Appendable-block Blockchain Evaluation over Geographically-Distributed IoT Networks

Abstract

In the last few years, different researchers presented proposals for using blockchain in the Internet of Things (IoT) environments. These proposals consider that IoT environments can be benefited from different blockchain characteristics, such as: resilience, distributed processing, integrity and non-repudiation of produced information. However, researchers faced some challenges to use blockchain in IoT, e.g., latency, hardware and energy constraints, and performance requirements. One of the prominent solutions is the appendable-block blockchain, which uses a hierarchical peer-to-peer (p2p) gateway-based architecture. Additionally, current proposals present simplified evaluation scenarios, usually performed in controlled environments, which do not include important network features, for example, latency. Consequently, a model to evaluate a geographically distributed environment, for example, in a situation in which health data have to be collected from different countries in a pandemic situation, can help to understand the behavior and possible flaws of blockchains. In order to evaluate appendable-block blockchains in a realistic scenario, this paper presents an analysis of different consensus algorithms in geographically distributed hosts, in which latency can impact the performance of main operations in a blockchain, such as block and transaction insertion.

performance evaluation of permissionless distributed ledger technology ethereum

“…Hyperledger has more throughput and less latency compared to Ethereum, but Ethereum can handle more concurrent transactions than Hyperledger. Dinh et al [9] used BLOCKBENCH to conduct a comparative performance analysis on three mainstream private blockchains, namely Ethereum (Geth v1.4.18), Parity (v1.6.0), and HLF (v0.6.0-preview). Their findings can be summarized as follows: 1) HLF consistently outperforms Ethereum and Parity across all macro (e.g., throughput and latency) and micro (e.g., IOHeavy) benchmarks, but it cannot scale to more than 16 nodes; 2) The consensus protocols have been identified as major bottlenecks for HLF and Ethereum, while transaction signing is a bottleneck for Parity.…”

“…Their findings can be summarized as follows: 1) HLF consistently outperforms Ethereum and Parity across all macro (e.g., throughput and latency) and micro (e.g., IOHeavy) benchmarks, but it cannot scale to more than 16 nodes; 2) The consensus protocols have been identified as major bottlenecks for HLF and Ethereum, while transaction signing is a bottleneck for Parity. The authors then compared the performance of two different versions of HLF v0.6.0 and v1.0.0 [9]. Rouhani and Deters [10] investigated Ethereum's performance on a private blockchain by examining two of the most popular Ethereum clients: PoW-based Geth and PoAbased Parity.…”

Multi-level consensus algorithm for appendable-block blockchains in IoT Environments

“…After that -during the design, development and improvements on appendableblock blockchain -we also published our research in conferences and journals [Lunardi et al 2019a, de Arruda et al 2020, Lunardi et al 2022b]. Some of these papers were produced in collaboration with other international research groups, in particular with researchers from the University of New South Wales (UNSW) and Commonwealth Scientific and Industrial Research Organisation (CSIRO) -both from Australia -and researchers from Newcastle University -from United Kingdom.…”