Model-Based Testing of Multi-Agent Systems
Loading...
Date
2022
Journal Title
Journal ISSN
Volume Title
Publisher
Faculty of Sciences
Abstract
This thesis uses a method that encompasses all levels of higher agent abstraction to address the challenge of formal model-based testing for multi-agent systems. The agent level, the agent company level, and the system level are the three levels of abstraction. The proposed method is ethical when it comes to model-based testing. By establishing a process that incorporates requirement management, test case creation automation, and test case execution. We first propose an agent-based system that uses reference networks to represent a mobile agent. The example was created using the MULAN (multi agent system net) architecture and the Java programming language. Second, we provide a model-based (network-to-network) testing approach that works in conjunction with the JUNIT tool for developing and testing multi-agent systems. Unlike other test methodologies, where instrumentation is primarily related to the program's source code, our approach keeps the tested application intact by instrumenting at the model level, allowing it to retain its initial behavior