Physical synchronisation for anomaly detection in epidemic blockchain consensus.

Area of Science:

• Complex Systems Physics
• Network Science
• Blockchain Technology

Background:

• Gossip protocols in peer-to-peer networks share similarities with epidemic spreading, but this analogy lacks formalization.
• Blockchain consensus mechanisms rely on information propagation, making them susceptible to disruptions and attacks.

Purpose of the Study:

• To formalize the analogy between gossip protocols and epidemic spreading within the context of blockchain consensus.
• To develop a physics-based framework for analyzing blockchain consensus dynamics and detecting anomalies.

Main Methods:

• Formalized gossip protocols using phase-encoding, reducing dynamics to coupled-oscillator synchronization on complex networks.
• Mapped block preferences to oscillator phases, communication latencies to natural frequencies, and network topology to coupling graphs.
• Developed a static phase proxy from block-attribution statistics to analyze real blockchain data and detect consensus disruptions.

Main Results:

• The order parameter, a synchronization measure, tracks simulated network consensus, correlating with phase transitions.
• Consensus disruptions manifest as phase-coherence disturbances, serving as candidate anomaly signals for network attacks.
• The proxy-based detector identified the 2013 Bitcoin chain fork with high statistical significance on unmodified data.

Conclusions:

• The physics-based framework provides a novel approach to understanding and securing blockchain consensus mechanisms.
• The methodology offers a complementary anomaly detection signal for blockchain operators and a testbed for epidemic modeling.
• The generalized phenomenology validates the applicability of coupled-oscillator synchronization to diverse blockchain protocols.