Even if Bitcoin ultimately secures its on-chain transactions against quantum threats, five distinct components of the Lightning Network’s off-chain payment infrastructure would still demand separate modifications, a Sept. 12 research preprint indicates.
The study introduces PQLN, a hybrid post-quantum framework built as a rust-lightning research prototype. It focuses on gossip data (which shares node and channel details), encrypted peer-to-peer links, signed invoices, reusable payment offers, and onion routing packets designed to hide payment routes.
Currently, Lightning relies on secp256k1-based ECDSA, Schnorr signatures, or ECDH for these tasks. PQLN retains these traditional methods while incorporating ML-DSA signatures into invoices, offers, and gossip. It also integrates hybrid ML-KEM key exchanges into transport channels, payment onions, and BOLT 12 privacy pathways.
The researchers created a feature-gated fork of rust-lightning alongside an adjusted ldk-sample node. Their matrix of 12 scenarios demonstrated that standard and post-quantum binaries derived from the same rust-lightning core can run side by side. Depending on specific settings and routing paths, mixed-mode configurations either fell back to classical security or safely closed down.
Such interoperability does not mean the broader network is prepared for deployment. The codebases are explicitly categorized as research prototypes, and the evaluation omitted testing against alternative Lightning implementations.
Communication overhead, rather than pure cryptographic calculation, proved to be the primary factor in the findings. When performing a static synchronization of an entire network topology, the baseline ML-DSA-44 and ML-KEM-768 architecture caused gossip download volumes to multiply by 10.2 and graph data storage requirements to increase by 8.8 relative to standard Lightning.
Employing a smaller, Falcon-based FN-DSA setup lowered the download multiplier to 4.2, though the National Institute of Standards and Technology (NIST) still classifies that specific algorithm as a work in progress rather than an official standard.
Routing private payments introduced additional bandwidth costs. A static list of post-quantum ciphertexts sized at 21.8 kilobytes added between 19 and 53 milliseconds of delay per hop across an emulated 10 Mbit/s connection, and between 160 and 187 milliseconds per hop on a 1 Mbit/s link.
Conversely, signature generation using ML-DSA-44 took an average of 327 microseconds on the investigators’ hardware setup, which featured a 16-core AMD Ryzen Threadripper PRO 3955WX workstation paired with 64 GB of RAM. This hardware-specific metric for a single primitive should not be interpreted as overall payment latency.
Furthermore, PQLN leaves vulnerabilities rooted in the main Bitcoin blockchain unaddressed. It fails to protect the on-chain keys securing commitment outputs, funding, penalty transactions, or HTLCs, meaning that safeguarding channel assets still fundamentally depends on post-quantum updates at the base Bitcoin layer.
The paper works under the assumption of a future quantum adversary capable of breaking current cryptography; it does not document any active quantum attacks against Lightning or Bitcoin today. While the prototype indicates that Lightning’s off-chain defenses can undergo evaluation independently of foundational Bitcoin upgrades, it does not imply the network is primed for migration. Ultimately, the primary operational challenge highlighted by the tests is the volume of extra storage, gossip data, and transaction traffic that node operators will be forced to handle.
