Toward Verification of DAG-Based Distributed Ledger Technologies through Discrete-Event Simulation.

Area of Science:

• Computer Science
• Distributed Systems
• Network Simulation

Background:

• Directed acyclic graph (DAG)-based distributed ledgers show promise for IoT systems.
• Discrete event simulations offer high-fidelity protocol evaluations.
• A gap exists in public discrete event simulators for DAG-based distributed ledgers.

Purpose of the Study:

• Introduce a novel discrete-event simulator for DAG-based distributed ledgers.
• Investigate the dynamics of IOTA's Tangle protocol using the simulator.
• Assess the efficiency and convergence properties of DAG architectures.

Main Methods:

• Developed a custom discrete-event simulator in Python.
• Implemented and simulated IOTA's Tangle DAG protocol.
• Analyzed transaction dynamics, random walk efficiency, and node convergence.

Main Results:

• Revealed Poisson processes in Tangle's transaction dynamics.
• Quantified the efficiency and intricacies of the random walk mechanism.
• Assessed node convergence and the impact of weight updates.
• Provided insights into the role and challenges of coordinators/validators.

Conclusions:

• The developed simulator is crucial for understanding DAG ledger dynamics.
• Findings for Tangle underscore the need for adaptable DAG simulators and tip selection methods.
• The research highlights the significance of weight updates and coordinator roles in DAGs.