10 Who CT1 writes to, and what the letters count

10.1 What a liaison statement is

Working groups cannot change each other's documents, so when one needs something from another it writes a letter. A liaison statement is exactly that: a question, an answer, or a warning that something has changed on this side and the other side had better look.

Each letter is filed as a document like any other, with a number and an outcome, which is why they can be counted.

In the twelve months to 2026-08-04, CT1 sent 108 letters out ^[1] and received 135 in ^[2]. Over the whole record it has sent 3002 ^[3] and received 2507 ^[4].

10.2 The four busiest in each direction

Where the outgoing letters went, when they went to one group only:

To	Letters in the year
RAN2	44 [5]
SA2	19 [6]
CT4	8 [7]
SA1	5 [8]

Where the incoming ones came from:

From	Letters in the year
SA2	39 [9]
RAN2	15 [10]
SA3	10 [11]
CT4	7 [12]

Two groups dominate both tables. RAN2 and SA2 are the ones CT1 writes to and hears from most, and CT4 is the busiest of CT1's own sibling groups CT1 — Mobility, call control, session management (WG1_mm-cc-sm_ex-CN1).

The direction reverses between them, which is the most interesting thing in the two tables. CT1 sent 44 letters to RAN2 ^[5] and 19 to SA2 ^[6]; it received 39 from SA2 ^[9] and 15 from RAN2 ^[10].

Two groups appear in one table and not the other. SA1 is fourth among the destinations with 5 letters ^[8]; SA3 is third among the sources with 10 ^[11].

Since both tables are cut at four, that is not evidence of a one-way relationship. It only means each of those two fell below fourth place in the other direction.

There is one more thing to notice about the outgoing table before using it. It counts letters that went to one group only.

A letter addressed to two groups at once is in the total of 108 and in none of the four rows, so the rows and the total are not the same kind of measurement. Adding the four rows together, or subtracting them from the total, counts nothing.

10.3 What the letters said

Nothing in the register. The counts are counts of letters; the text of any of them is not held here in a form this data touched.

So no chapter of this course can say what CT1 asked RAN2, what SA2 answered, or whether anybody agreed. What the numbers support is one narrow statement — who CT1 corresponds with, and how often — and nothing about the content.

10.4 Why those two groups, read from the document

The register does not say what any other group does. The specification, however, shows exactly where TS 24.501 stops and hands off, and reading the two side by side is instructive as long as it is understood for what it is: a reading, not a recorded fact.

The NAS depends on a radio layer it does not specify. When an idle device needs to send its first message, the document is explicit:

Only when the lower layers report success does the device consider its N1 NAS signalling connection established §5.3.1.1. There is a whole subclause on mapping the document's own access categories onto the establishment cause the radio layer expects §4.5.6, which is a seam if anything is.

The NAS depends on an architecture and a set of procedures it does not specify. The overview places the NAS at reference point N1 and sends the reader to TS 23.501 for what that means §4.1. The same overview says the sequences its elementary procedures are assembled into can be found in TS 23.502 §4.1. Neither of those documents is on CT1's own list of 229.

The NAS depends on a security specification it does not own. Primary authentication produces keying material "as specified in 3GPP TS 33.501" §5.4.1.1, and security mode control points at the same document for the rules on running security procedures concurrently §5.4.2.1.

The reference list at the front makes all of those part of this document by reference §2 — which is a strong dependency, and dependencies are exactly what generate letters.

The NAS carries other people's protocols. The third main function the overview names is a transport service: NAS messages carry short messages, positioning protocols, location services, policy containers, roaming information and parameter updates for the device §4.1. Every one of those is specified somewhere else, and every one of them is a reason for somebody else to need something changed here.

There is a fourth seam of a different kind, easy to miss because it looks like an implementation note. When the device works out which access category applies to a request, it then has to map that category onto the establishment cause the radio layer expects §4.5.6.

A whole subclause exists to translate this document's vocabulary into another document's. That is what a boundary looks like when somebody writes it down.

10.5 What that does and does not license you to say

Fair: CT1 corresponds most with RAN2 and SA2, and its own document leans heavily on a radio layer, on two architecture documents and on a security document that it does not own.

Not fair: any statement that a named group owns a named document. The register carries the working group for CT1's own specifications and for no others, so TS 23.501, TS 23.502 and TS 33.501 have no owner in this data at all.

Also not fair: any comparison of size. The register was built for CT1 alone. It holds no specification count, document count or meeting count for CT3, CT4, SA2 or anybody else, so nothing here can say CT1 is the largest, the busiest or the fastest of anything What this record cannot tell you.

10.6 The one asymmetry worth noticing

Over the whole record CT1 has sent more letters than it has received: 3002 out ^[3] against 2507 in ^[4]. Over the last twelve months the balance runs the other way, 135 in ^[2] against 108 out ^[1].

That reversal is real but it has to be read with a caveat that applies to every year-long number in this course: 7181 CT1 documents carry no upload date at all ^[13], so the all-time counts include undated rows and the twelve-month counts cannot.

The two are not measured the same way, and a trend drawn between them is not supported.

10.7 Where this meets the rest of the course

Who the sibling CT groups are, and what the record does and does not say about them, is The neighbours — and how little the record says about them. The work items behind most of this correspondence are The work items that fill a CT1 year.