A System for Analyzing Diversified Software Binaries

In recent years, software diversification has become a viable security assurance measure to prevent the proliferation of reusable exploits. Lacking, however, are tools that focus on comparative analysis of diversification techniques on memory usage and CPU usage, as well as energy consumption, along with metrics associated with the persistence of return oriented programming ROP gadgets across diversified binaries. In this paper, we present an open source tool-set system to streamline binary diversification. In addition, we use our system to diversify several Linux binaries, and present an analysis of the impacts of several diversification algorithms on CPU clock cycles and on the total number of ROP gadgets between variants.