A multi-dimensional mimic voting method based on process elements

Abstract

This invention provides a unified control algorithm for executing bodies in a heterogeneous virtual machine platform for mimicry defense, aiming to address the lack of standardized algorithm definition and implementation issues in the dynamic scheduling of executing bodies. The algorithm considers both local and global heterogeneity, assuming that the probability of identical vulnerabilities among equivalent heterogeneous entities can be neglected. It defines and standardizes optimal local replacements based on Manhattan distance for local heterogeneity. For global heterogeneity, it utilizes information entropy to resolve the trade-off between local and global optimal solutions. Designed with a targeted approach, the algorithm prioritizes maximizing local heterogeneity and subsequently global heterogeneity, fundamentally enhancing the defense capability of mimicry defense systems.

Images