In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
will defend his dissertation
Schedulability, Response Time Analysis and New Models of P-FRP Systems
Functional Reactive Programming (FRP) is a declarative approach for modeling and building reactive systems.
FRP has been shown to be an expressive formalism for building applications of computer graphics, computer vision, robotics, etc. Priority-based FRP (P-FRP) is a formalism that allows preemption of executing programs and guarantees real-time response. Since functional programs cannot maintain state and mutable data, changes made by programs that are preempted have to be rolled back. Hence in P-FRP, a higher priority task can preempt the execution of a lower priority task, but the preempted lower priority task will have to restart after the higher priority task has completed execution. This execution paradigm is called Abort-and-Restart (AR).
Current real-time research is focused on preemptive of non-preemptive models of execution and several state-of-the-art
methods have been developed to analyze the real-time guarantees of these models. Unfortunately, due to its transactional nature where preempted tasks are aborted and have to restart, the execution semantics of P-FRP does not fit into the standard definitions of preemptive or non-preemptive execution, and the research on the standard preemptive and non-preemptive may not applicable for the P-FRP AR model. Out of many research areas that P-FRP may demands, we focus on task scheduling which includes task and system modeling, priority assignment, schedulability analysis, response time analysis, improved P-FRP AR models, algorithms and corresponding software.
In this work, we review existing results on P-FRP task scheduling and then present our research contributions: (1) a tighter feasibility test interval regarding the task release offsets as well as a linked list based algorithm and implementation for scheduling simulation; (2) P-FRP with software transactional memory-lazy conflict detection (STM-LCD); (3) a non-work-conserving scheduling model called Deferred Start; (4) a multi-mode P-FRP task model; (5) SimSo-PFRP, the P-FRP extension of SimSo - a SimPy-based, highly extensible and user friendly task generator and task scheduling simulator.
Date: Thursday, June 22, 2017
Time: 10:00 AM
Place: PGH 550
Advisor: Dr. Albert M. K. Cheng
Faculty, students, and the general public are invited.