Efficient input space exploration for falsification of cyber-physical systems
Date
2022
Authors
Savaliya, Meetkumar, author
Prabhu, Vinayak, advisor
Pasricha, Sudeep, committee member
Ghosh, Sudipto, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
In recent years black-box optimization based search testing for Signal Temporal Logic (STL)specifications has been shown to be a promising approach for finding bugs in complex Cyber Physical Systems (CPS) that are out of reach of formal analysis tools. The efficacy of this approach depends on efficiently exploring the input space, which for CPS is infinite. Inputs for CPS are defined as functions from some time domain to the domain of signal values. Typically, in black-box based testing, input signals are constructed from a small set of parameters, and the optimizer searches over this set of parameters to get a falsifying input. In this work we propose a heuristic that uses the step response of the system – a standard system characteristic from Control Engineering – to obtain a smaller time interval in which the optimizer needs to vary the inputs, enabling the use of a smaller set of parameters over which the optimizer needs to search over. We evaluate the heuristic on three complex Simulink model benchmarks from the CPS falsification community, and we demonstrate the efficacy of our approach.