OpenForensics: A digital forensics GPU pattern matching approach for the 21st century

Pattern matching is a crucial component employed in many digital forensic (DF) analysis techniques, such as file-carving. The capacity of storage available on modern consumer devices has increased substantially in the past century, making pattern matching approaches of current generation DF tools increasingly ineffective in performing timely analyses on data seized in a DF investigation. As pattern matching is a trivally parallelisable problem, general purpose programming on graphic processing units (GPGPU) is a natural fit for this problem. This paper presents a pattern matching framework - OpenForensics - that demonstrates substantial performance improvements from the use of modern parallelisable algorithms and graphic processing units (GPUs) to search for patterns within forensic images and local storage devices.