VoIP and International Call Termination - Peering into the future

The previous article in this series outlined the issues and opportunities associated with terminating international calls to mobile operators and how portability was impacting that. This article shifts to a topic that is related in some ways – what is the role of Peering in international termination?

Firstly, a definition: Peering means many things, but for the sake of this article, I am defining peering as being the direct VoIP interconnect between two service providers where the calls are routed just to the customers of the terminating provider. This involves several things – a VoIP interconnect between the partners and some way of sharing numbering information from the terminating partner to the originating partner, and a way of keeping that up to date as customers come and go. The reason that the second part of this definition is important is that just having the VoIP interconnect on its own is just a technology change – there are no directly assigned TDM ports but it is still just a normal wholesale interconnect between carriers.

The advantage of routing directly to the service provider serving the end user comes in three areas: price, set-up time and quality. A service provider normally has no cost to terminate a call to its own customers. If they are on a physical network, the network will already be in place; and if it is a VoIP connected customer, there are no out-payments to other carriers to reach that line. As a result, it is possible to reach a commercial deal with such a carrier for them to accept calls to their own customers at reduced (or no) cost. This can be reciprocal – i.e. we exchange calls between our own customers at no cost – but there is no reason that it has to be. Call set-up time is reduced because there are no intermediate carriers in the call path (each of which could have been trying other carriers to terminate the call). Finally, quality should be as good as it gets, as there are no intermediate carriers or conversions from one technology to another in the call path.

With those clear benefits, peering looks like a no-brainer. But is the international (or national) world moving to the direct peering of all end service providers with each other? The short answer is no. There is very little evidence of that at present for several reasons. Firstly, an interconnect is not cost free. There are contracts to be agreed, equipment to be provisioned, invoicing and settlement processes to be established, testing and troubleshooting arrangements created. Then a mechanism for exchanging customer numbers needs to be agreed. If customers migrate to this provider on a one by one basis (by porting their numbers from other carriers), then that requires the exchange of individual numbers as the customers ebb and flow. That is fine for two companies, but doing it between 100 or 1,000 on a series of bilateral exchanges is impossible. Finally, how much will a company save? International call charges to fixed lines are already very inexpensive – sub 1c in many cases, and so is the effort worth the reward in financial terms? Most service providers are more worried about the problems of rapidly adding customers and handling growth. The saving of a fraction of a cent is lower down the priority list.

So what is the future of peering? Well, it is here now in the routing plans of wholesale carriers. A wholesale carrier aims to offer a full international termination plan giving the best mix of price and quality, and is continually seeking to optimize its termination approaches. Arbinet, for instance, has established its own internal database of the customer numbers of some of its 1000-plus carrier partners, and queries each call to those countries to first decide if that call is headed for an “on-net” customer before routing the call over a direct path to that service provider and customer. This enables a better price/quality mix to be offered back up the chain to the originating service provider. This “enabled peering” approach gives VoIP service providers the financial and quality benefits of peering without the complexity of setting up those 1,000 relationships themselves. As Arbinet also handles the contracts, invoicing and settlement functions, this provides the best mix between the DIY approach to peering and a pure wholesale offering. IP

Steve Heap is chief technology officer for Arbinet. He can be reached at sheap@arbinet.com.