Defense Date

2024

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Electrical & Computer Engineering

First Advisor

Robert Klenke, PhD

Second Advisor

Carl Elks, PhD

Third Advisor

Smitha Gautham, PhD

Fourth Advisor

Milos Manic, PhD

Fifth Advisor

Irfan Ahmed, PhD

Abstract

In the past two decades, Unmanned Aerial Systems have progressed from expensive military hardware or one-off custom builds, to include off-the-shelf drones that can be purchased for a rather affordable price and flown by nearly anyone. As the technology and performance have improved, the door is opened to applications that require operation in environments where the consequences for failure are high, such as operating in the navigable airspace or in urban environments, or with human passengers. This requires a great deal of trust in the reliability and integrity of the control systems of the aircraft. A method of monitoring the execution of autopilot software is described, which detects incorrect execution caused by software bugs or cyber-attacks using common hardware interfaces on an existing autopilot. An external monitor using existing buses allows execution monitoring capabilities to be added without the expense of a custom processor with built-in monitoring hardware. The implemented design makes use of debug hardware within the autopilot processor to make instrumentation lightweight. Monitoring is conducted externally using a System On a Chip (SOC) that integrates an FPGA with an ARM processor. The value of any execution monitor is limited by its adoption; A framework which is difficult to write specifications in is less likely to provide much benefit. The use of regular expressions as the language for the primary monitor type that makes up this work adds to its value as they are already well known and commonly used by programmers.

Rights

© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

12-5-2024

Download

Share

COinS