10 SA1 to SA5, the two RAN groups, and the plenaries above them

10.1 Why the deciding groups have the smaller numbers

The map: fourteen groups, 309 meetings and 374 names showed the whole table and one fact from it does the work of this chapter: the groups that decide what IMS is have far fewer documents than the groups that write down how it behaves on the wire.

That is not a measure of importance. Stage 1 says what the system must do, stage 2 says how it is put together, stage 3 says what the bytes are. A requirement written in one line can produce years of stage 3 writing, and a document count shows the writing and never the line.

So read this half of the map for what each group owns rather than for how much it filed.

10.2 SA2, the architecture group

SA2 carries 429 IMS documents over 30 meetings, S2-107 to S2-175, with its most recent stored document dated 2026-06-02 SA2.

It owns TS 23.228, "IP Multimedia Subsystem (IMS); Stage 2" — how the system is put together. The register stores a version of it for every release from Rel-5 at 5.15.0 to Rel-20 at 20.0.0, and 92 IMS documents change it TS 23.228.

The acronyms the register names against SA2 are a good short history of the subject: the WebRTC work, emergency calls over the packet core, the first enhanced-IMS-to-5GC study and the work item that implemented it, and the Rel-20 architecture study SA2.

The first of the two 5G core studies is SA2's; its second phase was CT's, and both are laid out in Releases 15 to 17: IMS moves onto the 5G core.

10.3 SA4, the media group

SA4 carries 186 IMS documents over 27 meetings, S4-84 to S4-136 SA4. Its subject is what the pictures and the sound actually do.

Four of the five acronyms the register names against it are media work items in this subject's own history:

• Telepresence media, IMS_TELEP_S4 and later IMS_TELEP_EXT, which produced TS 26.223, telepresence media handling and interaction, with 38 IMS documents against it TS 26.223.

• Split rendering, SR_IMS, with 84 documents SR_IMS ^[2] — drawing part of a picture in the network instead of on the device. It produced TS 26.567, "Split rendering over IMS", which the register holds in one release only, Rel-19 TS 26.567.

The last of those shows how a subject reaches IMS from outside it. The work item says where the same capability had already been done:

A capability is built once somewhere else, and then a work item opens to put it inside IMS. That is the pattern behind most of SA4's row.

• Machine learning inside a call, AIML_IMS-MED, "Media aspects for AI/ML in IMS services", a Rel-20 item with 32 documents AIML_IMS-MED ^[3].

SA4's acronyms run from the Rel-13 telepresence media work to a Rel-20 machine learning item, and its newest stored document is dated 2026-05-21 SA4. Whatever a call is being asked to carry next, this is the group whose row it lands in.

10.4 SA3, security, and SA1, requirements

SA3 carries 96 IMS documents over 10 meetings, S3-100-e to S3-128 SA3. Two acronyms are named against it. The first is the security assurance work SCAS_IMS, which produced TS 33.226, security assurance for IMS, with 48 IMS documents TS 33.226.

The second is FS_IMSRE_SEC, the Rel-20 study on security aspects for IMS resiliency, and it needs both of its facts written or neither.

Its own name in the work plan begins with the word "Deleted", and it also has an approved study item document — SP-251237 from SA WG3 at plenary SP-109 — and 16 meeting documents FS_IMSRE_SEC SP-251237.

Either of those two facts on its own reads as a conclusion. Together they read as what they are: a record that says two things and does not reconcile them.

SA1 carries 88 IMS documents over 7 meetings, S1-80 to S1-105 SA1. It owns TS 22.228, the stage 1 service requirements for IMS, with 31 IMS documents against it TS 22.228.

Seven meetings for the group that says what the system must do, against CT1's 66 for the group that writes the bytes CT1: one subject, two very different amounts of paper, and the point at the top of this chapter in one comparison.

The three acronyms named against SA1 are the two real-time-services items of Releases 15 to 17: IMS moves onto the 5G core and one much later row: IMSDCDataOff, "PS Data Off for IMS Data Channel Service", a Rel-19 item with 9 documents IMSDCDataOff.

SA1's first stored document is dated 2017 and its last 2024 SA1, and those three acronyms are what stands between them.

10.5 SA5, charging and management

SA5 carries 72 IMS documents over 7 meetings, S5-100 to S5-123 SA5.

The acronyms named against it are all charging or management items and they reach back to Rel-6 and Rel-8: IMS charging, real-time tariff information, roaming charging, and charging for IMS over the 5G system CH8_IMS 5GS_Ph1_IMSCH.

The newest stored SA5 document in this subject is dated 2019-01-25 SA5. That is a fact about this store and about no group's intentions: the register carries no status field for a work item and nothing that says whether a group has finished with a subject.

10.6 RAN5, conformance testing, and RAN2's single row

RAN5 carries 147 IMS documents over 14 meetings, R5-74 to R5-109 RAN5. Conformance testing is the work of proving that a device behaves the way the specification says it does.

