Benchmarking Post-Quantum Cryptography in Lightweight Virtualization Environments on Embedded Hardware
Study finds container overhead is tiny for PQC, while algorithm choice can swing TLS energy by 1000x.
Researchers measured post-quantum signature and key-encapsulation primitives on embedded ARM under native execution, Docker, and a Unikraft unikernel on QEMU, covering about 70 parameter sets plus classical baselines. Container overhead was negligible for primitive computation; unikernel overhead was small for most families, moderate for BIKE, HQC, and MAYO (1.28–1.53×), and extreme for Falcon signing (17.8–19.2×). Per-operation energy tracked runtime. In TLS 1.3, environment differences faded once expensive post-quantum algorithms dominated, and algorithm choice changed handshake energy by up to three orders of magnitude.