Long ago, many people viewed VoIP as simply a low-cost, low-quality alternative to POTS. People were prepared to accept uncertainty about connections, audible distortions and conversation-impeding delays in order to talk to people on the other side of the world via the Internet for a tiny fraction of the cost they would have paid their telco service provider.
Today the second wave of VoIP, known as IP telephony, is transforming the network and communications infrastructure of the enterprise. However, the enterprise demands something quite different from those VoIP pioneers. The enterprise has a business to run, customers to keep satisfied and a balance sheet to maintain. Reliable connections, sufficient quality, rapid detection of service-affecting faults and ready identification of root-cause to enable laser-focused remedial action are prerequisites. Typical deployments still use dedicated IP phones that resemble the familiar phones used with POTS for the last hundred years.
Second-generation IP telephony deployments are not rolling out without incident, and enterprises routinely are demanding service level agreements (SLAs) from their hosted service providers or IT departments in terms of metrics that are relevant to their businesses and end users quality of experience (QoE).
QoE depends on both the IP network performance and application-level factors, such as: echo, noise, delay, speech levels and speech waveform distortion. Edge devices are critical to ensuring that application-level performance is satisfactory, and the set of approved measurements can identify the contribution of the edge device to overall performance and fault issues.
Looking forward, what will become of IP edge devices? Will we need a PC, mobile phone and IP phone? The answer is almost certainly no.
Heres why: IP telephony today is marketed principally on the basis of cost savings (lower call costs promise an operational savings for on-net voice calls and voice conferencing). This is only a precursor to the next phase of enterprise communications, which builds on the technological capabilities of IP and wireless systems to provide an integrated and seamless suite of communication, knowledge and collaboration tools, or what we now call unified communications. While cost savings provide the business case for IP telephony, unified communications enables increased productivity through flexibility, mobility, advanced features and video.
Comparing the pros and cons for the three main classes of edge device PC, mobile phone and IP phone leads to a seemingly inevitable conclusion: The mix of the future will be a combination of wireless/ mobile devices and PCs. The mobile device will provide the utility we require when on the move, and the PC will offer the power of full-screen video, high fidelity and easy user interface when we are at our desks.
When we are at our desks, the advantages of the PC are overwhelming. Big is good for screens, keyboards, processing and available software when portability and power consumption are not the determining factors. PC soft clients that can exploit the capabilities of the PC are emerging and, when used with regulated audio peripherals, can offer higher fidelity than POTS. It also is possible to achieve higher resilience to network errors and flexible integration with other applications, including video.
When we are on the move, however, we need mobile devices to provide communications and information. In terms of mobile devices, their small size enables maximum portability but limits screen and keyboard real estate. Power consumption must be low and processing resources are correspondingly limited. High call costs can be controlled somewhat by having the mobile device revert to the campus VoIP network via Wi-Fi/WiMAX when in proximity.
Performance management and the resolution of service-affecting faults on a per-call basis become even more critical across this mix of technologies in which multiple vendors can be expected for a majority of deployments. In addition to monitoring at media gateways and commercial demarcations, such as third-party WAN providers, there is a growing need to incorporate performance measurement capability into the edge devices. This allows user-specific QoE information to be collected for SLA management and also effective diagnosis of any service-affecting issues, helping to minimize the cost of resolution.
In the future, the costly, dedicated IP phone rarely will be used, except perhaps to provide basic telephony to locations without PCs. It will be fascinating to watch how the UC revolution will change a hundred years of telephony history.
Dr. Mike Hollier is a technical and commercial pioneer in perceptual engineering.
He directed BTs research into audio, video and multimedia performance assessment for nearly a decade before leading the incubation of Psytechnics Ltd. in 2000. He left BT to become the companys CEO and since October 2002 has served as its CTO.
|Psytechnics Ltd. www.psytechnics.com|