Virtual Shield: A Novel Approach to Defending Cyber Attacks

ABSTRACT:In this proposed effort, we propose a virtualization-based framework, called virtual shield, to protect the end systems in an enterprise network. Virtual Shield follows the diversity, moving target defense and fight-through strategies to defend against cyber attacks. The Virtual Shield architecture adopts various resilient techniques to make the protection approaches more agile and adaptive to cyber attacks. Virtual Shield also employs a virtual machine introspection technique to monitor the execution of an end system and detect malware without deploying agents in the monitored system. Moreover, Virtual Shield fully utilizes virtualization to implement the capabilities of freezing, duplicating and moving an end system in the network, thus making moving target defense feasible. Finally, these proposed techniques will be integrated in a workable framework prototype to show the feasibility of our proposed approach.BENEFIT:The proposed solution is well-suited for application protection in an enterprise network. Both commercial and military markets will benefit from the development of the proposed solution for infrastructure security, particularly in a network environment under advanced cyber attacks. In addition, it can also be applied to protect valuable data asset or mission critical services, providing a fight-through protection and failover. The proposed solution also has great commercial market potential. DoD efforts to consolidate and reduce data center via virtualization are anticipated to save $3 billion annually by 2015. The global cyber security market is growing at 11.3% CAGR and is estimated to reach $120.1B by 2017. The addressable market size for the proposed solution is estimated to be more than $200+ million during the first 10 years of commercialization. Potential end users include security tool providers (e.g. Symantec, McAfee, SourceFire); Banks (e.g., Bank of America); and virtualization software developers (e.g., VMWare, Microsoft).