Sustainability and Performance Trustworthiness of IoT Monitoring Software Architectures in the Edge

This repository contains the raw and synthesized data used in the paper “Sustainability and Performance Trustworthiness of IoT Monitoring Software Architectures in the Edge”, published in the Journal of Systems and Software.

The article analyzes sustainability as a key attribute for assessing the trustworthiness of IoT monitoring software architectures deployed at the Edge. In the context of Industry 4.0 and Green software development, the study explores the relationship between sustainability and performance in Edge-based IoT monitoring systems, where energy consumption and response time are critical concerns. An exploratory experimental study is conducted using an indoor environmental monitoring IoT system as a case study. Four different IoT monitoring software architecture configurations are evaluated. Based on thirty experimental executions, the study examines how the distribution of monitoring and processing activities between Edge nodes and servers affects energy consumption and response time. The results demonstrate that balancing responsibilities between the Edge and the server enables the construction of trustworthy IoT monitoring software architectures that reduce both energy consumption and response time.

This repository includes three compressed files containing the experimental datasets and figures used in the study:

1. EnergyConsumption_IoTMonitoringEdgeArchitectures.zip
   This file contains the datasets related to the energy consumption evaluation. It is composed of four folders:

   • AdditionalMetrics: Raw data obtained from 24 experiments that were not directly used to calculate energy consumption but were generated by the measurement tools.

   • BasalEnergyConsumption: Raw data from experiments conducted to measure the basal energy consumption of the Smart Gateway, along with an Excel file containing the basal energy consumption calculations.

   • DataSynthesis: Analysis and synthesis of the energy consumption results.

   • EnergyMeasurementExperiments: Raw energy consumption data obtained from the 24 experiments.

2. ResponseTime_IoTMonitoringEdgeArchitectures.zip
   This file contains the datasets related to response time measurements. It includes five Excel files:

   • Architecture1.xlsx, Architecture2.xlsx, Architecture3.xlsx, and Architecture4.xlsx: Each file contains the raw response time measurements collected for the corresponding software architecture configuration. The data were obtained from 24 experiments per architecture, including measurements obtained at both the Server and the Edge Node (Smartgateway). It has to be noted that to compute the response time of the IoT data monitoring process, a set of timestamps was recorded during each experiment. As shown in Figure 5 of the Figure.zip file, five timestamps (t0 to t4) were defined, allowing the calculation of the computation time at the Smart Gateway, the computation time at the Edge/Server, and the total end-to-end response time when a data frame is transmitted through each architecture.

   • TimesArchitecturesDataSynthesis.xlsx: An Excel file that provides the synthesis and analysis of the response time data across the four evaluated architectures.

3. Figures.zip
   This file contains all the figures included in the paper, provided to facilitate result interpretation and reproducibility.