"An Analysis of Proof-of-Work based Blockchains under An Adaptive Double-spending Attack" - Blockchain Research Monthly Talk - Ramezan Gholamreza

Blockchain@UBC

G. Reza Ramezan is a research assistant and Ph.D. candidate at The University of British Columbia (UBC). His research interest focuses on telecommunications and Blockchains. Reza has received two Master’s degrees in Electrical and Computer Engineering (ECE) and IT Management (ITM). He also has extensive work experience as a network security consultant for more than 10 years in different telecommunication companies including AT&T and Telus.
In this talk, we first propose an adaptive strategy for double-spending attack on blockchains. The attacker in our strategy observes the length of the honest branch when a submitted transaction becomes available in the blockchain, and then updates the attack strategy accordingly. This provides a stronger strategy than conventional double-spending attack. We then derive closed-form expressions for the probability of a successful attack and the expected reward of attacker miners. Our analysis shows that the probability of a successful attack by convincing the network nodes to follow the counterfeit branch under the proposed attack strategy is 60% higher than what is expected from the conventional attack strategy when the attackers acquire 40% of the total network processing power. To counter this increase in the probability of attack, the network nodes are required to use a bigger number of confirmation blocks for validating any transaction in the blockchain. We computed the expected reward of an attacker for mining a counterfeit branch on a blockchain and observed that the expected reward drops to zero after a few number of  block confirmations.