There seems to be little question that SIP is the way to go for PSTN access trunking. New UC solutions as well as legacy PBXes are increasingly using SIP as their access. The big question is: which SIP trunking provider is the best for your organization? There are over 200 companies in the US that provide some form of SIP trunking, which one fits your requirements best? To help decide, there are five key questions that should be asked of any SIP trunking provider. These questions cover areas that are vital to the overall operation of the system and should be considered by any organization moving to SIP. Making the right choices based on your current and future needs will assure that the solution fits well and avoids longer term contractual issues.

Question 1: Speed of Installation/Change

How does the SIP service provider manage the initial installation and ongoing change processes? Most SIP service providers choose one of two paths: 1) adding SIP onto an existing TDM infrastructure and methodology or 2) building a system designed from the ground up as a SIP solution. Those based on the former are often tied to legacy operational time frames and options, which can take weeks or months to be installed or changed. A big advantage of a well-implemented SIP solution is that, while there are hardware components, the actual services are on-demand and can be implemented through self-management and rapid implementation. Modern SIP architectures are predominately software-based, with many of the elements operating in virtualized systems. The advantage here is rapid provisioning and operations.

You should test how your potential service provider reacts to your needs and changes. For your initial install, test the SIP providers operating methodology by specifying a relatively short ramp or asking if they can meet a ramp of a few days (even if you actually have more time). Requesting a few trial SIP ports is another way to test their reaction time. Evaluating the longer term ability of your SIP service provider to meet your changing needs can be challenging, but specifying a short initial install period will test their ability to respond in a timely manner. Be wary of service providers who will not respond or who specify traditional telecom time frames that are measured in weeks or months. Getting references from your potential providers of how they have reacted to time-critical needs of existing customers can enable you to understand their ability to react to your needs. If a potential SIP service provider can only quote delivery in weeks or months and cannot show clear examples of rapid reaction to customer needs, they will clearly not be ready to meet your needs as they change.

Question 2: Broad Interoperability with Communications Equipment

Many times, the only interoperability consideration is existing SIP equipment that will be connected to the trunks. Unfortunately, an evaluation methodology based only on existing equipment can potentially be a major issue. If a SIP service provider was chosen based on interoperability with a given VoIP PBX system, a subsequent addition of a different communications platform or an SBC may introduce challenges. While a broad interoperability list is not a guarantee of interoperability with all potential SIP equipment, it is a strong indicator of the flexibility of the service provider to enable new equipment of solutions elements to be deployed. With the rapid changes in business communications and UC, having the flexibility to optimize solution elements and equipment without interoperability concerns is important. Every organization will face rapid changes in both needs and equipment, which makes it critical that they support the widest range of equipment. A broad interoperation list also shows the range of the SIP service provider's user base and the capability to rapidly adapt to market changes and integrate new vendors.

Question 3: Coverage

Geographic coverage capabilities can make or break an SIP solution. Traditional TDM access was limited to physical trunks, therefore, a typical solution was local only, extended over distance through the PSTN. SIP enables the virtualization of both the infrastructure and numbering plans, providing a solution that can include a global range of connections and locations. In the case of SIP, having direct dial numbers in a range of geographies is not only possible, but often comes as part of the overall solution and may be crucial for centralization of both trunking and communications systems.

The coverage footprint of a SIP service provider can be evaluated through the IP connection locations location of SIP trunks and the range of local access number locations. Most SIP trunking providers have a dedicated IP network that moves the real-time traffic off the open Internet to improve quality. A key to service quality is having a relatively close local transition from the service provider's optimized IP network into that local environment. For example, if you have an office in Lithuania, is the SIP trunk connection to your Lithuania office running over the open Internet from the U.S. or somewhere else? Or is it travelling on a dedicated parallel IP infrastructure with higher quality goals and guarantees? In other words, is the Point Of Presence (POP) close enough to the end point? By examining where a SIP service provider has its points of presence, it is possible to assure that your overall services are better managed. In the TDM PSTN world, coverage was not a factor because the underlying network was used - albeit at relatively high cost. For SIP, local presence and short final IP hops are significant contributors to quality.

Another key factor is the ability to have local sales or support service numbers in a multiplicity of geographies. The other side of SIP is using a single SIP trunk to represent your organization around the globe. To do this effectively, the SIP service provider must have a global local presence. This is critical for a variety of reasons - from assuring that your organization is able to "be global and act local," to reducing costs for your customers to interact with you. While your day one deployment may be simple domestic SIP trunks, the ability to use the same infrastructure for local dial numbers or toll free calling outside your home country should be a consideration in your design to support future international expansion plans.

For both of these values, your SIP solution provider should have a national and global footprint. Enabling your solution to take advantage of the virtual nature of SIP often becomes an important part of driving future business value from your investment.

