1 A tag with no battery, and the gap it was meant to fill

1.1 The gap somebody wrote down

Every radio 3GPP had built up to 2022 assumed one thing: the device has its own power. A phone has a battery you charge, a meter has a battery you replace, a tracker has a battery that lasts ten years and then does not.

In March 2022 a study opened by saying that this assumption was leaving whole classes of use unserved:

The same sentence names the three gaps: places under extreme conditions such as high pressure or extreme heat and cold; devices that must be a millimetre thick, maintenance-free and long-lived; and places where a battery-driven device simply cannot be used ^[1].

Nothing about that list is about speed or capacity. It is about where you can put the thing and whether anybody ever has to go back to it.

1.2 What the thing is

The study answers its own question in the next sentence, and this is the definition the whole area is built on:

Three years later, when the architecture had actually been specified, the published document said the same thing in one line:

That specification also gives a name to the patch of network in which such a device can be reached — an AIoT Area, carried by an identifier of its own, and one area may hold several radio nodes and several readers ^[8].

1.3 What the first study asked for

The founding study is FS_AmbientIoT, unique id 950004, run by SA1, the group that writes what a system must do FS_AmbientIoT. Its own objective is short:

Note what is left open. The energy may come from radio waves, light, motion or heat, and the study explicitly refuses to close the list. That is a requirements document doing its job: it says what must be possible, not how.

It was approved as document SP-220085 at plenary SP-95-e, sent in by SA WG1 SP-220085. It ran from 10 March 2022 to 28 May 2023 and its report is TR 22.840, against which 497 documents were filed FS_AmbientIoT TR 22.840.

1.3.1 What the four named sources imply

Radio waves, light, motion and heat are not interchangeable, and the choice decides where a device may live.

Light needs a lit place. Motion needs something that moves. Heat needs a temperature difference. Radio waves need a transmitter nearby — which is the one case where the network itself can supply the energy, and it is the case the radio work went on to build around ^[6].

The study refusing to close the list matters for the same reason. A document that named one source would have decided the design.

Naming four, and adding "any other power source that could be seen suitable", left the decision with the groups that came later ^[4].

1.4 The harder question the radio people asked

Wanting a battery-less device is one thing. Building a radio for it is another, and the radio side opened with a study that asked whether it could be done at all:

That study is FS_Ambient_IoT_RAN, unique id 970078 — the only Rel-18 item in the whole area. It was approved as RP-232404 at plenary RP-101, sourced from Huawei and HiSilicon, and its report is TR 38.848 FS_Ambient_IoT_RAN RP-232404 TR 38.848.

The same objective adds that the work must "provide clear differentiation" from what 3GPP already had — it names "existing 3GPP LPWA IoT technology" and does not spell those four letters out, so neither does this course ^[5].

1.5 How small is small

The follow-on study, the one that looked for actual radio designs, put numbers on the target. This is the sentence to remember, because every later argument about what is buildable traces back to it:

A microwatt is a millionth of a watt. The same scope names a second, larger device at up to a few hundred microwatts, and the first of the two backscatters its answer on a carrier wave provided from outside rather than generating one ^[6].

1.6 What was finally specified

Five studies and three years later, real binding text exists. Its scope states plainly what it covers, and — just as importantly — what it does not:

The second sentence is the one people miss: in this release the Ambient IoT system is an isolated private network, one that does not interact with a public network ^[7].

The security specification's title says the same from the other side — security for isolated private networks TS 33.369.

So the first release of real Ambient IoT work is not "battery-less tags on the public mobile network". It is a self-contained network of its own, and the architecture that describes it is TS 23.369 TS 23.369.

That scope also shows how much of this area is borrowed rather than invented. It complies with requirements written elsewhere, in TS 22.369 TS 22.369, and it defers for its device types, traffic types, use cases and topologies to a radio document, TS 38.300 TS 38.300 ^[7].

Three documents, three groups, one system. That interlocking is the reason this course is organised by group rather than by document: no single specification here is readable on its own.

1.7 Where this leads, and where to read on

The first Ambient IoT document of any kind was filed on 2022-04-29 ^[9], soon after the founding study started on 10 March 2022 FS_AmbientIoT.

What happened between then and now is the subject of Study first, then build, then do it again and Seventeen groups, and where the weight sat.

Four documents carry the thinking behind this chapter. This school does not hold the text of any of them, so the hop-off points are these:

• TR 22.840 TR 22.840 — the founding study report: use cases and requirements, written by SA1.

• TR 38.848 TR 38.848 — whether a radio design was feasible at all.

• TR 38.769 TR 38.769 — the radio solutions themselves, and the source of the microwatt figure.

• TS 23.369 TS 23.369 — the architecture, and the only one of the four that is binding text rather than a study.