Sarisht Wadhwa (Duke University), Jannis Stoeter (Duke University), Fan Zhang (Duke University, Yale University), Kartik Nayak (Duke University)

Hashed Time-Locked Contracts (HTLCs) are a widely used primitive in blockchain systems such as payment channels, atomic swaps, etc. Unfortunately, HTLC is incentive-incompatible and is vulnerable to bribery attacks.
The state-of-the-art solution is MAD-HTLC (Oakland'21), which proposes an elegant idea that leverages miners' profit-driven nature to defeat bribery attacks.

In this paper, we show that MAD-HTLC is still vulnerable as it only considers a somewhat narrow set of textit{passive} strategies by miners. Through a family of novel textit{reverse-bribery} attacks, we show concrete textit{active} strategies that miners can take to break MAD-HTLC and profit at the loss of MAD-HTLC users. For these attacks, we present their implementation and game-theoretical profitability analysis.

Based on the learnings from our attacks, we propose a new HTLC realization, He-HTLC (Our specification is lightweight and inert to incentive manipulation attacks. Hence, we call it He-HTLC where He stands for Helium.) that is provably secure against all possible strategic manipulation (passive and active). In addition to being secure in a stronger adversary model, He-HTLC achieves other desirable features such as low and user-adjustable collateral, making it more practical to implement and use the proposed schemes. We implemented He-HTLC on Bitcoin and the transaction cost of He-HTLC is comparative to average Bitcoin transaction fees.

View More Papers

RoVISQ: Reduction of Video Service Quality via Adversarial Attacks...

Jung-Woo Chang (University of California San Diego), Mojan Javaheripi (University of California San Diego), Seira Hidano (KDDI Research, Inc.), Farinaz Koushanfar (University of California San Diego)

Read More

How to Count Bots in Longitudinal Datasets of IP...

Leon Böck (Technische Universität Darmstadt), Dave Levin (University of Maryland), Ramakrishna Padmanabhan (CAIDA), Christian Doerr (Hasso Plattner Institute), Max Mühlhäuser (Technical University of Darmstadt)

Read More

Machine Unlearning of Features and Labels

Alexander Warnecke (TU Braunschweig), Lukas Pirch (TU Braunschweig), Christian Wressnegger (Karlsruhe Institute of Technology (KIT)), Konrad Rieck (TU Braunschweig)

Read More