A unifying framework for dynamic monitoring and a taxonomy of optimizations

Publikationstyp
Conference Paper
Date Issued
2020-10-27
Sprache
English
Author(s)
Jakobs, Marie-Christine  
Mantel, Heiko 
TORE-URI
http://hdl.handle.net/11420/13826
First published in
Lecture notes in computer science  
Number in series
12477 LNCS
Start Page
72
End Page
92
Citation
Lecture Notes in Computer Science 12477 LNCS: 72-92 (2020)
Contribution to Conference
9th International Symposium on Leveraging Applications of Formal Methods, Verification and Validation, ISoLA 2020  
Publisher DOI
10.1007/978-3-030-61470-6_6
Scopus ID
2-s2.0-85096541871
Publisher
Springer International Publishing AG
Reducing the performance overhead of run-time monitoring is crucial for making it affordable to enforce more complex requirements than simple security or safety properties. Optimizations for reducing the overhead are becoming increasingly sophisticated themselves, which makes it mandatory to verify that they preserve what shall be enforced. In this article, we propose a taxonomy for such optimizations and use it to develop a classification of existing optimization techniques. Moreover, we propose a semantic framework for modeling run-time monitors that provides a suitable basis both, for verifying that optimizations preserve reliable enforcement and for analytically assessing the performance gain.
DDC Class
004: Informatik