Performance Perspective on Private Distributed Ledger Technologies for Industrial Networks

Abstract

Blockchain-based Distributed Ledger Technology (DLT) is a novel paradigm to create tamper-resistant execution environments and data storage for distributed applications on top of a peer-to-peer network. This technology has shown to be of interest in many use-cases, especially in industrial processes where multiple shareholders would like to process and share data in a secure and accountable way. In this work, we evaluate the performance of a DLT-based system via modeling and a quantitative performance evaluation, focusing on the impact of the underlying communication network. Our numerical evaluation is based on the Hyperledger Fabric DLT framework, its benchmarking tool Caliper, and a dedicated test bed, where network properties such as latency or packet loss can be artificially influenced. Our experiments show that the validation of the transactions in a DLT-based system is the main contributor in the transaction latency. We also demonstrate that the properties of the communication network can influence the performance largely, even in the case where only one of the participants in the DLT system has poor network access.