It owns TS 34.229-1, the protocol conformance specification for the IMS call control protocol, with 72 IMS documents against it TS 34.229-1. Two of the work items that changed it define test cases for the IMS data channel.

Its four acronyms carry two oddities worth knowing.

• IMS_CONV_IP_COM-UEConTest is marked Rel-5 in the work plan and has 59 documents in a store that begins in 2015 IMS_CONV_IP_COM-UEConTest. A row's release is what the plan says the row is for; it never dates anything.

• IMSProtoc17_dataCH-UEContest is the one acronym in this whole subject whose name begins with the word STOPPED IMSProtoc17_dataCH-UEContest ^[4]. Like "Deleted", that word is typed into the title by hand and is not a status field.

RAN2 has exactly one IMS document, at meeting R2-92, tagged with the work item IMS-CODEC RAN2.

10.7 The three plenaries

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

Plenary	Documents	Meetings	First and last
CT	221	44	CP-67, CP-112
SA	34	17	SP-68, SP-112
RAN	25	13	RP-75, RP-110

Every work item and study in this subject was opened by a document filed at one of those three. The CT plenary's 221 documents over 44 meetings is the shape of a body approving protocol work meeting after meeting; the SA plenary's 34 documents over 17 meetings is a thinner trail across the same stretch of years.

10.7.1 What a plenary week actually looks like

The recent plenary entries carry a registration count, and it is worth putting beside the IMS document count.

Meeting	Place	Days	Registered	IMS documents
CP-110	Baltimore	2025-12-08 to 2025-12-09	277	10
CP-111	Fukuoka	2026-03-09 to 2026-03-10	297	2
CP-112	Singapore	2026-06-08 to 2026-06-09	282	5
SP-112	Singapore	2026-06-09 to 2026-06-12	568	3

Two or three days, a few hundred people, and a handful of IMS papers on a very long agenda. IMS is one subject among many at a plenary, and the numbers say so plainly. SP-108 at Prague, with 488 registered, has no IMS document at all SP-108.

10.8 One document each, and what that is worth

Every group in this chapter can be summed up by the document it holds the pen on, which is the fastest way to know where to go with a question.

Group	The document it owns here	IMS documents against it
SA1	TS 22.228 service requirements, stage 1	31
SA2	TS 23.228 the system, stage 2	92
SA3	TS 33.226 security assurance	48
SA4	TS 26.223 telepresence media handling	38
SA4	TS 26.567 split rendering over IMS	73
RAN5	TS 34.229-1 protocol conformance testing	72

Read the last column with care. It counts IMS documents naming that document as the one they change, so it measures how much this subject wrote about it and not how important it is.

TS 22.228 has the smallest count in the table and holds the service requirements from the user's and the operator's point of view ^[5] — the requirements the rest of the work exists to meet.

The RAN plenary's own row says the same thing about its half of the map: 25 documents, and the three acronyms named against it are all conformance test items RAN. RAN's part in this subject is almost entirely testing that devices do what CT1 wrote down.

10.9 The letters groups actually sent

45 IMS documents in this record are of the type "LS out" — a liaison, which is a letter one group sends another ^[6].

The register gives an entry of its own to fifteen of them: the ones whose outcome word is approved or agreed, which is the record of a letter really sent rather than merely drafted C4-244485.

Four of those fifteen are one exchange, and following it is the best available picture of how a decision travels between groups.

• C4-244485, "FS_IMS_RES outcome and future work plan", from China Telecom at meeting C4-125 C4-244485 — CT4 telling other groups what its restoration study concluded and what it thinks should follow.

• C1-246921, a reply from Nokia at CT1#152 C1-246921.

• S2-2413007, a reply from SA WG2 at S2-166 S2-2413007.

• C3-250657, a reply from Ericsson at CT3#139 C3-250657.

One group asks, three answer. What any of those four letters actually said is not on this machine — the register carries the number, the title, the sender and the meeting, and no text.

10.10 Three things this half of the map is silent about

• Whether a group is still working on IMS. SA5's newest stored IMS document is from 2019 and SA2's from 2026 SA5 SA2. Neither is a status: the register carries no field that says a group has finished, and no date for any release.

• What a liaison said. Fifteen letters have an entry apiece, each with a number, a title, a sender and a meeting, and no text C1-246921.

• Why anything was decided. The stores record the outcome word a document was given and never the reasoning behind it, so no sentence here may say a group chose one thing because of another.

10.11 Where this meets the rest of the course

The four CT groups that write the stage 3 documents these groups decide on are CT1, CT3, CT4 and CT6: where most of it happens. The whole table these two chapters take apart is The map: fourteen groups, 309 meetings and 374 names.

For what SA2 and SA4 were doing in the releases where their counts are highest, read Releases 15 to 17: IMS moves onto the 5G core and Releases 10 to 14: the browser, telepresence and the middle years.