Vulnerabilities in Ethereum Smart Contracts : An Analysis of Testing and Verification Techniques

This bachelor’s thesis investigates the critical need for robust testing and verification techniques for smart contracts, which play a key role in blockchain ecosystems. Despite previous advances in tool development and vulnerability identification, consistent benchmarking for evaluating tool effectiveness remains limited. This study builds on existing research by systematically analyzing current methodologies and proposing recommendations to strengthen smart contract security. The investigation begins with an overview of smart contract architecture and their inherent vulnerabilities, particularly those stemming from properties such as immutability, decentralization, and autonomous execution, which hinder post-deployment error correction. The thesis evaluates a range of testing and verification approaches, including fuzzing, symbolic execution, systematic modeling, and machine learning-based methods. In addition, it assesses the effectiveness of selected analysis tools in real-world scenarios using a defined set of benchmarking criteria, focusing on detection accuracy, resource efficiency, and code coverage. The findings comprise a categorized overview of common vulnerabilities, an empirical evaluation of prominent tools, and actionable recommendations for improving testing practices. By introducing practical benchmarks and insights, this research contributes to the development of more reliable smart contracts and supports the advancement of secure blockchain infrastructures.