Post-Quantum Forward-Secure Signatures with Hardware-Support for Internet of Things

ICC 2023 - IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS(2023)

引用 1|浏览1
暂无评分
摘要
Digital signatures provide scalable authentication with non-repudiation and therefore are vital tools for the Internet of Things (IoT). IoT applications harbor vast quantities of low-end devices that are expected to operate for long periods with a risk of compromise. Hence, IoT needs post-quantum cryptography (PQC) that respects the resource limitations of low-end devices while offering compromise resiliency (e.g., forward-security). However, as seen in NIST PQC efforts, quantum-safe signatures are extremely costly for low-end IoT. These costs become prohibitive when forward security is considered. We propose a highly lightweight post-quantum digital signature called HArdware-Supported Efficient Signature (HASES) that meets the stringent requirements of resource-limited signers (processor, memory, bandwidth) with forward security. HASES transforms a key-evolving one-time hash-based signature into a polynomially unbounded one by introducing a public key oracle via secure enclaves. The signer is non-interactive and only generates a few hashes per signature. Unlike existing hardware-supported alternatives, HASES does not require a secure-hardware on the signer, which is infeasible for low-end IoT. HASES also does not assume non-colluding servers that permit scalable verification. We proved that HASES is secure and implemented it on the commodity hardware and the 8-bit AVR ATmega2560 microcontroller. Our experiments confirm that HASES is 271 similar to and 34 similar to faster than (forward-secure) XMSS and (plain) Dilithium. HASES is more than twice and magnitude more energy-efficient than (forwardsecure) ANT and (plain) BLISS, respectively, on an 8-bit device. We open-source HASES for public testing and adaptation.
更多
查看译文
关键词
Authentication,Internet of Things,post-quantum security,embedded devices,lightweight cryptography
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要