3 How the work is cut up, and who carries it

3.1 Two kinds of document, and the order they come in

3GPP writes two kinds of paper about a subject. A study report, numbered TR, is homework: a group investigates a question and writes down what it found. A specification, numbered TS, is a rule other people have to follow.

The order is almost always the same. A study item opens the homework and produces a report; a work item opens the rule-writing and changes specifications. A study's short name starts FS_, for feasibility study, and that prefix is the quickest way to tell the two apart in a list.

Each of them is opened by one document that a plenary — the three-monthly meeting above the working groups — approves. The work plan records that document against the row, so the shortest statement of what a piece of work promised is always one paper away.

Not every row has one. The Release 8 row in this subject names no describing document at all ANSS.

Once a specification is published it is never simply edited. It is altered by a change request — one proposed edit, filed as its own numbered paper.

A report still being drafted gets the same treatment under a different name, a proposed change request. That is why the paper trail of a finished feature is so much longer than the feature itself.

3.2 One feature, three or four rows

Radio work is split by what kind of requirement it writes, and each split piece gets its own short name and its own row:

• The core part, name ending -Core, writes the feature itself.

• The performance part, ending -Perf, writes what the equipment then has to achieve.

• The conformance part, ending -UEConTest, writes how a device is tested for it. Those rows sit in RAN5 RAN5.

Above them usually sits a feature row carrying the bare name. So one piece of radio work can be four short names before anybody has written a line of it.

Some short names then appear on more than one row of their own — the Release 17 architecture item for the internet-of-things core network sits on five IoT_SAT_ARCH_EPS.

That is why the subject's 155 short names ^[1] spread across 183 work plan rows ^[2].

Those 155 names are what the documents are filed under. 35306 meeting documents carry one of them ^[3], and 550 separate meetings have had one of those documents on the table ^[4].

3.3 The release spread, and the strays in it

The rows are spread over nine releases, and the shape of that spread is the shape of the subject.

Release	Rows of this subject	What sits there
Rel-8 Rel-8	1	one stage 1 positioning row, no documents
Rel-10 Rel-10	8	two old North American band items and their parts
Rel-12 Rel-12	2	reaching the telephony core over a satellite television return channel
Rel-15 Rel-15	1	the first study on the new radio over satellite
Rel-16 Rel-16	2	the solutions study, and the first requirements study
Rel-17 Rel-17	23	the first rules, on both radios, and the core network behind them
Rel-18 Rel-18	47	enhancements, the first bands, backhaul, security, management
Rel-19 Rel-19	58	phase 3 everywhere, and most of the band work
Rel-20 Rel-20	41	phase 4, and the radio work broken into narrow items

The first three rows of that table are not the same subject as the rest, and it is worth knowing why they are in the list at all.

The Release 8 row is ANSS, a stage 1 row for positioning by satellite navigation ANSS. Navigation is deliberately kept out of this subject, and this row survived the filter only because its own name does not repeat the word. It carries no document.

The Release 10 rows are two old band items for the ancillary terrestrial component of mobile satellite service in North America: one L band L_Band_LTE_ATC_MSS and one at 2 GHz S_Band_LTE_ATC_MSS.

Both name a describing paper RP-101284 RP-110775, and neither carries a single meeting document.

The Release 12 pair is IMSSat, reaching IMS — the part of the core that carries calls — over a satellite television return channel IMSSat, with a stage 1 row beside it IMSSat-SA1. Neither carries a document.

3.4 Eighteen group codes carry it

Each 3GPP group meets on its own and numbers its papers with its own prefix, so counting documents by that prefix says where the work was actually done.

Group	Documents	Its meetings on the record
RAN4 RAN4	7490	34
RAN2 RAN2	6714	44
RAN1 RAN1	6492	47
SA2 SA2	3588	50
RAN5 RAN5	2455	20
CT1 CT1	2164	37
RAN3 RAN3	1725	40
RAN plenary RAN	1273	36
SA5 SA5	751	36
SA1 SA1	711	29
SA3 SA3	551	23
SA6 SA6	402	16
SA4 SA4	272	14
CT3 CT3	202	28
CT4 CT4	180	30
CP, the CT plenary CT	144	23
SP, the SA plenary SA	119	27
CT6 CT6	73	16

Five RAN working groups, six SA working groups, four CT working groups and the three plenaries above them. That is eighteen codes, and the top three are all radio.

RAN4 carries more of these documents than any other group RAN4. It owns the radio requirements, and its register entry names it in the leadership line of the band items — which is where that weight comes from.

Counting meetings tells a different story. SA2 has spent more separate meetings on this subject than any group here, 50 of them SA2, and RAN1 has spent 47 RAN1 against RAN4's 34 RAN4.

3.5 What the plenaries do, and do not do

The three plenary rows are small. A plenary is where work is opened and signed off, not where it is done, and the numbers show it: the RAN plenary carries 1273 of these documents RAN against RAN4's 7490 RAN4.

The register holds the plenary paper that the work plan points at for each row that names one, which is how you get from a short name to the promise behind it in one step.

That is also the practical way to search this subject. Every meeting document carries the short name of the work it belongs to, so the name is the handle: it finds the papers, the plenary document that opened the work, and the list of specifications the work promised to change.

Searching for the word instead does not work. 5663 documents say satellite or GNSS, carry some other work item, and say neither non-terrestrial nor NTN ^[5] — satellite navigation above all, which is a different subject.

One caution about the group table above. The counts are documents filed at that group's own meetings, not a measure of who decided anything, and no store here records which group led a piece of work except in the free text of the work item's own leadership line.

3.6 Where this leads

The two halves of the subject sit in different rows of that table. The radio half is RAN1, RAN2, RAN3, RAN4 and RAN5; the system half is the SA and CT groups.

The radio half, release by release follows the first of them release by release, and Why anyone wanted it holds the stated reasons behind the larger pieces.