7 Proving it works — the conformance shelf

7.1 Why there is a third row

A radio requirement is a sentence about what a transmitter and a receiver have to do. Somebody then has to be able to put a real device on a bench and find out whether it does it — which measurement, in what conditions, with what allowed error, and what counts as a pass.

Those are different documents written by a different group, and in the work plan they are a different row. Bands, and why RAN4 carries more of this than anybody showed the first two rows of the pattern, -Core and -Perf. This chapter is the third, -UEConTest.

RAN5 owns it. It carries 2455 documents of this subject over 20 of its meetings, and every short name in this register ending -UEConTest names it RAN5.

7.2 The conformance rows, largest first

Five of them are worth walking through: four of the heaviest, and one small one that is the point of the whole chapter.

NR_NTN_solutions_plus_CT-UEConTest is the Release 17 one, and it is the heaviest: 1008 documents over 33 meetings ^[5].

It was planned from 2022-06-15 to 2025-12-12, stands at a completion value of 1, and sits under NR_NTN_solutions in the work plan's own ordering NR_NTN_solutions_plus_CT-UEConTest. Its objective says what a conformance row is for, in one clause:

The quotation stops where the register's stored text stops, at the first item of a list. What is visible is the shape of the sentence: this work exists to enable testing of that work, and it names the other work item as its subject.

LTE_NBIOT_eMTC_NTN_req-UEConTest is next with 553 documents over 19 meetings ^[6].

It is Release 18, it was planned to finish 2024-03-03, it stands at 1, and the work plan names TS 36.521-4 as the document it produces LTE_NBIOT_eMTC_NTN_req-UEConTest.

NR_NTN_enh_plus_CT-UEConTest is the Release 18 NR one: 205 documents over 13 meetings, planned from 2024-09-09 to 2026-03-03, standing at 1 NR_NTN_enh_plus_CT-UEConTest. Its objective is the fullest description of the job anywhere in this register:

Read the list of verbs in that sentence and the size of the shelf stops being a surprise.

One feature means analysing what the requirement says, building an environment to test it in, writing the procedure, writing the messages, working out how much measurement error to allow, writing down which devices the test applies to, and then editing every document that carries any of those.

IoT_NTN_enh_plus_CT1-UEConTest carries 159 documents over 16 meetings, Release 18, planned to 2025-12-12 and standing at 1 IoT_NTN_enh_plus_CT1-UEConTest.

NR_NTN_Ph3-UEConTest is the Release 19 NR one, and it is the small one: 34 documents over three meetings, planned from 2026-03-03 to 2027-03-03 NR_NTN_Ph3-UEConTest. Its completion value is 0.14, a long way short of the 1 its own core and performance rows carry.

7.3 Every conformance row on one page

Sixteen short names in this register end -UEConTest. Here they are, largest first, with what the work plan says about each.

Row	Release	Documents	Completion	Planned finish
NR_NTN_solutions_plus_CT-UEConTest	Rel-17	1008	1	2025-12-12
LTE_NBIOT_eMTC_NTN_req-UEConTest	Rel-18	553	1	2024-03-03
LTE_NBIOT_eMTC_NTN_plus_EPS-UEConTest	Rel-17	330	1	2024-03-01
NR_NTN_enh_plus_CT-UEConTest	Rel-18	205	1	2026-03-03
IoT_NTN_enh_plus_CT1-UEConTest	Rel-18	159	1	2025-12-12
NR_IoT_NTN_req_test_enh-UEConTest	Rel-19	150	0.34	2026-12-12
IoT_NTN_TDD-UEConTest	Rel-19	59	0.39	2026-12-11
IoT_NTN_FDD_LS_band-UEConTest	Rel-18	35	1	2024-12-12
NR_NTN_Ph3-UEConTest	Rel-19	34	0.14	2027-03-03
IoT_NTN_extLband-UEConTest	Rel-18	32	1	2025-06-06
IoT_NTN_FDD_S_band-UEConTest	Rel-19	28	1	2025-12-12
LTE_TN_NR_NTN_mob-UEConTest	Rel-19	26	0.46	2026-12-12
IoT_SAT_UEConTest	Rel-18	26	1	2025-09-09
IoT_NTN_Ph3_plus_CT1-UEConTest	Rel-19	0	0	2027-12-12

The work plan stores those completion values as long decimals; the column above cuts each one at two places and changes nothing else.

Two Release 10 rows are left out of the table because they belong to a different subject: the ancillary terrestrial component work for mobile satellite service in North America, which carries two documents between them L_Band_LTE_ATC_MSS-UEConTest S_Band_LTE_ATC_MSS-UEConTest.

Now read the completion column down the page. Every row standing below 1 is a Release 19 row, and every one of those five is planned to finish in 2026 or 2027.

Everything from Release 18 and earlier stands at 1, and one Release 19 band row does too — IoT_NTN_FDD_S_band-UEConTest, planned to 2025-12-12 IoT_NTN_FDD_S_band-UEConTest.

That is the shape of the whole chapter in one column. The rule rows of Release 19 are planned to 2025-09-09 and 2026-03-03; their test row is planned to 2027-03-03.

7.4 The shelf itself

Behind those rows sit the documents. They split cleanly in two: the ones this work created for satellites alone, and the general test documents that already existed and had satellite material added to them.

