Daniel Perez (Imperial College London), Benjamin Livshits (Imperial College London, UCL Centre for Blockchain Technologies, and Brave Software)

Metering is an approach developed to assign cost to smart contract execution in blockchain systems such as Ethereum. This paper presents a detailed investigation of the metering approach based on emph{gas} taken by the Ethereum blockchain. We discover a number of discrepancies in the metering model such as significant inconsistencies in the pricing of the instructions. We further demonstrate that there is very little correlation between the gas and resources such as CPU and memory. We find that the main reason for this is that the gas price is dominated by the amount of emph{storage} that is used.

Based on the observations above, we present a new type of DoS attack we call~emph{Resource Exhaustion Attack}, which uses these imperfections to generate low-throughput contracts. Using this method, we show that we are able to generate contracts with a throughput on average 50 times slower than typical contracts. These contracts can be used to prevent nodes with lower hardware capacity from participating in the network, thereby artificially reducing the level of centralization the network can deliver.

View More Papers

Detecting Probe-resistant Proxies

Sergey Frolov (University of Colorado Boulder), Jack Wampler (University of Colorado Boulder), Eric Wustrow (University of Colorado Boulder)

Read More

Automated Discovery of Cross-Plane Event-Based Vulnerabilities in Software-Defined Networking

Benjamin E. Ujcich (University of Illinois at Urbana-Champaign), Samuel Jero (MIT Lincoln Laboratory), Richard Skowyra (MIT Lincoln Laboratory), Steven R. Gomez (MIT Lincoln Laboratory), Adam Bates (University of Illinois at Urbana-Champaign), William H. Sanders (University of Illinois at Urbana-Champaign), Hamed Okhravi (MIT Lincoln Laboratory)

Read More

Data-Driven Debugging for Functional Side Channels

Saeid Tizpaz-Niari (University of Colorado Boulder), Pavol Černý (TU Wien), Ashutosh Trivedi (University of Colorado Boulder)

Read More

BLAZE: Blazing Fast Privacy-Preserving Machine Learning

Arpita Patra (Indian Institute of Science, Bangalore), Ajith Suresh (Indian Institute of Science, Bangalore)

Read More