The term on this page
- PQCpost-quantum cryptography
- algorithms that run on ordinary computers and are believed to resist quantum attack
Also mentioned
RSARSARivest, Shamir and AdlemanA widely used public-key algorithm for encryption and digital signatures whose security relies on the difficulty of factoring large numbers.Read the full entry (new tab), ECDSAECDSAElliptic Curve Digital Signature AlgorithmElliptic Curve Digital Signature Algorithm, a widely deployed signature scheme based on elliptic-curve cryptography, offering strong security with compact keys.Read the full entry (new tab), ML-KEMML-KEMModule-Lattice-based Key Encapsulation MechanismModule-Lattice-Based Key-Encapsulation Mechanism, the NIST-standardized post-quantum KEM derived from CRYSTALS-Kyber and specified in FIPS 203.Read the full entry (new tab), ML-DSAML-DSAModule-Lattice-based Digital Signature AlgorithmModule-Lattice-Based Digital Signature Algorithm, the NIST-standardized post-quantum signature scheme derived from CRYSTALS-Dilithium and specified in FIPS 204.Read the full entry (new tab), SLH-DSASLH-DSAStateless Hash-based Digital Signature AlgorithmStateless Hash-Based Digital Signature Algorithm, the NIST-standardized signature scheme derived from SPHINCS+ and specified in FIPS 205.Read the full entry (new tab), FIPSFIPSFederal Information Processing StandardFederal Information Processing Standards, publicly announced standards developed by NIST for use in U.S. government computer systems, including cryptographic algorithms and modules.Read the full entry (new tab), NISTNISTNational Institute of Standards and TechnologyThe U.S. National Institute of Standards and Technology, the agency that develops and publishes cryptographic standards, including the FIPS series and post-quantum algorithms.Read the full entry (new tab) are defined in the glossary.
Read this first
On 19 November 2026, the IESG approved draft-ietf-pquip-pqc-hsm-constrained for publication as an Informational document. Revision 12 is now in the RFC Editor queue. The guidance turns constrained-device migration into measurable engineering work across RAM, storage, signing latency, link fragmentation and firmware trust. It is not an RFC yet, does not define protocol behavior or choose algorithms for a product, and its worked memory figures are examples rather than universal minimums.
The approval moves Adapting Constrained Devices for Post-Quantum Cryptography past IESG evaluation and into editorial processing. The IESG decision log records approval on 19 November. The Datatracker history shows revision 12 entering the RFC Editor queue that day and returning to reference checking and formatting on 20 November after brief author input. No RFC number has been assigned.
QuantaKrypto covered revision 08 when it entered IESG review in October. The new event is not a new algorithm or implementation release. It is the IETF's approval of the engineering guidance itself. For a PQC migration roadmap, that makes the document a stronger planning reference while leaving product-specific protocol choices, interoperability and validation to other specifications.
What the approval changes
| Question | 15 October coverage | 22 November status |
|---|---|---|
| Document state | Revision 08 in IESG evaluation | Revision 12 approved by the IESG and in the RFC Editor queue |
| Intended status | Informational Internet-Draft | Approved for Informational publication, still awaiting final RFC publication |
| Operational meaning | Useful review-stage guidance | Approved guidance that implementers can use for planning and procurement |
| What it does not prove | No final RFC or product deployment | Still no released device support, conformance result or field deployment evidence |
The distinction matters. IESG approval means the document cleared the IETF's technical and process review for its intended Informational status. It does not make the recommendations mandatory, assign protocol code points or certify an implementation. The draft explicitly says it does not specify protocol behavior or algorithm selection for constrained networks. Those decisions stay with the relevant protocol specifications and product threat model.
The resource budget is now an explicit migration gate
Revision 12 keeps the concrete resource accounting that makes the guidance useful. Its straightforward ML-KEM-768 example holds a 4,608-byte matrix plus three 1,536-byte polynomial vectors, totaling 9,216 bytes. A lazy-expansion example reduces the live set to about 2.5 KB by generating matrix and secret polynomials as needed. For ML-DSA-87 signing, the draft calculates at least 79 KB of RAM before per-iteration vectors.
