Situation-Aware IoT Data Generation towards Performance Evaluation of IoT Middleware Platforms

Area of Science:

• Computer Science
• Software Engineering
• Data Science

Background:

• The proliferation of Internet of Things (IoT) applications necessitates robust performance assessment methods for IoT middleware platforms.
• Existing performance testing methodologies for databases and Big Data are insufficient for the complex and heterogeneous nature of IoT applications and their data.

Purpose of the Study:

• To introduce a novel situation-aware IoT data generation framework (SA-IoTDG) to overcome limitations in current IoT performance testing.
• To enable the generation of situation-specific data tailored to the requirements of event-driven IoT applications.

Main Methods:

• SA-IoTDG employs a situation-based approach, incorporating a situation description system and a SysML model for IoT application requirements.
• A novel Markov chain-based method facilitates dynamic transitions in IoT data generation according to specific situations.
• The framework was demonstrated using a real-world IoT traffic monitoring scenario.

Main Results:

• Experimental evaluations confirmed SA-IoTDG's capability to generate realistic IoT data comparable to real-world data.
• The generated data effectively supported performance evaluations of IoT applications on different middleware platforms.
• Promising outcomes validated the framework's efficacy in simulating IoT environments.

Conclusions:

• SA-IoTDG provides a valuable tool for researchers and developers to generate realistic IoT data for application testing and performance analysis.
• The situation-aware approach enhances the relevance and utility of generated data for evaluating IoT middleware performance.
• The framework facilitates crucial initial testing before the actual deployment of IoT applications.