Question 4: Quality of Service

An initial SIP service install is generally easy. The challenge is delivering consistent quality communications over time. In an evaluation of any SIP service provider, it is important to examine how their network is architected and operated to deliver the levels of quality you require. Take time to understand how a vendor's architecture and implementation, as well as their operational model, will impact call quality and availability.

How the service provider peers with both IP and SIP networks is an indication of how well the system is architected. IP peering is the number and locations of how the service providers IP infrastructure connects to other IP networks around the globe. The quality of IP peering is critical, as these are the points where the SIP service provider interfaces with the other IP providers and the open Internet. To maximize quality, a SIP service provider should have a range of IP peering points, generally focused to both the Internet core locations (in major peering points) and the correct local POP connections to the correct IP network providers for each region. The closer these IP peer points are to the end customer locations, the better for end voice quality (generally speaking). When the IP peer points get further away, there is a higher probability that the variability of the open Internet may have random impacts on the quality of service. By choosing from a wide range of IP peer points, a SIP service provider should be able to demonstrate that they will not get in the situation of having traffic back-track and loops - a serious issue that can cause extended latency and quality degradation. As part of the evaluation of a SIP service provider, ask how they will peer for SIP traffic into other providers and for IP traffic into the Internet. To better understand this, define some of your most complex scenarios (involving international calling or using different ISPs at different locations) and discuss how the vendor would handle the traffic.

SIP peering is potentially more important to the overall quality and experience. Implementing proper SIP peering assures that there are a minimal number of hops (for minimal latency), as well as minimal transcoding. As the SIP standard does not require a specific codec or implementation, peering to a limited set of other SIP networks often means that the traffic is actually going through other and/or multiple peering points on the path to the final destination. As these peering points often are implemented with transcoding to match different carrier's standards, the issue of multiple transcoding's can be introduced, resulting in substantially degraded quality of experience. A SIP provider that has optimized direct SIP (or PSTN) interconnections around the globe will dramatically reduce the potential of multi-hop. This is because the SIP service provider will directly peer into a broad range of other SIP networks (and PSTN termination networks). As part of your analysis, have the potential vendors explain how many SIP transitions a typical call that leaves your network has before getting to the other end of the SIP session (or PSTN line). Also discuss how the vendor deals with different encoding standards and transcoding in their network and matching to their partners.

Ultimately, no solution is better than the team behind it. As part of your decision, it is imperative to understand if the provider team has the capability to provide appropriate support. In the past, TDM PSTN and trunking was often provided by large organizations that had huge staffs. In the case of SIP, those extensive staffs and tools often do not translate over into the SIP/IP world. Smaller providers often do not have the staff or tools to provide the 24/7 levels of support which organizations require. As part of the procurement process, evaluating the support infrastructure, staff training, and response methodologies is crucial. As part of your evaluations, discussing how the service provider has responded to past incidents is a good way of evaluating their methodology and systems.

Question 5: Management Tools

Transitioning from a physical TDM infrastructure to a virtualized SIP environment can be a challenge for both the users and the IT team in traditional telephony organizations. In a traditional TDM environment, operations were very limited and tended to be focused on whether a link was up or not. If a TDM link was up, the very structure of the PSTN generally assured both call completion and basic quality. There are several factors that make operations a much greater management challenge, including: the underlying IP network and infrastructure, the packetized VoIP nature of the traffic and corresponding QoS management, and the interaction of all of the operational components that can cause a variety of operational and quality issues.

The management platform provided by your SIP service provider plays a vital role in helping you manage the system on a daily basis. The system should provide you with a simple dashboard to see your traffic, the current quality of service and any issues. This dashboard should be available to both your NOC staff as well as your support team. When there are issues with voice quality, the solution should enable your team to see key performance indicators like latency and lost packets that can contribute to degradation of MOS scores.

In addition to providing activity and other information, the management platform should enable a reasonable level of flexibility in direct configuration by the end user organization. The ability of the platform to automate changes and provisioning will enable you to manage you overall system more effectively. For example, the tools should allow changing the end-point of an inbound phone number from one contact center to another easily to respond to an MPLS or other IP outage at one site. The solution should also allow you to scale up or down the available trunk lines/ports at your locations based business requirements. The capability to automate and simplify the management of capacity and configuration is a major capability of SIP and legacy platforms and business system often do not enable rapid response.

Make the right choice

SIP trunking is a valuable tool in the evolution of modern communications systems. However, picking the right SIP trunking vendor will have a huge impact on the success, cost, and ROI of your communications platform. As part of this decision process, it is best to start with a clean slate, open your acquisition process to arrange of providers, and evaluate based on clear criteria. Transitioning to SIP with a known legacy vendor may seem expeditious, but it may result in a solution that has significant quality and operational issues and does not enable you to take advantage of the virtual nature of SIP to the fullest. To help you in your analysis process, the five areas have been summarized as a checklist that can be downloaded in pdf form.

By Gaetan Brichet, Chief Operations Officer, Voxbone