Daily Dose of 6G

The deadline 3GPP did not set

Every study this newsletter has covered is a draft at version zero-point-something, aimed at a release that freezes in 2028. Today’s is finished. TR 33.703 closed in June, and it is the one piece of 6G-era security work that could not wait for 6G, because its clock belongs to somebody else.

Transcribed from 3GPP’s published Release 21 timeline (© 3GPP 2026). Editions 14 to 17 held the highlight on Rel-21 Stage 2; today it moves back. Post-quantum work has left the study phase, and the normative documents it is aimed at are Release 20 specifications, whose Stage 3 freeze is the highlighted row — not Release 21’s.

A study with an end date that has passed

TR 33.703 is “Study on transitioning to Post Quantum Cryptography (PQC) in 3GPP”. Responsible group SA3, Release 20 — and its status is not draft. It is under change control, the state a 3GPP document reaches once a plenary has approved it and further edits arrive as change requests. Nothing else this series has covered has got there.

Both records were readable this time. The specification record names two rapporteurs, Zhongding Lei of Huawei Technologies and Virendra Kumar of Qualcomm Technologies Int. The work item record for FS_CryptoPQC, UID 1080045, names one of them — listed there as Zander (Zhongding) Lei, Huawei — and adds what the specification record does not: WID SP-250858, and a study running from 5 June 2025 to 6 June 2026.

The version archive tells that year as a size curve, with a shape the drafts in Editions 14 to 17 never reached: it stops.

Sixty-three times its opening size, and the sixteen-fold step arrives in October 2025 — early, where the 6G studies grow late. The step from v1.0.0 to v20.0.0 is the approval: one becomes twenty because twenty is the release. One oddity, recorded without a theory attached — a v0.6.0 was also uploaded, on 25 June 2026, the same day as v20.0.0 and three weeks after the version sent for approval.

G-agnostic, and that is the whole argument

Stawros Orkopoulos, an SA WG3 delegate at Nokia, wrote 3GPP’s own account of this work for Highlights Issue 12. He describes the study as “G-agnostic” — deliberately not attached to a generation — and its purpose as evaluating “the potential updates to cryptographic profiles”. The reason is in his framing of the threat: “The threat of Harvest Now Decrypt later (HNDL) refers to a long-term surveillance strategy in which sensitive data is collected today with the expectation that it can be decrypted in the future.”

Traffic captured this year is a liability whenever the machine to decrypt it arrives, which makes the migration date a function of how long the data stays sensitive, not of when 6G ships. Orkopoulos sets the clock against national agencies — he names NCSC, ANSSI and BSI: “The common recommendation is to complete the migration for high priority systems by around 2030 and for all systems by approximately 2035.”

Where the cryptography actually lands

The study is closed and the work is not, and the SA3 email-discussion tree shows exactly where it went. The PQC folder holds twenty-six files, the newest from August 2026, and its recent additions are not study material. Two are titled as living document skeletons: one for “IKEv2 and TLS for TS 33.210”, one for “vendor credentials for TS 33.310”, both filed on 6 August and revised the following day. A third, from AT&T, updates a TLS profile.

Those two numbers are the point. TS 33.210 and TS 33.310 are network-domain security and the certificate framework, and have carried 3GPP’s transport security and PKI since long before 5G. Orkopoulos names the dependency: “3GPP SA3 Security standards extensively integrate widely used protocols and framework defined by IETF. For example, TLS, IPsec, PKI, JOSE.” Post-quantum migration here is therefore mostly a profiling exercise on other people’s protocols, plus the places 3GPP invented its own — 5G-SUCI calculation in TS 33.501, MIKEY-SAKKE for mission critical in TS 33.180.

Beside the PQC folder sits a Cryptographic Inventory one from 2025, holding spreadsheets of specifications and a protocol list: the audit that has to precede a migration, and visibly the working material behind TR 33.938, 3GPP’s published cryptographic inventory.

Two work items in one week

SA3#129 in Prague, 24 to 28 August 2026, is the meeting Edition 17 read for its security areas. Two of its drafts belong here instead, and both are WIDs rather than contributions: S3-263245, “New WID revision PQC migration”, and S3-263227, “WID Support of 256-bit algorithms”, filed seventeen minutes apart on the Monday morning. A work item description is the document that turns a finished study into normative specification work.

They are also the answer to a question this newsletter’s own queue got wrong. Post-quantum and 256-bit are not one subject. They are two, on separate documents and separate clocks — and the 256-bit thread splits again, into TR 33.700-41 on the transition to 256-bit algorithms and TR 33.771 on the AEAD algorithms themselves. Three TR numbers, one loose idea of “stronger crypto”.