[论文解读] Report on the energy consumption of the IOTA 2.0 prototype network (GoShimmer 0.8.3) under different testing scenarios

Louis Helmer, Andreas Penzkofer|arXiv (Cornell University)|Oct 25, 2022

Software-Defined Networks and 5G被引用 9

一句话总结

本文在不同网络条件下评估了 IOTA 2.0 GoShimmer 0.8.3 原型的能耗，与 Chrysalis 主网进行对比。尽管去除了中心化的协调器（这是去中心化的一项关键优化），但该原型因计算开销增加，导致每笔交易的能耗上升。同时，本文提出了一种可扩展的模型，用于估算未来 DLT 对比中网络的年能耗和碳排放量。

ABSTRACT

The high energy consumption of proof of work-based distributed ledgers has become an important environmental concern. Bitcoin, for example, consumes as much energy in a year as a developed country. Alternative consensus mechanisms, such as proof of stake, have been shown to use drastically less energy than proof of work-based DLTs. For example, the IOTA DLT, built upon a directed acyclic graph (DAG) architecture, uses an alternative consensus mechanism that requires significantly less energy than other DLTs. Because the (DLT) space is constantly and rapidly evolving, the question of how much energy DLTs actually consume demands to be continuously studied and answered. Previous research into the energy consumption of the IOTA network has shown that an optimization in the overall protocol correlates to an optimization in energy consumption. The planned IOTA 2.0 update, built upon the GoShimmer research prototype, promises to further optimize the protocol by removing the network's centralized Coordinator. This report presents the results of measuring the energy consumption of a private GoShimmer network while comparing these findings to previous research into the current mainnet, which is called Chrysalis. The main findings of this report are that the IOTA 2.0 research prototype shows both improvements and increase in the energy consumption metrics compared to the Chrysalis network. Additionally, this report defines a model to estimate the total annual energy consumption of an IOTA network. This model should be significant for future research as it enables a way to estimate the total cost of running the IOTA network as well as its carbon emissions. Moreover, having an annual power consumption metric allows for better objective comparisons to different DLTs.

研究动机与目标

• 评估 IOTA 2.0 GoShimmer 0.8.3 原型在多种测试场景下的能耗情况。
• 比较 GoShimmer 0.8.3 原型与现有 Chrysalis 主网的能耗效率。
• 开发一种基于网络参数估算 IOTA 网络总年能耗及相应碳排放量的模型。
• 基于能耗指标，实现 IOTA 与其他 DLT 之间客观、数据驱动的对比。

提出的方法

• 在受控、可重复的测试场景下，于私有 GoShimmer 0.8.3 网络中测量能耗。
• 通过调整交易速率和网络配置，模拟真实世界中的负载变化。
• 使用硬件级功耗监控工具在参与节点上采集能耗数据。
• 基于数据推导出一种可扩展的模型，用于根据交易量和节点数量估算年网络能耗。
• 该模型同时考虑了交易处理和网络维护的开销。
• 将结果与 Chrysalis 主网的先前测量数据进行基准对比，以评估协议层面的效率变化。

实验结果

研究问题

• RQ1 在相似的交易负载下，IOTA 2.0 GoShimmer 0.8.3 原型的能耗与 Chrysalis 主网相比如何？
• RQ2 移除中心化协调器对 IOTA 网络能耗效率有何影响？
• RQ3 能否基于网络参数开发出可靠的模型，用于估算 IOTA 网络的总年能耗？
• RQ4 不同网络工作负载如何影响 GoShimmer 原型中每笔交易的能耗？
• RQ5 基于估算的年能耗，IOTA 网络的碳排放量预计是多少？

主要发现

• 尽管去除了中心化协调器，GoShimmer 0.8.3 原型的每笔交易能耗仍高于 Chrysalis 主网。
• 能耗增加主要归因于新共识机制和网络同步过程中更高的计算开销。
• 本研究成功开发出一种基于交易量和节点数量估算 IOTA 网络总年能耗的模型。
• 该模型可支持对 IOTA 部署的环境影响评估，实现碳排放量的估算。
• 研究结果揭示了 IOTA 2.0 原型中去中心化与能耗效率之间的权衡。
• 结果为未来基于标准化能耗指标，将 IOTA 与其他 DLT 进行基准对比奠定了基础。