No Jitter is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Charting the SIP Journey

SIP trunking offers significant cost savings, scalability and flexibility benefits. Yet despite significant service provider investments and enterprise interest, many IT departments have not begun the migration from TDM to SIP. The good news is they can address the technical and organizational obstacles delaying deployment through effective planning and a step-by-step approach enabled by testing and quality assurance.

What are the options? A "do nothing" strategy is not really viable: The PSTN (meaning, TDM) network comprises aging, legacy infrastructure, managed by a workforce nearing retirement age. The TDM-based PSTN will inevitably disappear, by FCC edict if not by market forces, and at some point carriers will start charging a premium for TDM as they phase out and discontinue services.

An all-in, leap of faith commitment to SIP is not the only alternative, however. We've observed successful implementations in which enterprises gradually move toward a hybrid model, moving traffic to SIP in stages to ease the transition. This incremental approach can be particularly effective in addressing the concerns of telephony technical staffers, who are often loathe to make changes in their external-facing connectivity.

How & Where to Start
In one example, a healthcare provider was transitioning a network of critical facilities and numbers to SIP trunking. Concerned about reliability and resiliency issues and reluctant to move too quickly, the company started with non-critical traffic that it could move as a test case. By building experience and establishing a comfort level within the organization, the company was able to gradually move more critical traffic. The transition approach included failover strategies that allowed for routing critical call types around the IP network in real time in the event of an interruption of the company's data network.

In terms of identifying the starting point of a SIP migration, a customer-facing contact center can be -- perhaps counterintuitively -- the lowest-risk option. Because contact center architectures comprise multiple carriers and both toll and toll-free numbers, a company can easily move traffic between SIP and traditional TDM terminations, using either carrier tools or by changes at the Service Management System (SMS/800) layer. This allows a shift in traffic to SIP for testing and then a quick return to TDM if issues arise, and back again to SIP. This ability to build, monitor, and tweak is in fact a key part of the SIP value proposition.

Implementation timeframes can vary, but the transition can be done quickly given proper planning and preparation. In another example, a client seeking to reduce costs began a SIP migration and faced a peak season later in the year, from September through December. During this peak season, an excess of 10,000 toll-free numbers would be in operation, with 10,000 to 20,000 concurrent calls taking place. The SIP project started in January, and we installed, tested, and converted the trunks in less than nine months, in time for the peak season.

Understanding Architectural Choices
SIP trunking can follow either a centralized or distributed structure. A centralized approach consolidates disparate sites into a central data center to deliver efficiencies in architecture and sizing. Pooling and aggregation can allow an enterprise to buy fewer concurrent channels and to support bursting and least-cost routing. A centralized approach is also easier to manage and better suited to 911 calls.

Many of these benefits are lost under a distributed approach to SIP trunking, which typically only makes sense if the WAN provider also offers SIP. Otherwise, moving to SIP requires creating a separate network. In most instances we see enterprises moving to a centralized model with some distribution for resiliency.

From a carrier perspective, not all SIP trunking is created equal. Significant differences exist in packaging, pricing, and capabilities, making an apples-to-apples comparison difficult. As a result, assessing offerings against requirements is imperative. For example, some carriers allow for aggregation of concurrent SIP sessions at an enterprise or centralized session termination level with bursting at individual locations. While this fee is generally higher than alternatives that don't include localized bursting, the more expensive option may yield a better ROI if an enterprise's traffic profile lends itself to this approach.

An optimal SIP trunking strategy requires a clear understanding of concurrent session and bandwidth requirements, including peaks and valleys in demand. To identify the sweet spot where capacity meets demand, many enterprises rely on the Erlang B formula, an algorithm that utilizes busy hour call volume information and desired blocking percentages to determine concurrent trunk or session requirements.

Codec selection is another critical technology consideration, as different compression standards can impact quality. For example, typical compression standards minimize bandwidth consumption while trying to preserve the sound quality of a human voice; however, even relatively minor losses in sound fidelity coupled with seemingly insignificant network impairments can lead to issues impacting overall call quality. Further, many voice applications require the use of uncompressed codecs -- speech recognition applications and music on hold are examples. The SIP provider must have the capability to switch codecs mid call as needed when calls transition to different endpoints or applications with differing requirements and capabilities. Because the traditional TDM network employs one standard exclusively, the telecom team managing the transition may not be aware of these subtleties when moving to SIP. If the telecom team is not prepared, it risks learning things the hard way.

Staffing for SIP
The staffing and skills impact of the migration can be significant, as TDM requirements and protocols are very different from VoIP and, particularly, SIP -- and that can be intimidating. People need appropriate training as the underlying infrastructure is the data network, of which voice has traditionally not been a part.

While migrating to SIP presents some technical and organizational challenges, experience shows that they can be overcome through an incremental, phased approach that mitigates risk and accounts for the appropriate staff training and tools needed to be successful.

Mark Allen is a director with Alsbridge, a consulting and advisory services firm.