An Unmanned Aerial Vehicle (UAV) is often an aircraft with no crew that can fly independently by a preprogrammed plan, or by remote control. Several UAV applications, like autonomously surveillance and traffic monitoring, are real-time applications. Hence tasks in these applications must complete within specied deadlines.
Real Time Calculus (RTC) is a formal framework for reasoning about realtime systems and in particular streaming applications. RTC has its mathematical roots in Network Calculus. It supports timing analysis, estimating loads and predicting memory requirements.
In this thesis, a formal analysis of real-time stream reasoning for UAV applications is conducted. The performance analysis is based on RTC using an abstract performance model of the streaming reasoning on board a UAV. In this study, we consider two dierent scheduling methods, first-in-first-out (FIFO) and fixed priority (FP). In the FIFO scheduling model the priorities of the tasks are assigned and processed based on the order of their arrival, while in the FP scheduling model the priorities of the tasks are preassigned. The Quality of Service (QoS) of these applications is calculated and analyzed in a proposed design space exploration framework.
QoS can be defined dierently depending on what field we are studying and in this thesis we are interested in studying the delays of the real-time stream reasoning applications when (i) we fix jitters and number of instances and vary the periods, (ii) we fix the periods and number of instances and vary the jitters, and (iii) we fix the periods, jitters and vary the number of instances.