Regular readers of this blog probably remember that I'm a fan of Voice over LTE via GAN (VOLGA). For those who don't, have a look here on more details on why I think it has a good chance of becoming THE voice solution for LTE. It's amazing how fast the Volga-Forum is pushing out the specifications. In May, they published the stage 1 specification document, which contains a high level architecture and the requirements. Now only a month later, a first version of the stage 2 specification is available. Stage 2 specifications as per 3GPP contain a detailed architecture description and all procedures required from connecting to the network, originating and terminating calls, doing handovers, etc.
While their speed is incredible, maybe it should not be that surprising, because VOLGA is based on the already existing 3GPP GAN (Generic Access Network, i.e. GSM over Wi-Fi) specification. That's a good thing because that means that VOLGA could thus be developed quite quickly as it's likely that existing products can be modified instead of being designed from scratch. In addition, this should also mean that the first version of the standard is already quite mature as many areas were already verified during implementation and rollout of GAN in current networks.
I did a quick comparison between the two stage 2 specs and as I expected, many parts are very similar. While the GAN stage 2 specification has 126 pages, the current VOLGA stage 2 specification has 87 pages. This is probably because VOLGA is simpler than GAN. There are fewer handover procedures and most of the handover details are part of the 3GPP Single Radio Voice Call Continuity (SR-VCC) specification (for IMS) so they don't have to be included in the VOLGA spec. In addition to fewer handover scenarios, handovers are a bit more simple with LTE from a VOLGA perspective, as the network takes care of it unlike with GAN, where the mobile has to force the network into a handover. Also, there's no need to support the packet switched part of the network which also significantly lowers the complexity.
Well done, I am looking forward to the stage 3 specification which will contain the details on all messages and information elements used.