|
|
Whitepaper: The evolving role of hardware as a key enabler of open source telephony in the business
2006-09-11 Article Rating:  / 1
Foreword: This whitepaper examines the market potential for hardware add-in cards that interface open telephony systems with the PSTN (public switched telephone network). It also looks at the trends driving businesses to adopt IP telephony systems, and at various popular open source PBXs and other tools.
The paper was prepared by J Arnold & Associates, an independent telecom analyst and marketing consultancy, at the request of Sangoma, which makes a variety of telephone network add-in cards supported under Linux. Enjoy . . . !
Introduction The business telephony market is undergoing radical transformation due to the advent of disruptive technologies such as Internet Protocol and open source software. Traditional PBX systems have dominated the enterprise market, while Key systems have been the primary solution for small and medium sized businesses. IP-based alternatives are now gaining acceptance, and open source-based telephony solutions are opening up new markets for the functionality that previously required an expensive PBX system. Open source telephony solutions are largely software-based but require a hardware component for PSTN connectivity. This whitepaper addresses the drivers behind open source telephony and how open source solutions are evolving. There is a continuing trend for telephony functions to be software-based, but for now, functions such as PSTN connectivity will be hardware-based. Why enterprises are migrating from TDM to IP In the enterprise and small business market, telephony has long been the domain of equipment vendors, primarily for PBX or Key systems. Reliability has been the hallmark of these systems, and businesses have come to rely on them for their voice connections to the outside world, one of their most mission-critical functions. In the TDM (time division multiplexing) world, voice was a dedicated service, and telephony systems were built around this one application. Until recently, phone systems were not integrated with data networks, and there were no alternatives to TDM for voice. While these conditions have given rise to telephony systems that work very well for their intended purpose, businesses have paid a high price. Aside from being costly, these systems were proprietary and closed. Each vendor had their own technology and competing systems were not compatible with each other. Vendors retained full control, leaving customers totally dependent on them for fixes, enhancements and upgrades. In short, acquisition costs were high, support was expensive and hard to find, and feature sets were fixed. The advent of IP (Internet Protocol) has changed the business telephony market on many levels. First and foremost, it offered an alternative to TDM for voice, and with it, IPbased solutions for telephony systems. One of the reasons large enterprises are drawn to IP telephony is the potential efficiency gained from combining the voice and data functions in an organization. All businesses -- both large and small -- recognize the economic benefits of IP telephony, especially in areas such as toll bypass, reduced trunking costs and eliminating MAC costs -- moves, adds and changes. By nature, IP is flexible, and has enabled the development of a wide variety of innovative telephony solutions that are gaining acceptance among businesses. It is now possible for businesses to have the richness of a PBX feature set without paying PBX prices. Legacy vendors have introduced less costly IP PBX systems to help their installed base migrate to IP, as well as bring PBX functionality to a new set of customers they could not previously reach with a TDM-based PBX. Similarly, service providers are now offering IP Centrex as a way for traditional Centrex customers to get more for less. Carriers are also offering IP Centrex on a hosted basis to smaller businesses that could not justify a full PBX system. Getting more features and performance at a lower cost is attractive for any business, and these developments indicate that IP technology has matured to the point where the days of the traditional PBX are now numbered. The rise of open source telephony Most IP-based solutions have been targeted at the PBX market for a number of reasons. The installed base is quite large, and these enterprises have been accustomed to investing significant capital in high performance telephony systems. Eventually, TDM PBX deployments will transition to IP, but this process is expected to take several years. The enterprise PBX market is substantial, especially in terms of the revenue opportunity for PBX vendors. However, there is another substantial portion of the business telephony market that does not use PBX systems. For these businesses the capital investment is not justified, and less expensive TDM-based systems can provide the functionality they need. IP-based alternatives exist for this market, but both vendors and carriers have yet to develop strong channels to educate and support these businesses with IP. These conditions set the stage for open source telephony. Over the past two years, open source software has gained considerable acceptance throughout the enterprise environment and more recently has been applied to telephony. The inherent appeal of open source is lower cost, and with the advent of PC-based PBX solutions, the addressable market opportunity is substantial. There are millions of small businesses that cannot afford a PBX but would certainly desire its feature set. Open source telephony is largely software-based, and is by nature less costly than hardware-based PBX systems. Aside from cost, however, there are several other factors that make open source attractive for business telephony. These are summarized in Table 1 below.
