When planning a migration strategy from TDM to IP, it’s important to note that service providers migrate for different reasons. Some are migrating as part of a strategy to replace TDM with IP across the board. Others are planning to grow interconnects using IP rather than TDM, but are not looking to close down or stop using existing TDM networks. Because of these different paths, timelines across the industry for migration are not very well defined.

In some regions, operators are still building out their networks using TDM technology. For providers that are on a strong growth path, increasing coverage provides a better return on investment than the optimization, decreased costs and additional services enabled by IP.

Migration will be a long-term exercise – we at Tata Communications expect we’ll still serve some customers on TDM for years to come.

-Claus N., Marketing

One of the architectural features of the Internet is that all packets are treated with equal importance, no matter where they originate or what they contain. However, not all packets are in fact equal to end-users, particularly when it comes to latency. Some, such as VoIP traffic or streaming video, are much more time-sensitive than email or web browsing.

Providers also see that not all packets have the same monetary value, with some customers willing to pay a premium for better than best-effort delivery. One solution to both of these problems would be to define differentiated quality of service depending upon the type of traffic – with latency-sensitive packets receiving priority over others – using the same approach currently in place on many private, IP-based networks.

Prioritizing some packets raises red flags for advocates of “net neutrality,” who worry that such efforts could place ISPs in the position of favoring traffic from preferred providers over others, and potentially crimp the free-wheeling exchange that they see as a crucial feature of the Internet as it exists now.

-Sylvie L., Operations

One of the current challenges with carrying voice traffic over the public Internet is optimizing path and latency to the destination. The current Internet architecture provides a high level of assurance that data will eventually reach its destination, but a low level of control over the path it will take.

For the type of data traditionally carried over the Internet, such as email or browser traffic, this best-effort delivery serves well. However, for voice traffic, high latencies can impact call quality. Managing latency still remains a challenge that has to be managed when using the public Internet to transport IP voice traffic.

One way for providers to address latency is to ensure that they are widely connected. Many ISPs are undertaking efforts to expand their connectivity reach with direct connections to as many networks as possible. Emerging markets, which are often underserved and not widely connected when it comes to Internet access, are an area where VoIP providers using the public Internet need particularly to focus.

A wide range of direct connections enables both efficient and redundant routes, meaning traffic can be routed around areas of congestion and high latency.

-Sylvie L., Operations

One of the most important lessons we learned from the trials is the importance of a checklist to use to prepare for a new IP interconnection. Within the TDM world, the definitions that need to be established and the parameters that need to be exchanged are well defined, but in this trial, we didn’t know where we’d encounter configuration issues we’d have to address.

We completed the trials on time within several months, end-to-end, but with the knowledge of the process we have now, we expect we can reduce time frames to several weeks, just from knowing what information to share. So building out that checklist of parameters was one of the very positive results to come out of the trial.

For engineering teams, having opportunities to prepare procedures when there’s not commercial pressure and no financial impacts to mistakes is an invaluable opportunity. With the experience we gained in this dry run, we’re looking forward to supporting our customers’ migration strategies from TDM to IP.

-Claus N., Marketing

We started with the GSMA recommendations to plan our test case, focusing on the items we knew we could execute, and that made the most sense in light of future commercial rollouts.

During the course of the trials we solved several interesting technical challenges, but we were encouraged to discover that these were configuration differences rather than interoperability issues. Converting from the ANSI specifications used by our US partner to the ITU variant used by the European operator was initially a challenge, which we addressed once we had a full picture of each environment.

We also found the test environment contained different implementations of SIP I, so we modified or eliminated some parameters that were causing problems. Attention to detail was an important component of our success – for instance, resolving a difference in the way two SIP environments treated a plus sign from the E.164 specification in the URI field was one of the fixes we implemented.

We were tremendously pleased with the results we achieved. We showed that interconnections could work on private MPLS networks, and the public Internet as well. We also showed we could effectively monitor and troubleshoot the interconnection to provide the right quality for end customers. Most important for commercial implementations, we also now know we can manage billing off of the interconnect.

-Jim K., Engineering

We recently announced the completion of a trial of the IP eXchange protocol proposed by GSMA and now handled by the IPIA. Over the course of the coming week, we’ll be posting some of the lessons – both business and technical – that our teams learned in the course of the trials.

More information about the proposed IP eXchange protocol is available at the GSMA website.

Tata Communications volunteered for the trials because we saw it as an important opportunity to track and influence the evolution of a key model for the transition to IP – a model that’s going to be used for a good percentage of our customers’ business in the near future. Because we identified strong partners at an early stage of trial, we were able to quickly come up to speed and contribute our successful results to the GSMA working group.

Our active involvement in industry groups such as IPIA (the IP interconnect offshoot of the GSMA) and the i3 Forum has offered tremendous payoffs. With the experience we gained through participation in the trials, we’ve been able to get a head start on developing a commercial offer that will align with the proposed IPX standard.

-Christian Michaud
Senior Vice President, Product Marketing and Business Development, Global Voice Solutions