Mar 20, 2009

Diameter adoption and network geography

Diameter is used everywhere in the NGN core network – it is in the mobile, wireline, WiMax, Cable, from the ASN’s, GGSN’s and BRAS up to Application Servers. There isn’t a network core entity without at least Diameter interfaces, and some of them network functionalities have up to 4 different Diameter interfaces.
There are dozens of Diameter interfaces defined, over 110 interfaces are actually defined for Mobile, Wireline, Cable and WiMax and it seems also that the number is growing rapidly, new interfaces are introduced all the time, either for LTE (dozen of new interfaces), Lawful interception, interoperability, Roaming, IPTV, advanced policy and other needs.

But in reality most of those interfaces are still on the drawing boards or in the R&D labs, and not deployed and used yet in operators networks.
So what we see in the market, what places have the highest Diameter maturity, here is my view, top to bottom
I believe that the highest maturity is around the Charging and Billing systems ( what is called Rf, Ro, Gy, Gz, CCA interfaces)
It could be the connection from the Application to the Charging system, or from the GGSN to the Billing, or other network functionality, but it seems always to involve some charging component.
2nd place is given to the AAA servers and subscribers databases area, either HSS, NASS or other AAA Server, (interfaces like Sh, Dh, Cx, Dx), it could be CSCF connectivity to HSS, HSS to Application Server or one of many other scenarios of connecting to a AAA and subscriber databases using Diameter.
The 3rd place closely behind is taken by the policy and enforcement area. Policy is a spaghetti of Diameter with more than 9 Diameter interfaces (Gx, Rx, Gq, Gq’, Tx,Ty…) and we see almost any combination you can imagine of those policy related Diameter interfaces nd sometimes requests for all of them.

The rest of the network Diameter interfaces are defiantly behind, we see growing number of requests for almost all the 100+ defined interfaces, but they are still tailoring behind and you can find them today mainly in the vendors and operators labs and not in deployed operational networks.
Actually one of the emerging areas is the LTE related Diameter interfaces, although not finalized yet, and bearing a temporary name those Diameter interfaces are in high demand with strong interest by vendors starting to developing and testing LTE related equipment.

Mar 3, 2009

LTE and Diameter

It seems LTE is moving to the manufacturing halls.
The industry has aligned behind it (leaving far behind some potential alternatives such as Mobile WiMax), and in recent weeks we saw some new announcements on first deployments in Tier-1’s in 2010.
What place does Diameter take in LTE architecture ? well it seems much bigger than in the past, even compared to IMS.

There are about 15 new Diameter interfaces (on top of the existing 30 inherited from previous 3GPP versions and that are also used in LTE), and Diameter is getting out of the network core, closer to the edge and to the interconnectivity and roaming connections between operators.
However there are some warning signs – LTE standards are still work in progress, and the same for the LTE Diameter interfaces – so the potential for interoperability, connectivity and vendor lock-in issues is stronger than ever before.
Here in Traffix we supplied a number of network equipment vendors with LTE Diameter interfaces – and surprisingly each one wanted a different version of the still forming LTE (3GPP Rel. 8) specifications.
In some places where the standards and the Diameter interfaces are still too young to do the job, the equipment vendors are using earlier Diameter interfaces – again this wont contribute to connectivity.
So what’s the conclusion – well that the interoperability issues that operators are facing today won’t disappear with LTE, actually I believe they will be even bigger than in the past, especially in the first 2-3 years of LTE deployments.
I believe this might also slow LTE adoption – operators will understand the immaturity and that the “dream” is far from reality, and the related costs and will prefer to wait a bit until LTE will be more mature.
Actually it sounds a bit like IMS all over again (just replace LTE with IMS in the lines above :-) )