Analysis of binary code topology for dynamic analysis

Abstract

A better understanding of binary code topology is essential in making execution path exploration method. The execution path exploration is closely related to code coverage in binary code dynamic analysis. Since the number of execution paths in a program is astronomically high, the efficient exploration strategy is needed. In this paper, we analyze binary code topology in a viewpoint of basic blocks. We find that the incoming edges show unbalanced distribution which follows power law, instead of balanced distribution. This unbalanced distribution of incoming edges can help understanding binary code topology, and we propose our study for deciding efficient execution path exploration strategy of dynamic binary code analysis.