MaskAuct: Seller-Autonomous Auction With Bidder Anonymity and Bidding Confidentiality
Siqin ZHOU Li LI, Kun He, Jing Chen, Min Shi, Meng Jia, Ruiying Du, Ling Han · IEEE Transactions on Information Forensics and Security · 2024
Electronic auctions, popular in the digital era, raise great privacy concerns that may impact participant interests. However, traditional privacy-preserving auction systems fall short in facilitating seller autonomy, particularly in identifying and excluding previously mischievous anonymous bidders. In this paper, we propose MaskAuct, a seller-autonomous auction system with the privacy of bidder identity and bidding price. To enable seller autonomy without compromising bidder privacy, we present a new cryptographic primitive, called Zero-Knowledge Blacklistable Group Signature (ZKBGS), which can invalidate signatures from users in the blacklist without opening user identity. We construct MaskAuct from fully homomorphic encryption and ZKBGS, and introduce the distributed privacy server provider to address the collusion problem. The experimental results show ZKBGS has a smaller signature size (8320 bytes) and running time (635 ms for signing and 24 ms for verification) than the linkable ring signature, even when the length of the blacklist is$2^{9}$. In contrast to the sealed-bid auction scheme SEAL, MaskAuct provides better communication complexity, and is$27\times $faster on bidder computation.