This paper shows a method targeted at software execution interpretation and fault detection using recurrence plot analysis. In in the proposed approach recurrence plot analysis is applied to software execution trace that contains executed assembly instructions. Results of this analysis are subject to further processing with PCA (Principal Component Analysis) method that simplifies number coefficients used for software execution classification. This method was used for the analysis of five algorithms: Bubble Sort, Quick Sort, Median Filter, FIR, SHA-1. Results show that some of the collected traces could be easily assigned to particular algorithms (logs from Bubble Sort and FIR algorithms) while others are more difficult to distinguish.
© (2015) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Citation
M. Mosdorf
"
Using recurrence plot analysis for software execution interpretation and fault detection
", Proc. SPIE 9662, Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 2015, 96623X (September 11, 2015); doi:10.1117/12.2205963; http://dx.doi.org/10.1117/12.2205963