Date of Award

2014

Document Type

Thesis

First Advisor

Yuan, Xiaohong Dorothy

Abstract

To engineer secure software, it is imperative to understand attackers’ perspectives and approaches. This information has been captured by attack patterns. The Common Attack Patterns Enumeration Classification (CAPEC) repository hosts over 450 attack patterns that contain information about how attacks have been launched against software. Researches have indicated that attack patterns can be utilized for developing secure software; however, there exists no systematic methodology to address this concern. This research proposes a methodology for utilizing CAPEC attack patterns for developing abuse cases at the requirements stage of the secure software development lifecycle (SDLC). In previous research, a tool for retrieving attack patterns (TrAP) was developed to retrieve CAPEC attack patterns according to Microsoft STRIDE threat categories. This tool also features a search function using keywords. The proposed methodology starts with a set of initial abuse cases developed through brainstorming. Microsoft SDL threat modelling tool is then used to identify and rank possible security threats in the system. The SDL tool generates a series of questions for each threat and these questions are used to extract keywords that serve as input to the TrAP tool to retrieve attack patterns relevant to the abuse cases. Keywords can also be system prerequisites or any technology being implemented in the system. From the list of retrieved attack patterns, the most relevant attack patterns are selected and used to extend the initial abuse cases. New abuse cases can also be discovered through this process.

Share

COinS