PhD Defense: Analyzing the Privacy and Security of Proof-of-Work Cryptocurrencies

Link to register and connect to the online PhD defense 

Members of the defense committee:

Prof. Dr Gilbert FRIDGEN, Université du Luxembourg, Chairman

Prof. Dr Peter Y A RYAN, Université du Luxembourg, Vice-Chairman

Prof. Dr Marcus VÖLP, Université du Luxembourg, Supervisor

Prof. Dr Jeremie DECOUCHANT, TU Delft, Netherlands, Member

Prof. Maria POTOP-BUTUCARU, Sorbonne Université, France, Member 

Abstract: 

In the past decade, we have witnessed the growth of cryptocurrencies. Nowadays, these currencies have generated significant impact in economy and society. Their

main innovation relies on the fact that they combine many techniques (e.g., cryptographic primitives and fault tolerance methods) to decentralize online payment

system. While cryptocurrencies have been developed since 2009, evaluating their privacy and security still remains challenging due to the complexity of the systems and the interplays between their different layers. In this thesis, we focus on proof-of-work (PoW) based cryptocurrencies (e.g., Bitcoin, Ethereum, Monero). We contribute three novel studies to improve the understanding of the privacy and security of PoW cryptocurrencies. Our first two studies focus on the impact of the network layer. First, we show that network properties are important to the system’s privacy and security. We analyze PoW cryptocurrency peer to peer networks and characterize the impact of network properties on privacy and security. Precisely, we design tools to measure the Monero peer to peer network, and conduct experiments to reveal that this network was vulnerable to some network level attacks despite its strong use of cryptographic primitives in the consensus layer. Second, we de fine new metrics to link the network layer and consensus layer, which allow us to evaluate the system’s security by considering network delays. We design a tool to measure the Bitcoin peer to peer network, and provide an empirical analysis. We conduct experiments to evaluate the impact of network delays on the consensus layer. As a consequence, we show that network delays have a significant impact on the system’s security. For instance, introducing relatively small network delays already simplifies double spending and selfish mining attacks, allowing adversaries to receive significant gains. We propose an abstraction, which we call dual private chain (DPC), in our third study that further challenges Bitcoin’s security and its consensus layer in the presence of Byzantine, Altruistic, and Rational nodes that are practical. We design a Markov Decision Process model for dual private chain attacks, and conduct Monte Carlo based simulations to evaluate the adversary’s success rate and revenue. Our results indicate that dual private chain attacks are more threatening to Bitcoin’s security compared to traditional temporary block withholding attacks. We suggest countermeasures to mitigate the effects of such attacks.