These are examples, not device requirements
Representation, parameter set, compiler, hardware acceleration and operation all change the result. Verification can need less memory than signing. Lazy expansion saves RAM but can add computation. Measure the exact implementation on the target device instead of copying the draft's figures into a requirement sheet.
The draft primarily targets Class 1 constrained nodes and above, plus cryptographic modules with limited RAM, storage or processing capacity inside otherwise capable hosts. It says Class 0 devices generally lack enough RAM and storage for the covered algorithms and would typically offload PQC operations to a dedicated cryptographic module. That is a procurement boundary, not an optimization detail.
A device test must cover more than a successful handshake
A desktop library test answers whether an algorithm can run. A device migration test must answer whether the product can run it inside its actual power, memory, timing and update constraints. Add the following evidence to the cryptographic inventory before approving a rollout:
- Peak RAM and stack use for key generation, encapsulation, decapsulation, signing and verification.
- Persistent storage for seeds, expanded private keys, certificates, trust anchors and rollback images.
- Median and tail latency, especially for ML-DSA signing where rejection sampling makes the iteration count variable.
- Energy use and watchdog behavior under repeated operations rather than one warm benchmark.
- Packet and frame counts for public keys, ciphertexts, signatures and certificate chains on the real transport.
- Behavior under fragmentation, loss, retransmission and multi-hop mesh routing.
- Backup, import, rotation, zeroization and recovery for seeds and expanded keys.
The link layer matters because some post-quantum artifacts are one or two orders of magnitude larger than elliptic-curve equivalents. On a small-frame network such as IEEE 802.15.4, fragmentation can increase latency, energy consumption and loss exposure. An implementation audit should therefore capture complete wire artifacts and device telemetry, not only cryptographic return codes.
Firmware trust is part of the same boundary
The approved guidance treats secure boot, firmware updates and device attestation as migration surfaces. A device that negotiates a post-quantum session but accepts firmware signed only with RSA or ECDSA retains a future trust failure in its update path. The draft discusses HSS/LMS, XMSS, SLH-DSA and ML-DSA as candidate tools, each with different state, size and implementation tradeoffs.
That does not create one universal firmware profile. Stateful hash-based signatures require correct one-time-key state and backup discipline. SLH-DSA avoids that state but brings larger signatures and slower operations. ML-DSA is faster and shorter, but lattice implementations still need careful side-channel and randomness review. A firmware-signing test method should verify the full boot and recovery path with the exact format and key lifecycle selected.
Turn the guidance into acceptance criteria
For new hardware, post-quantum requirements should influence component selection before the board, secure element or HSM is fixed. For deployed hardware, the same requirements determine whether software optimization, offload or replacement is realistic. A migration team can convert the document into a short acceptance record for each device class:
- Named algorithm, parameter set, protocol profile and implementation version.
- Measured RAM, flash, latency, energy and transport overhead with reproducible fixtures.
- Key and seed protection equivalent to the private material they derive.
- A tested firmware and trust-anchor migration path, including recovery after a failed update.
- An algorithm-change path that can survive parameter or recommendation changes during the device lifetime.
- A stated limitation for operations the device cannot perform locally.
A conformance assessment can verify that ML-KEM or ML-DSA behavior matches the applicable standard. It cannot by itself prove that a battery-powered node survives fragmented traffic, that a watchdog tolerates signing tails or that a bootloader can recover from a rejected image. Those system properties need separate evidence.
What remains unfinished
The document still has no final RFC number and can receive editorial changes during publication. It requires no IANA actions and does not define a wire protocol. Implementers must still select the relevant protocol specification, wait for stable identifiers where required, verify released library support and test interoperability on the target hardware.
Use it as a planning baseline
Add constrained-device resource budgets and firmware trust to the migration roadmap now. Do not cite IESG approval as proof that a product supports PQC. Product evidence still requires released software or firmware, exact configuration, reproducible measurements and successful negative tests on the deployed hardware.
References
- IETF Datatracker, draft-ietf-pquip-pqc-hsm-constrained-12, 15 November 2026
- IESG decision log, approval recorded 19 November 2026
- Datatracker document history, RFC Editor processing on 19 and 20 November 2026
- NIST FIPS 203, ML-KEM
- NIST FIPS 204, ML-DSA
- IETF RFC 9019, A Firmware Update Architecture for Internet of Things
Where this goes next