7.4.1 The four the work created

Document	What it tests	Documents	Change requests	Owner
TS 38.181	the satellite access node, on NR	374	220	RAN4
TS 38.521-5	the device, on NR	371	265	RAN5
TS 36.181	the satellite access node, on the older radio	123	90	RAN4
TS 36.521-4	the device, on the older radio	291	136	RAN5

The ownership column is the part people get wrong. The node's test document belongs to RAN4, the same group that wrote the node's requirements. Only the device's test document belongs to RAN5.

The two NR ones say plainly what each covers. The node document names three kinds of node and adds a second radio:

Those type names are the document's own and it is the document that defines them; this register carries the scope sentence and not the definitions behind it.

The device document is one sentence:

7.4.2 The six that were already there

These are the general test documents of 5G and of the older radio. None of them is about satellites; each of them now carries satellite material.

• TS 38.533 — the device's radio resource management tests, 299 documents and 298 of them change requests.

• TS 38.523-1 — the device's protocol tests for 5G, 213 documents, 212 change requests.

• TS 36.523-1 — the same for the older radio and its core network, 183 documents, all of them change requests.

• TS 36.508 — the shared test environment for the older radio, 145 documents, all change requests.

• TS 38.508-1 — the shared test environment for 5G, 118 documents, all change requests.

• TR 38.903 — how test tolerances and measurement uncertainty are worked out, 79 documents, all change requests.

Notice how nearly every document in that list is change requests and almost nothing else. That is what "adding a feature to an existing test document" looks like in the record: no discussion papers, no study, just the edits.

7.4.3 And the ones the change tables name but the register does not hold

A conformance work item's change table is longer than either list above. The Release 17 NR one names nine documents, and the Release 18 one eleven NR_NTN_solutions_plus_CT-UEConTest NR_NTN_enh_plus_CT-UEConTest.

Beyond the ones already named, those tables reach a second ring of documents, and the register holds most of them:

• TS 38.522 — which radio test cases apply to which device, 78 documents.

• TS 38.508-2 — the form a maker fills in to declare what its device implements, 70 documents. Its older-radio twins are TS 36.521-2 with 50 and TS 36.523-2 with 90.

• TS 38.523-3 — the protocol test suites themselves, 55 documents, with TS 36.523-3 at 61 on the older side.

• TR 38.905 — where in a requirement's range a test should measure, 56 documents, and TR 36.903 with 30 for the older radio's tolerances.

One document those change tables name is not in this register at all: the special testing function a device has to provide so a tester can drive it.

The register keeps a specification only where a work item lists it or enough documents name it, and 149 further specifications fall below that line ^[7]. Chasing the last of the shelf means opening the work item description itself.

7.5 The gap the conformance rows make visible

This is the reason the chapter exists. Take Release 19's NR phase 3.

• T1. NR_NTN_Ph3-Core stands at a completion value of 1, planned from 2023-12-13 to 2025-09-09, with 2807 documents behind it NR_NTN_Ph3-Core.

• T2. NR_NTN_Ph3-Perf also stands at 1, planned to 2026-03-03 NR_NTN_Ph3-Perf.

• T3. NR_NTN_Ph3-UEConTest stands at 0.14, planned to 2027-03-03, with 34 documents NR_NTN_Ph3-UEConTest.

Three rows of one feature, planned to finish in 2025, 2026 and 2027 respectively. The same shape appears further down the same list: IoT_NTN_Ph3_plus_CT1-UEConTest carries no document at all, stands at 0, and is planned to 2027-12-12 IoT_NTN_Ph3_plus_CT1-UEConTest.

The work plan carries no status word at all — no running, no stopped, no finished. It carries planned days and a completion value between 0 and 1, and 116 of the 155 short names in this subject carry a value of 1 ^{[8] [9]}.

That is the whole of what the record says about how far anything has got.

7.6 The older rows, and one that is easy to miss

Three more conformance rows are worth knowing by name, because a reader following the internet-of-things side will meet them:

• LTE_NBIOT_eMTC_NTN_plus_EPS-UEConTest — Release 17, 330 documents, planned to 2024-03-01, standing at 1 LTE_NBIOT_eMTC_NTN_plus_EPS-UEConTest.

• NR_IoT_NTN_req_test_enh-UEConTest — Release 19, 150 documents, planned to 2026-12-12 and not yet at 1 NR_IoT_NTN_req_test_enh-UEConTest.

• IoT_SAT_UEConTest — Release 18, 26 documents, standing at 1, and the one row here with no parent row above it in the plan's ordering IoT_SAT_UEConTest.

None of the NR band items of Bands, and why RAN4 carries more of this than anybody carries a conformance row of its own. Their change tables do name the node's test document TS 38.181, so the node half of the testing is written inside the band item.

The device half has no row of its own anywhere in this register, and the register does not say where it went — while the phase item beside them, NR_NTN_solutions_plus_CT-UEConTest, carries 1008 documents ^[5].

7.7 Where this sits in the rest of the course

The conformance shelf is the last of the radio half. From here the subject crosses into the system groups.

• R1. What the system was asked to do — SA1, and what the system was required to be able to do in the first place.

• R2. TS 38.521-5 and TS 38.181 are the two documents to open if you actually have to test something. This school carries the text of neither.

• R3. The overview course covers this ground in a paragraph rather than a chapter.