Dataflow Analysis for Known Vulnerability Prevention System

Usually, a new exploit for a single vulnerability can readily be turned into worms which compromise hundreds of thousands of machines within only a few minutes. In order to protect the host from malicious attacks, we propose a new approach for automatic defense mechanism: dataflow analysis for known vulnerability prevention system (shortly for DA-VPS), which has properties with easy deployment, accurate detection and low overhead. In our paper, we present the principle, architecture, implement, and deployment of DA-VPS.