11 How a radio feature travels from a question to a rule

11.1 The journey, in one paragraph

Somebody thinks the radio should do something it does not do. They write a study item description and put it to the plenary. If it is accepted, a working group studies it and writes a study report, in public, over several meetings.

If the study says the thing is worth doing, a work item description turns it into drafting work. The drafting produces change requests against existing specifications, or a new specification. The plenary approves the change requests in bundles. Then it is a rule.

Every one of those steps is a document type with a name and a count in the local database, and this chapter walks them in order.

11.2 Step one: the study item description

737 study item descriptions carry a RAN document number ^[1]. Each one is a proposal put to the plenary saying: this is worth looking at, here is the scope, here is who would do it.

Not every one is accepted. RP-261197, a proposal to survey the regulations that apply to satellite operation in terrestrial bands, is recorded in the database as "not pursued".

That is all the record says. The status of a document is approved, revised, noted, withdrawn or not pursued, and it never says why. This course does not guess at reasons, and neither should you when reading a status column.

11.3 What gets opened

3230 separate pieces of work were opened by a RAN plenary document ^[2]. 217 of them are studies ^[3].

So the great majority of what the plenary opens is not a study at all. It is work: drafting, with a deliverable and a deadline.

By release, the last three break down as 335 for Release 18 ^[4], 195 for Release 19 ^[5] and 100 for Release 20 ^[6].

The newest is the smallest because it has only just started.

11.4 Step two: the study report, written in public

A study produces a study report, and the report is drafted over several meetings with every intermediate version tabled as a document. The database holds 2684 draft study reports ^[7] and 23061 proposed changes to draft study reports ^[8].

The ratio is the interesting part: roughly nine proposed changes for every draft tabled. A study report is not written by one person and reviewed; it is assembled out of contributions, meeting by meeting.

The 6G study is at exactly that stage now. At RAN#112 the plenary agreed version 0.4.1 of TR 38.914, tabled as RP-260870.

That single agreed draft is everything this course can tell you about 6G. There is no date, no scope and no content in the data behind it — see What the record does not tell you.

11.5 Step three: the work item description

When a study is finished, or when nobody thought a study was needed, a work item description turns the idea into drafting work. 3627 of those exist ^[9].

RP-260923 is one: a new work item description from Ericsson, approved at RAN#112, opening conformance work on the gateway device function for mission critical communication.

Notice what that example shows. A work item can be opened purely to produce test cases, in which case its output is RAN5 documents and nothing else. Not every work item changes a requirement.

11.6 Step four: the change request

Once a specification is published, it changes only by change request. 190421 RAN documents are change requests ^[10], and another 40530 are drafts of change requests ^[11].

243160 RAN documents in total name the specification they change ^[12] — more than the change requests themselves, because drafts, packs and revisions all point at a document too.

And yet the commonest RAN document is none of these. 257341 are marked simply "discussion" ^[13], which is more than all the change requests together. A RAN meeting is mostly argument, and the change request is what the argument produces when it ends.

11.7 Step five: the pack, and the plenary

The plenary does not approve change requests one at a time. 7012 documents are change request packs ^[14] — a collection of what a working group agreed, put to the plenary as one item.

That is why the plenary can dispose of a quarter's drafting in four days. It is deciding on bundles that the working groups have already argued to a conclusion.

11.8 The paperwork that holds it together

Three kinds of document exist purely so that the process can be followed.

Status reports. Every working group reports to the plenary at every plenary. RP-260832 is RAN1's status report at RAN#112 and RP-260834 is RAN2's. Separately, 7249 work item status reports are on record ^[15] — one per piece of work rather than one per group.

Minutes. A working group writes its own. R5-261705 is RAN5's record of its previous meeting, filed by the secretariat at the next one. The plenary's own record is written by the 3GPP secretariat: RP-260831 is the draft report of RAN#111, tabled at RAN#112.

Agendas. The plenary's agenda is written by its chair — RP-260830 at RAN#112 — and each working group has the equivalent one level down, such as R1-2603480 for RAN1#125.

Those four documents, taken together, are the fastest way into a meeting you did not attend: the agenda says what was meant to happen, the status reports say where each group was, and the report says what was decided.

11.9 When one group needs something from another

Groups that need something from each other write a liaison statement — a letter from one group to another. 20742 outgoing letters carry a RAN document number ^[16].

R1-2603888 is one, drafted by CATT in RAN1, about how two-sided machine-learning models are paired. R2-2602996 is a draft reply on the same question from RAN2, at 3GPPRAN2#134 in the same week — one subject, two groups, two letters.

That the two groups meet in the same week is not a coincidence, and what it means for how quickly a question can be settled is The meeting rhythm — plenary, working-group week and bis.

11.10 The whole path, as document types

Step	Document type	RAN documents of that type
propose a study	SID new	737 [1]
draft the study report	draft TR	2684 [7]
change the draft report	pCR	23061 [8]
open real work	WID new	3627 [9]
draft a change	draftCR	40530 [11]
propose the change	CR	190421 [10]
approve a bundle	CR pack	7012 [14]
ask another group	LS out	20742 [16]
report progress	WI status report	7249 [15]
argue about any of it	discussion	257341 [13]

Every figure in that table is a row count in the local meeting database on 4 August 2026, taken from the type field, for documents with a RAN number.

11.11 Where this meets the rest of the course

Which documents the change requests land on is The documents that attract the most change, and what that says. When each of these steps happens in the calendar is The meeting rhythm — plenary, working-group week and bis.

What the record cannot tell you about any of it — who decided, why, and when a release closed — is What the record does not tell you.