Projects Using KUSP

The KUSP software has been developed over many years, in the contet of many different projects, and has been rewritten, revised, and ported to new versions of Linux many times. A continuing theme throught this history, however, has been a constant improvement in (1) the precision with which computation behavior can be described, (2) the precision with which described behavior can be controlled and ensured during execution, and (3) the precision with which execution behavior can be measured and analyzed. At this point, the KUSP software provides a number of interesting capabilities and can be useful for a number of different purposes.

Current Projects

Semantic Domain Integration for Embedded and Hybrid Systems :

Application software intended to run in a computer system can be viewed under three major semantic domains: it can be formally described; it is specified as software under a given programming model, transformation into an executable form, and executes according to how it is controlled by the system; finally, its behavior can be measured at run-time. Most commonly, these three views of an application are considered separately, which causes there to be less correspondence between the formal modeling, the execution behavior and the measured execution behavior than is desirable. This project considers how the reatment of these three semantic domains can be integrated to create a stronger correspondence between modeled behavior, execution behavior, and mesured behavior. The result is intended to be a system in which application QoS requirements, system software QoS semantics, resource management, and behavioral information are integrated through (1) mutually consistent formal and verifiable models of each semantic domain; (2) novel policies and mechanisms for exerting precise run-time control across semantic domains; and (3) detailed, efficient, and timely collection and dissemination of behavioral information to improve run-time control fidelity. Through rigorous integration of these semantic domains we aim to achieve a much greater correspondence among their respective semantics, and establish a foundation for revolutionary improvements in the state of the art, particularly for increases in system accuracy and reliability in producing desired behaviors (and in preventing undesired behaviors) in complex mission-critical engineered systems.

Non-Bypassable Kernel Services :

Society is dependent on many engineered systems whose increasing complexity and inter-connectedness have, in turn, increased their vulnerability to adversarial attacks. In many of these systems, protecting the execution of their computations is as crucial as ensuring the security of their data. This research investigates how to maintain survivable operation of such systems, even in the face of invasive attacks where computations are intentionally subverted to interfere with other computations’ execution constraints. THis research is done in collaboration with researchers at Washington University, who concentrate models of application software development to describe behavior, or to look for signatures of security problems in system behavioral data. At the University of Kansas we concentrate on the portion of the problem related to precise control and measurement of execution behavior. These two approaches will converge and synergize, incresing the dgreee to which we can provide execution security to applications through increased control precision, as well as increasing our ability to detect challenges to execution security by observing many details of system behavior. a prominenet part of the KU based research is the ability to use a stream of events collected from the system that describe actions by valrios threads to discover the full set of components comprising a computation, including all system or server components that even the authors of the application software may be unaware are required for the proper execution of the application. We consider discovery of computation structure fundamental to exeuction security because it is fundamentally impossible to fully control a computation unless you are ware of all of its components.

Historical Projects

A wide range of projects have used KUSP software for support to achieve specific goals, concentrated on imporiving the system capabilities, and contributed to the development of the software and the range of applications that it can support.

  • ATM Reference Traffic System
  • KU Real-TIme (KURT) Linux
    • Utime - High resolution event time keeping
    • Explicit Plan Scheduling
  • DARPA ANTS Project
  • DARPA PCES Project

Table Of Contents

Previous topic

Subsystems Comprising KUSP

Next topic

Non-bypassable Kernel Services for Execution Security

This Page