MCiANT: A Monte Carlo inspired auction mechanism for health data NFT trading

Abstract

Auctions play a vital role in modern commerce by offering a transparent, structured, and competitive method for trading goods, services, and data. However, traditional in-person auctions are limited in terms of accessibility, convenience, security, and efficiency. However existing online auction platforms, while addressing some of these limitations, still face challenges such as limited transparency, centralized control, and insufficient security and privacy protections. Moreover these issues become extremely critical in case of sensitive applications like healthcare, defense and finance. To address these challenges, this article proposes a three-phase trading framework. In the first phase, data generation, anonymization, and storage are performed. In the second phase, an ensemble learning-based price forecasting approach is employed to estimate the asking and bidding prices, which depend on the volume and type of data. Finally, in the third phase, a Monte Carlo-inspired auction-based Non-Fungible Token (NFT) trading mechanism (MCiANT) is incorporated to enable efficient trading between buyers and sellers. The efficacy of the proposed MCiANT framework is compared with three distinct auction algorithms: the Vickrey auction, the Markov-Inspired Stationary Distribution Auction (MISD), and the Two-Phase English–Dutch Hybrid Auction (TPEDHA). The results demonstrate that the MCiANT framework significantly outperforms the others, achieving success-rate improvements of 5%, 5%, and 1% over the Vickrey, MISD, and TPEDHA auctions, respectively. Furthermore, the proposed framework is evaluated using health data by measuring anonymization time, encryption time, InterPlanetary File System (IPFS) upload time, and I/O performance.