A systematic model for iterative deduction of reverse engineering

Abstract

Reverse engineering constitutes a fundamental pillar of cybersecurity, digital preservation, and compatibility engineering, with historical precedents dating to World War II cryptanalysis efforts that precipitated the development of early computational machines including Alan Turing’s Bombe. Despite its critical role in malware analysis, vulnerability research, legacy system maintenance, protocol reconstruction, and cultural preservation, reverse engineering has received limited systematic academic treatment. The field is frequently characterised as an intuitive craft requiring “wizardry” rather than a rigorous, teachable discipline , resulting in a significant methodological gap wherein practitioners rely predominantly on tacit knowledge, ad-hoc techniques, tool-specific workflows, and individual expertise rather than generalisable theoritical frameworks. This informal and unregulated practice leads to variability in approaches, inconsistent results, and barriers to knowledge transfer between practitioners.

This thesis addresses this gap by proposing a systematic, domain-independent model for reverse engineering that formalises the iterative process of knowledge acquisition from black-box systems though structured observation of component-environment interactions. We present a formal framework that abstracts reverse engineering into an interactive cycle comprising six core phases: (0) initial model construction based on available knowledge, (1) model instantiation to create observable instances, (2) testing and trace generation through controlled interaction, (3) knowledge synthesis combining new observations with existing understanding, (4) model revision reflecting refined knowledge, and (5) iteration until knowledge reaches the desired completeness threshold. This methodology structures reverse engineering as the systematic observation and analysis of behavioural traces across controlled environmental variations, transcending domain-specific tools and techniques to provide a unified epistemological foundation applicable across diverse reverse engineering contexts.

More information

Identifiers

ORCID for Max Dylan Hayman: orcid.org/0000-0003-4825-810X

ORCID for Vladimiro Sassone: orcid.org/0000-0002-6432-1482

Catalogue record

Export record

Altmetrics