog phones available from a variety of sources and met all the other requirements. The purchase and implementation of that system also led to the relationships that eventually became Higher Ground Technologies.

 From the onset, Altigen also supported VoIP (Voice over Internet Protocol) technology. We used it in several situations to solve problems for which there was no other good solution. As VoIP gained popularity, the advantages I most often heard touted were "it will save you money", and "it will just plug into, and integrate with, your computer network". I found both of those arguments somewhat flawed. Many organizations were comparing their old $.20 - $.50 per minute long distance plans with the pricing being offered by VoIP trunk providers. What they failed to take into account was the nickel-a-minute (and less) plans being offered by the new generation of completive local exchange carriers (CLECs) created by the AT&T breakup. Moreover, these companies were offering very attractive rates on T1 and PRI circuits bundled with thousands of minutes of free long distance service. Add to that the fact that quality of VoIP service over the public internet is just now reaching a level I would describe as "pretty good" and I suggest that the "savings" promised by VoIP over the last few years was somewhat overstated.

 OK, what about the "integration" of voice and data systems provided by VoIP? Well, if your idea of integration was that the packets traveled over the same wire, you were probably well satisfied with the result.  I have a bit different view on what integration of these systems requires.  First of all, there must be an intuitive Windows-based GUI interfaces for end-users  and system administrators.  From that interface, end-users should be able to manage voice mail, setup call rules and inbound voice calls should  link to contacts in Outlook (or other contact manager applications)  to provide caller names, and, optionally, screen pop that contact's record on the first ring.  A user should be able to dial a number by simply clicking on a name on their computer screen and administrators and supervisors should be able to monitor the system and call queues in real-time from their desktop PC.  At that level, most of the first generation of pure VoIP systems fell short of the integration I had in mind.

 There is another consideration in the same-wire-integration concept: packet latency. Transferring a big file from the file server to your PC or running a data- base query and having a packet or two arrive a few seconds late may be a bit of an annoyance - if you even notice it at all - but let a voice packet arrive late and the sound quality of the voice on the other end starts to suffer in a very noticeable fashion. The solution, of course, is to guarantee that voice packets have the highest priority in moving across your network. One method is to isolate the voice network from the data - and where that is possible, it is a good idea - but it does sort of defeat the same-wire concept.    Another solution is to use a network protocol called Quality of Service (QoS) or Class of Service (CoS) whereby packets are "tagged" to indicate that they have a higher priority than "untagged" packets. Voice traffic is then "tagged" by the phone itself and the switches and routers in the network honor that tagging by giving that traffic "head of the line" priority. Switches and routers that have this feature are just now becoming reasonably priced.

 Finally, the most important criteria in selecting any technology is open standards between components. The opposite of open in this context is "expensive and proprietary". What do I mean by that? Expensive  is subjective and in the eye of the check-writer. Let me explain proprietary in this manner: ask your phone system provider to name five other manufacturers desk sets that will work with his system, and to name five other system providers who’s system will work with his desk sets. If he can't do that, show him the door. Would you buy a PC that only worked with XYZ brand servers or a server that only would talk to XYZ brand PCs? Of course not!

So what about that revolution I started talking about? Well, the fact is that the allure of "we now have you hooked" proprietary systems is such a ingrained culture in the PBX (Private Branch eXchange) market, that the VoIP systems offered by the legacy vendors still suffered the same flaw. Yes, it will run over the same wire as your computer network, but don't expect to go to the VoIP store and buy a new desk set. And furthermore, they will decide what features to offer and at what point you will be required to buy new hardware to enjoy those features. Even the current version of the Altigen system I recommended to my church 8 years ago now only works with their own brand of IP phones.

 The revolution began in 1992 when an Electrical Engineering student at Auburn University (War Eagle) decided to create a phone system in software to run on the Linux operating system. He named the software Asterisk, and much to his credit, he placed that software in the public domain and invited other programmers to improve it and create additional features. His name is Mark Spencer and his work in open source telecommunications software led him to found the Digium Company based in Huntsville, Alabama. Literally thousands of software developers have contributed to the code base and dozens of companies offer enhancements and add-on features. Today, there are more phone systems running some flavor of Asterisk than any other worldwide. Considering that fact in light of the deep-pockets and huge margins enjoyed by legacy PBX manufacturers and, yes, I would call that a revolution.

 Why am I telling you all this? Last year one of our large church customers gave me a briefing on their planned multiphase office relocation and new student building. This customer was running an Altigen analog system that was installed 6 years ago. Their question to me was: how do we make the office move go smoothly for our staff and what kind of voice service do we provide to our new, detached and somewhat distant, student building. (The detached and distant part is significant in the analog/copper wire world). Furthermore, they were approaching the maximum extensions supported by their Altigen system and, consequently, I started looking for new options. VoIP and fiber network cable makes the detached/distance problem go away, but this customer was already accustomed to a GUI interface for end users and administrators , reasonably priced desk sets from multiple vendors, and all the rest of the features I believe in so strongly.  My search led me to a company who markets a fully tested and supported implementation of the Asterisk PBX and that had recently become a part of the Digium organization. That company is Switchvox and we migrated our office to their system on the day before Christmas and have been amazed at how well their product works. As a matter of fact, when we show it to customers, the word we hear most often is "wow". Switchvox recommends desk sets made by Polycom but their system works fine with products from GrandStream, Snom, Cisco, D-Link and a host of other hardware and softphone companies. Systems from Fonality, Intutive Voice, Voice Mail Inc, FacetPhone, and hundreds of other Asterisk vendors would run on the same PC hardware and use the same Polycom desk sets as the Switchvox.   I think we can call the "Open Standard" requirement fully complied with.

 What about the user and administrator interface?   That is where Switchvox really shines.  They provide a web-based (Java) interface for both.  At the end-user level, you can setup you voicemail and call handling, including call-rules to route calls based on personal timeframes and call source (direct dial or queue calls), review call reports and create a personal phone book of both internal and external numbers.  End-users may also launch a "switchboard" that displays their phone book with  real-time status of all internal extensions including caller ID, barge-in, monitor and record if permissions are so granted.   From the administrator console (also web based), you can access all system functions, including IVR (Interactive Voice Response) management, extension and queue creation, system reporting and status, and automated backups.    All this runs on an industry standard PC hardware.   

 There are many things to consider before making a  move to a VoIP system.   Network infrastructure is the primary item.   In most situations switches will need to be upgraded and, if SIP trunks or external extensions are required, there are also firewall issues to be resolved. We  anticipate having several Switchvox systems installed by midsummer and will keep you posted on how it is going.  In the meantime, please  to give us a call if you are considering updating your telecommunications system - we will be glad to help.

 *Asterisk: The Future of Telephony, Second Edition
By Jim Van Meggelen, Jared Smith, Leif Madsen
Second Edition  August 2007 
ISBN 10: 0-596-51048-9 | ISBN 13: 9780596510480