Paper 2025/2032

TrX: Encrypted Mempools in High Performance BFT Protocols

Rex Fernando, Aptos Labs
Guru-Vamsi Policharla, University of California, Berkeley
Andrei Tonkikh, Aptos Labs
Zhuolun Xiang, Aptos Labs
Abstract

MEV (Maximal Extractable Value) remains one of the most corrosive forces in blockchain systems, enabling frontrunning, sandwiching, and other manipulations that directly exploit users. The core culprit is the transparent mempool: validators see transactions before they are ordered. Encrypted mempools are a promising solution by hiding transaction contents until after ordering. We present the first integration of encrypted mempools with a high-performance BFT protocol. Our system uses a cryptographic scheme based on recent work on batched threshold encryption, and improves on the cryptographic state of the art in this line of work. The system ensures confidentiality of transactions during ordering while sustaining performance on par with leading BFT designs. Specifically, the proposal-to-execution latency of our system yields only a 27 ms overhead (14%) compared to the baseline. The result is a practical consensus layer that simultaneously defends against MEV and delivers the throughput and latency needed for real deployments. This work closes the gap between cryptographic defenses and production-ready consensus, showing that robust MEV protection and high performance can, in fact, coexist.

Metadata
Available format(s)
PDF
Category
Public-key cryptography
Publication info
Preprint.
Keywords
batch threshold encryptionMEVmempool privacy
Contact author(s)
rex1fernando @ gmail com
guruvamsi policharla @ gmail com
andrei tonkikh @ gmail com
xiangzhuolun @ gmail com
History
2025-11-05: approved
2025-11-02: received
See all versions
Short URL
https://ia.cr/2025/2032
License
Creative Commons Attribution
CC BY

BibTeX

@misc{cryptoeprint:2025/2032,
      author = {Rex Fernando and Guru-Vamsi Policharla and Andrei Tonkikh and Zhuolun Xiang},
      title = {{TrX}: Encrypted Mempools in High Performance {BFT} Protocols},
      howpublished = {Cryptology {ePrint} Archive, Paper 2025/2032},
      year = {2025},
      url = {https://eprint.iacr.org/2025/2032}
}
Note: In order to protect the privacy of readers, eprint.iacr.org does not use cookies or embedded third party content.