Date of Award


Document Type


Degree Name

Master of Science in Computer Engineering


Department of Electrical and Computer Engineering

First Advisor

Jeffrey W. Humphries, PhD


Virtualization is a technique used to model and simulate the cyber domain, as well as train and educate. Different types of virtualization techniques exist that each support a unique set of benefits and requirements. This research proposes a novel design that incorporates host and network virtualization concepts for a cyber warfare training platform. At the host level, hybrid virtualization combines full and operating system virtualization techniques in order to leverage the benefits and minimize the drawbacks of each individual technique. Network virtualization allows virtual machines to connect in flexible topologies, but it also incurs additional processing overhead. Quantitative analysis falls into two sets of experiments. The first set of experiments evaluates traditional virtualization techniques against the hybrid approach. Results indicate that in some cases, performance of hybrid virtualization exceeds that of full virtualization alone while still providing an identical feature set. The second set of experiments examines the amount of overhead involved with network virtualization with respect to bandwidth and latency. Results indicate that performance over a local area network incurs two to four times the performance cost compared to physical connections. The benefit of this additional overhead is an increased flexibility in defining network topologies at the software level independent of the underlying physical topology.

AFIT Designator


DTIC Accession Number



Alternate title on SF298: Designing a Hybrid Virtualization Platform Design for Cyber Warfare and Simulation