Thunderlane Phase II

The Thunderlane Phase II effort defines a new hypervisor model that dissolves the boundary between hypervisor and operating system, and leverages state-of-the-art software diversity techniques for increased resistance to exploitation. The most widely used operating systems (e.g., Microsoft Windows and Linux) use monolithic kernel architectures that execute most system software with full root-level privileges. In contrast, microkernel architectures, which disaggregate the kernel into lower-privileged segments, are known for their stability and resistance to exploitation. Unfortunately, microkernels are not widely used, and often do not receive the development attention necessary to support mission critical hardware / software. The Thunderlane effort aims to solve these problems by leveraging hardware-based virtualization to define a new hybrid hypervisor / operating system architecture: the hyperkernel. Like a microkernel, each major component in the hyperkernel system executes in isolation, and communicates via well-defined channels that can be tested and verified. Unlike a microkernel, however, the hyperkernel can leverage existing software designed for monolithic kernels like Microsoft Windows and Linux providing both the reliability and security of a microkernel, with access to existing mission critical software. Furthermore, Thunderlane will use state-of-the-art compile-, load- and run-time software diversity techniques (i.e. moving target defense) to protect the hyperkernel.