LTE roaming needs IPX, Diameter

Ajay Joseph, CTO, iBasisiBasis CTO Ajay Joseph says IPX is a must for LTE roaming and interconnect and a Diameter hub is essential for rebuilding the roaming partner community as quickly and cost-effectively as possible 
LTE Insights: Why is it taking operators, which have launched LTE, so long to deploy LTE roaming?
Ajay Joseph: In general, there are two main obstacles causing delays in the deployment of LTE roaming today: timing and technical hurdles. In terms of timing, the progression most operators are going through begins with deploying LTE in. The next step is for operators to establish roaming agreements with other mobile networks within their home countries. After that they will look to establish LTE roaming agreements with a few operators in their primary correspondent countries, assuming those operators are ready to do so.
Only then will the broader scope of full international LTE roaming be addressed in a large-scale way. Of course, this progression is playing out at a different pace country-by-country, region-by-region.
In terms of the technical hurdles, in addition to the frequency fragmentation issue that will be addressed by the handset community soon, the all-IP nature of LTE is forcing a change in signaling protocols from SS7 to Diameter and SIP. That implies that the signaling interconnect infrastructure needs to be rebuilt, and physical interconnects between roaming partners need to be re-forged from scratch, using a relatively new protocol like Diameter.
In addition, the national Diameter infrastructure needs to be ready with the right Diameter Routing Agent (DRA) topology to support that interconnect. And it takes two to tango - the existing roaming agreements are only useable when both parties are technically ready for LTE roaming.
Last but not least, if the two parties have different IPX providers, they need to establish an IPX peering interconnect, which brings the normal demands and challenges of a new interconnect. That is why iBasis focuses on an end-to-end IPX roaming solution for LTE with wide reach, open peering, valuable services like hosted DRAs and a comprehensive platform for multiple services, including VoLTE.
When will we see widespread commercial agreements for LTE roaming?
The first LTE roaming corridors are being created this year, and next year we will begin to see the real take-up of commercial LTE roaming traffic as a critical mass of operators and their IPX carriers will be ready for LTE roaming.
iBasis has been at the forefront, preparing our IPX for LTE roaming and peering. So, from the perspective of "the man-in-the-middle", iBasis is ready to support widespread commercial adoption of LTE roaming right now. In fact, our hub, or exchange model, which we call the LTE Signaling eXchange will accelerate the spread of international LTE roaming by helping operators deliver that capability, as well as other new services, to their subscribers faster and more efficiently. We believe the pace will accelerate over the next 12 to 24 months as more operators complete their domestic deployments and home-country roaming agreements.
Is it now common for LTE roaming destinations to be connected via a Diameter hub?
Yes. The first few LTE roaming corridors were made with direct interconnects, but to be scalable and rebuild the roaming partner community as quickly and cost-effectively as possible, a Diameter hub is essential. The demand for LTE and LTE roaming is such that operators need to scale as soon as possible to satisfy that demand. The hubbing or exchange model is the fastest and most efficient way to overcome the interoperability and interconnect issues that arise from interconnecting hundreds of roaming partners.
What impact will IPX have?
IPX is a must for LTE roaming and interconnect due to the IP nature of LTE and the required quality differentiation that LTE voice, signaling, data and messaging requires from the underlying IP infrastructure. Technically speaking it is possible, for example, to run Diameter traffic over the GRX on a small scale for trials or the first few roamers.
But the GSMA IPX model is the right solution for meeting the varying quality of service demands across voice, video, data, signaling and messaging. As the complexity of the network increases and roaming margins come under pressure, operators will benefit from an IPX partnership with a vendor that can evolve all services, risk-free, to the new IP infrastructure.
What are the key challenges in launching VoLTE?
There are commercial and technical challenges with launching VoLTE, both domestically and internationally. Commercially, the focus of operators on domestic LTE data rollouts has reduced the focus on HD voice and VoLTE adoption and thus customer demand. In developed countries, where smartphone penetration is high, adoption will be quick. However, in the developing world, we will need to go through a cycle of equipment upgrades before the consumer is VoLTE-ready.
On the wholesale side, although existing agreements can be leveraged, the opportunity for local breakout and home routing add a level of richness to the inter-carrier relationship that did not exist before. Once LTE data networks are well established, we expect to see a huge push with VoLTE and HD voice in the next 12-24 months. Technically on the wholesale interconnect side, the entire TDM infrastructure has to be replaced by IPX and SIP-IMS signaling, which encompasses a large capital and operational investment.  
What are the main drivers behind VoLTE?
The main drivers are network efficiency and enhanced customer experience. The ability to pass voice and data over a single channel will enable the farming of 3G frequencies allowing mobile operators to leverage their scarce bandwidth in the most optimal way.
Customers will experience increased battery life, lower device cost due to a single radio and higher voice quality. Customers will experience the lower noise floor and greater fidelity of HD voice provided by G.722.2 (AMR-WB). This will once again give mobile operators the ability to differentiate their services for enterprise and consumer markets.
Do you think telcos need to run VoLTE on IPX?
In short, yes, but it depends on your definition of IPX and your commercial and operational expectations. The GSMA clearly lays out IPX transport for VoLTE as an all-IP infrastructure that is private, secure, and offers class of service and near real-time quality reporting and monitoring. In my view, all of these features are strategically important to differentiate mobile operator services from other services running over the public internet. While the public internet will eventually evolve to provide the robustness and quality required for real-time communications, it isn't there yet, globally.

This article originally published in Telecom Asia's LTE Insights e-report