Satellites — a base station in orbit

Easy to get wrong

One from each chapter that has one. The course carries 20 in all.

• 1.4 What a non-terrestrial network is, in 3GPP's own words

  A satellite access node is not a satellite. The definition puts the payload, the ground station and the base station functions inside one node. So "how many SANs does this constellation have" is not a question about how many spacecraft are in the sky, and a sentence that treats the two as the same thing has already gone wrong.

• 2.6 Why anyone wanted it, and who said so

  A justification is a claim, not a finding. It is what a group wrote to persuade a plenary to let it start. This record holds no vote and no reason for any decision, so a justification never tells you whether the room agreed with it, and a quoted phrase like "commercial deployment are ongoing" is evidence of what was argued, not of what is true.

• 3.3 How the work is cut up, and who carries it

  Never add the document counts of a feature and its parts together. They are four separate rows counted separately, and a document naming the core part is not also counted under the feature. The Release 17 NR core part carries 4346 documents and the feature row 781 ^{[2] [3]}; those answer different questions and their sum answers none.

• 4.3 NR over satellite, release by release

  Neither TR 38.811 nor TR 38.821 can be quoted here. This school holds no parsed text for either, so the two reports the whole NR line rests on can be named, counted and pointed at, and never read. Anything a chapter said about their contents would be invention.

• 5.3 The same idea on the older radio, for things that are not phones

  The internet-of-things satellite work lives in the older packet core, not in the 5G core. Four of these work items name TS 23.401 in their own change tables TS 23.401, which is why a reader who goes looking for it in TS 23.501 finds the 5G satellite architecture instead — a related but different piece of work, and the subject of a later part of this course.

• 6.2 Bands, and why RAN4 carries more of this than anybody

  "The busiest meeting" here means the busiest of the 550 meetings that carry a document of this subject ^[4], counted only by papers whose row names a satellite work item. It is not a claim about the busiest 3GPP meeting of any kind, and 6568 documents whose titles say satellite or NTN carry no work item short name at all and are in none of these figures ^[1].

• 7.5 Proving it works — the conformance shelf

  A completion value of 1 on a core row means one thing only: the work plan says the plan for that row is complete. It is not a statement that the feature is finished, that anything was published, or that a device can be tested against it. The conformance row is a separate row with its own number, and for Release 19's NR phase 3 that number is 0.14 with a planned finish of 2027-03-03.

• 8.12 What the system was asked to do

  A study report and a specification are not two versions of the same thing. TR 22.822, TR 22.865 and TR 22.887 are homework — nobody has to obey a word of them. TS 22.261 is the rule. When somebody says "3GPP requires X for satellites", the question to ask is whether X is in the specification or only in the report that led to it.

• 9.6.1 The architecture, phase by phase — and backhaul, which runs the other way

  "Serving satellite" is not a permanent relationship. The definition builds the limit into itself: depending on the orbit, that satellite covers a given area for a limited period only. Phase 2's whole work item is about what the system does when coverage comes and goes, which is that same limit stated from the other side.

• 10.3 What the phone says to the network

  The leadership line and the group list answer different questions. A work item led by SA2 can have its later papers written in CT working groups, and two large items here do. Reading only the leadership line makes the CT groups look smaller in this subject than they are.

• 11.4 Security, the bills, the management, the layer above and the voice

  A completion value of 1 on a work plan row does not mean a group has finished with a subject. The work plan carries planned days and that value and no status field at all, so the only honest reading is that the plan for that row is complete. Whether more work follows is recorded nowhere in this register.

• 12.5 The specifications this work created, and the ones it changes

  "102 specifications affected" sounds like 102 documents carrying satellite text. It is not that. Some are study reports nobody has to obey, some are on the list because a work item promised to change them, and the counts behind each one are counts of documents that carry a satellite work item short name — never of everything ever written about the document.

• 13.3 What the shape of the paperwork says, and where the work stands

  An outcome word belongs to one document at one meeting and says nothing about the work item it was filed under. "Approved" on a change request does not mean the feature is approved; "noted" does not mean it was rejected. No store behind this course holds a status word for a piece of work at all.

• 14.5.4 Reading the record yourself, and what it cannot tell you

  This course cannot tell you what any study concluded. The register holds titles, scopes, objectives, counts and the words 3GPP defines for itself. It holds no key issue, no candidate solution, no evaluation result and no requirement. Any statement about how well satellite access performs did not come from this course and cannot be checked in it.