Dieses Bild zeigt Maik Betka

Maik Betka

Herr

Wissenschaftlicher Mitarbeiter, Doktorand
ISTE - Institut für Software Engineering
ESE - Empirisches Software Engineering

Kontakt

+49 711 685 88352
+49 711 685 88380

Universitätsstraße 38
70569 Stuttgart
Deutschland
Raum: 1.024

I am looking for master's students. If you are interested in one of the following topics, in the area of software testing, then please write me an email. On interest, I will send you the full proposals with detailed descriptions and references. If you consider one of the topics suitable for you then we will set up a Webex meeting to discuss the process. Looking forward for your message.

Topic: Generating Property-based Tests for Fuzzing
Examiner: Prof. Dr. Stefan Wagner 
Institute, research group: Institute of Software Engineering, Empirical Software Engineering Group
Short description: Test input generators are powerful in exploring different program behavior but take noticable effort to write. Property-based tests tackle the test oracle problem by testing general properties of a piece of software instead of exact test outputs. However, these properties also have to be manually defined. This project targets to enhance existing software test suites by creating test input generators and property-based tests from existing unit tests. The created generators and properties shall be utilized by the JQF fuzz testing platform for Java with the Zest fuzzing algorithm. For that, existing software test suites of open source projects shall be analyzed and different approaches to extract generators and properties shall be evaluated.

Topic: Fuzzing Towards Non-Functional Properties
Examiner: Prof. Dr.-Ing. Steffen Becker
Institute, research group: Institute of Software Engineering, Software Quality and Architecture Group
Short description: Fuzzing techniques usually have code coverage guided feedback to explore different program behavior. Maximizing particular feedback metrics, like branch counters, can be used to guide input generation towards executing a program with worst-case complexity. Similar metrics for memory allocations exist that lead to heavy memory consumption of the software under test. This project targets to validate these domain-specific fuzzing concepts in Java. For that, the JQF fuzz testing platform shall be extended to support fuzzers that are built with FuzzFactory. The results should be validated by running several fuzzing campaigns on multiple open source projects.

Zum Seitenanfang