4 Who does the work: fourteen groups, and a core-network shape

4.1 Nobody owns this subject

3GPP splits a subject by rung, not by topic. Stage 1 is what the system must do, stage 2 is how it is put together, stage 3 is the bytes on the wire, and testing is its own world again. Each rung has its own group, its own meeting series and its own document numbers.

So a subject arrives at each group in turn, and a big subject arrives at a lot of them. Fourteen groups have IMS documents on the local record, and 309 separate meetings have had one on the table. ^[2]

That number is worth holding on to before any of the others. There is no single group to ask about IMS and no single meeting where it is decided.

4.2 The map

The counts below are documents carrying one of this subject's work item acronyms, filed at that group's own meetings. The meeting range is where the store's labelling reaches, not where the group's work began.

Group	What it does here	Documents	Meetings on the record
CT1 CT1	mobility, call control and session management; owns TS 24.229	2169	66, C1-90 to C1-162
CT4 CT4	core network protocols; both restoration studies	746	48
CT3 CT3	interworking with other networks; owns TS 29.163	719	47
SA2 SA2	the architecture; owns TS 23.228	429	30, S2-107 to S2-175
SA4 SA4	the media — telepresence, split rendering, AI in a call	186	27
RAN5 RAN5	conformance testing of the device	147	14
SA3 SA3	security	96	10
SA1 SA1	the requirements	88	7
SA5 SA5	charging and management	72	7
CT6 CT6	smart card aspects	6	2
RAN2 RAN2	one document	1	1
CT plenary CT	where CT work is opened and signed off	221	44
SA plenary SA	where SA work is opened and signed off	34	17
RAN plenary RAN	where the RAN test work is opened	25	13

A plenary is the meeting of a whole technical body — SA, CT, RAN — that approves work. A working group is where the text is actually written.

4.3 What the shape says

Look down the document column and the subject describes itself. The three largest counts belong to CT working groups: CT1 with 2169, CT4 with 746 and CT3 with 719. CT1 CT4 CT3

CT1 owns TS 24.229, the SIP and SDP profile, and that one document is changed by more IMS documents than any other in this register. TS 24.229

The other two CT groups divide the outside world between them. CT4 writes the core network protocols, and both of the restoration studies — what the system does when a part of it fails — are its work. CT4

CT3 handles talking to other networks, and owns TS 29.163, the rules for reaching circuit-switched networks. TS 29.163

SA2, with 429 documents over 30 meetings, is the architecture group and owns TS 23.228. SA2 TS 23.228

The split between the groups shows in one piece of work in particular. The first study on putting IMS on the 5G core, FS_eIMS5G, is SA2's work; the second phase, FS_eIMS5G2, is the CT groups'. FS_eIMS5G FS_eIMS5G2

The radio side is thin by comparison, and what there is of it is testing rather than radio design. RAN5 does conformance testing of the device and has 147 documents; the RAN plenary, where that test work is opened, has 25. RAN5 TS 34.229-1

At the other small end, CT6 has 6 documents over 2 meetings, and SA1 — the group that wrote the founding requirements of The founding: Release 5 and the ring of work around it — has 88. CT6 SA1

That last one is the clearest sign of what each rung costs in paper. Deciding what the system must do took 88 documents here. Writing the bytes it sends took 2169.

4.4 One work item, one row per group

The reason a subject spreads like this is written into the work items themselves. A work item description hands out the job group by group, and the work plan then carries one row per group.

IMS_IBCF, a Rel-9 item on the Ix interface of the Interconnection Border Control Function, has three rows and names CT3 and CT4 on them. IMS_IBCF

That is the whole machinery in one sentence — a named piece of work, a named group, a named rung of the ladder.

It also shows what the counts on this page cannot see. Not one document in this record carries the acronym IMS_IBCF, so a work item split across two groups adds nothing at all to either group's number. IMS_IBCF

4.5 The plenaries carry the decisions, not the work

The plenaries carry far less paper than the working groups do. CT has 221 IMS documents, SA has 34 and RAN has 25. CT SA RAN

That is what you would expect from what a plenary is for. A work item is opened by one paper and signed off by another; the hundreds of change requests in between belong to the working group.

4.6 The busiest room there has been

One meeting stands out. C1-112 — CT1's 112th meeting, in West Palm Beach, Florida, from 2018-08-20 to 2018-08-24 — handled 206 IMS documents, more than any other single meeting in this record. ^[3] C1-112

Release 15 is the release that put IMS on the 5G core, and its work item row carries 476 documents. The record does not say why that particular week in 2018 was the peak, though, and this course does not guess. 5GS_Ph1-IMSo5G

4.7 Who writes the documents

374 different names appear as the writer of an IMS document. ^[4]

As the documents spell them, Ericsson filed 602 of them ^[5], Nokia and Nokia Shanghai Bell 258 ^[6], Huawei 173 ^[7] and China Mobile 167. ^[8]

4.8 Where this goes next

The counts on this page are counts of paper: how many documents a group put on a table. They say nothing about who argued for what, because no store here records that.

They also say nothing about the first ten years of the work, because the document record does not reach back that far. The founding: Release 5 and the ring of work around it says where that floor is.

What all this paper actually changes — which documents are written, and which one absorbs the most of it — is the next part of the course.