Open source telephony solutions have two primary components -- software and hardware. The most widely deployed software platform is Asterisk, and is the most strongly associated with PC-based telephony solutions for business. In addition to Asterisk, SIPX, FreeSwitch and Yate are notable open source telephony software platforms. As the open source community has become more focused on the telecom market, these solutions have become more complete in their ability to meet the needs of businesses on many fronts:
Figure 1 below illustrates the basic components of an open source telephony system. Figure 1 -- Open source telephony topography (Click to enlarge) How open source telephony is evolving The evolution from DSP-based media gateways to DSP-free processing cards is an important one in the open source continuum, but is really just one of many. On a broader scale, open source is really moving along a path where the starting point for telephony was a system based on hardware-based components. These components were expensive, closed and proprietary. As software expertise developed, and as IP technologies matured, more and more hardware-based functions have become software-based, and telephony solutions have come to support a mix of both. At this time, PSTN connectivity still requires a hardware interface, and the PSTN is still the dominant mode for voice traffic. However, much of the PSTN voice processing is software-based, including signaling, switching, conferencing, voice-mail handling and compression. Only the functions that are very dependent on hardware are done on the voice card itself. These include HDLC handling, voice channelization and sometimes echo cancellation and voice compression. Many aspects of telephony are becoming software-based, and open source is continuing to reshape what is possible in telephony, especially so long as PC processing power continues to increase and costs continue to fall. One example of the mainstream embrace of the software model is Intel's Host Media Processing software. General purpose computer platforms can use Intel's DSP-free software to enable a wide range both voice and video applications such as IVR, unified messaging, video mail and video caller ID. These tools are enabling innovative, lower cost alternatives to hardware-based solutions, and increasingly, businesses will be less dependent on traditional vendors for their telephony and networking solutions. The business IP telephony market has clearly progressed since the days of purpose-built hardware for DSPs and media processing. Vendors such as Dialogic (now Intel), BrookTrout (now Cantata), AudioCodes and NMS developed high performing telephony components that interfaced with the PC only for control functions, all the voice processing being handled on board. Their technologies were proprietary and expensive, being priced generally beyond the reach of PBX customers and being used mostly for Interactive Voice Response and call center applications. PBX systems themselves have followed the same path, and the legacy telecom vendors have long controlled the market. That began to change with the advent of IP PBXs, and the entry of data vendors such as Cisco into the market. The open source movement has further accelerated the transition by introducing software-based PBX systems that bypass the hardware PBX vendors altogether. In terms of components, the market has also seen the rise of alternatives to purpose-built boards in the likes of Sangoma and Digium, who can provide PTSN connectivity using hardware built on general purpose computing platforms. New companies such as Fonality, with its low cost, do-it-yourself PBX model based on open source software, are showing the way towards highly integrated and easy to use PBX systems that never existed before. Open source has been the real catalyst behind this trend, and we feel it is one of the most important indicators as to where IP telephony is headed. Furthermore, we also view voice as the entry point for open source communications platforms. Once IP telephony becomes proven in open source, we anticipate more advanced voice applications such as text to speech and speech recognition. Building on that, video applications and mobility via WiFi and WiMax are logical extensions, all of which can be enabled with open source software. Conclusions Open source has made significant advances in telephony, and its developer community has proven that PC-based solutions can be viable alternatives to existing PBX systems. These platforms are not perfect, but in the right hands, can deliver tremendous value and performance for businesses of all sizes. As open source matures, standards will evolve, and the solutions will have more checks and balances to ensure the reliability and quality that businesses need for larger scale deployment. PC processing power will continue to improve as well, but for the time being, functions such as PSTN connectivity still need to be hardware-based. J Arnold & Associates, an independent telecom consultancy, produced this whitepaper. The contents herein reflect the conclusions drawn based on general research about Sangoma Technologies, as well as interviews with open source experts, and Sangoma customers. Reproduced by LinuxDevices.com with permission.  About the author: Jon Arnold is Principal of J Arnold & Associates, an independent telecom analyst and marketing consultancy with a focus on IP communications. Previously, he was the VoIP Program Leader at Frost & Sullivan, where he was responsible for managing their subscription service for Global VoIP Equipment Markets. He is a regular speaker at VoIP and telecom conferences, including VON, Globalcomm, Internet Telephony, Dow Jones Network Ventures, NYSSA, and the Canadian Institute. In addition, he is the Director of Industry Analyst Relations for the IMS Forum, and is a contributing editor to the IP Telephony Dictionary.Related stories:
|
news feed
