Resilience of IOTA Consensus

TL;DR

This paper evaluates the convergence and resilience of IOTA's consensus algorithms, revealing their limitations under various network topologies and adversarial conditions, and highlighting open challenges in IoT blockchain design.

Contribution

It provides an extensive simulation-based analysis of IOTA consensus algorithms' performance and resilience across different network topologies, offering insights for future blockchain design in IoT.

Findings

Poor convergence rates under low power adversaries

Limited scalability except in Watts Strogatz topologies

Design challenges remain for IoT-specific blockchains

Abstract

Blockchains are appealing technologies with various applications ranging from banking to networking. IOTA blockchain is one of the most prominent blockchain specifically designed for IoT environments. In this paper we investigate the convergence of IOTA consensus algorithms: Fast Probabilistic Consensus and Cellular Consensus, when run on top of various topologies. Furthermore, we investigate their resilience to various types of adversaries. Our extensive simulations show that both Cellular Consensus and Fast Probabilistic Consensus have poor convergence rates even under low power adversaries and have poor scaling performances except for the case of Watts Strogatz topologies. Our study points out that the design of IOTs dedicated blockchains is still an open research problem and gives hints design.