Mainstream To 3GPP Release Gap

A cryptic title today but here’s the thought: When looking at the UMTS landscape today, most networks are still 3GPP Release 6, i.e. HSDPA somewhere around 7.2 MBit/s and HSUPA 2 MBit/s. Agreed, there are some early deployments of Release 7 HSPA+ 21 MBit/s HSDPA to be found but these are not yet the norm. In contrast, 3GPP currently works on 3GPP Release 10 and some are already thinking ahead to Release 11. In other words, there’s a gap of 4 release cycles between theory and practice. In number of years, current networks trail the specification by around 5 years as according to the 3GPP website, work on Release 6 was complete in March 2005. So think about that when you hear or read something about features that have been specified lately.

Comments

[David Boettger]

One might argue that it took almost 10 years for the iPhone to justify the existence of 3G. ;)

[Kamil]

When it comes to infrastructure, it's pretty natural that operators are couple of years behind the specs - well, often they have to physically upgrade hardware.

But such a long delay for services, like IMS, is pushing them to being a bit pipe. Some comments from blogs around: "operators are moving at a glacier speed" ;)

[DavidL]

Compared with Rel-5/6, the number of new features in Rel-7/8 is such that it is more difficult for operators and vendors to make a decision, and operators may be reluctant to invest in UMTS beyond Rel-6 because of LTE.

Looking at LTE, this looks a bit better but there is still indeed a 2 release gap.

[Nasula]

Hmm. Interesting though. But Don't agree with DavidL that LTE is a reason for not investing in UMTS. Many of the new features in R7/8 are SW upgrades.

Many of them also give significant benefits to the operator like Dual Cell, CPC, MIMO etc.

I feel that a 2-release delay is understandable because the vendors have to implement the features, operators then have to test and only then deploy. Easily 2 years.

Here's a thought: How many terminals actually support features at R7/R8 levels? Could that have something to do with the longer than 2 year delay between spec and deployment?

[David Boettger]

Vendors don't develop new functionality in a vacuum. If the carriers aren't willing to pay for the new features, then the vendors won't develop them. (Of course, there's push from the vendor side as well, but ultimately the carrier has to be willing to pay.) So for carriers, there is almost certainly a need to rationalize the capital spend, i.e., to continue to develop UMTS or to focus on LTE. [Unless, of course, Huawei is the carrier's LTE vendor and is building the network for free. :)] Note that there are tons of EDGE and even GSM features that will almost certainly never see the light of day for this reason.

[DavidL]

From 2006 there have been roughly 4 times more people attending 3GPP RAN WG LTE sessions compared with UMTS sessions. So it also seems like most vendors decided to deprioritise UMTS enhancements 5 years ago already.

[Nasula]

DavidL: I think that might be a bit too much simplification to look at the number of people in the meetings. Naturally a new and largely undefined technology will require more people to work on it to get it work in general. Thus needs a high priority.

If you look at the advances that have happened in WCDMA Between the same timeperiod (Releases 6 through 9), I wouldn't say "lost interest".

Vendors BTW are currently shipping R8 networks and if you look at the advancements in 3G BTS technologies, RNCs and network architecture from 2006, you'll see the largest advancements so far.

[DavidL]

I meant terminal vendors. For Rel-7/8 features, you need terminals with new modems. Many terminal vendors just stopped attending UMTS sessions, that means some things.

The enhanced networks work with legacy terminals.