Implementation and Evaluation of a Spectrogram-based Methodology for Instruction and Memory Trace Analysis
Master’s Thesis / Student Research Project
Abstract
Abstract modelling of HW/SW systems is a fairly new research topic. The goal of this technique is it to capture only the essential parameters of software and hardware which influence their timing behavior. Memory accesses are a very important aspect as the amount of memory accesses as well as the memory-hierarchy-level on which data is located can have a significant impact on a software’s timing behavior.
Goal of this student project is to implement and evaluate a new spectrogram-based methodology with which an instruction and memory trace can be analyzed in order to create a trace-specific signature. Furthermore it is to be evaluated to what ends such a signature could be used in the context of very early software modeling and timing analysis.
Requirements
- Python
- Basic knowledge of signal analysis and FFT
- Willingness to explore a new area of research, come up with own ideas and take inspiration from other fields in science.