Title
The Rate-Based Execution Model
Keyword
I want to present a task model for the real-time execution of
event-driven tasks in which no priori characterization of the actual arrival
rate of events is known,only the expected arrival rate for event is known.This
task model is called rate-based execution(RBE),the RBE model is motivated
natually by distributed multimedia and digital signal processing
application
Background and outline
1. Background:
Real-time application frequently interact with external devices in an
event-driven manner.The delivery of a message ,or a hardware interrupt ,signals
the event arrival.A task is then dispached to process the event.In real-time
environments ,the system provides some form of guarantee that the task will
finish processing the event within d time units of the event occurrence.where di
is called the relative deadline of event ei,Hard-real-time systems guatantee
that every event ei will be processed within di time units of its
occurrence.Soft-real-time systems provids weaker guarantees of
timeliness
2.
Outline:
(1). RBE Task Model
(2).
Feasibility of RBE Tasks
(3). Discussion
(4). Summary and
Conclusion
Approach
1.The starting point for my work is to study the model of sporadic
tasks developed by Mok,and later extended by Baruah and Jeffay .
2.Alter the
formal model to account by extending the sporadic task to be a task that
executes at an average rate.
3.The RBA task model is most similar to the
linear-bounded arribal process(LBAP)model used in the DASH system.I want to
generalizes the LBAP model to include a more generic specification of rate and
adds an independent response time parameter to enable more precise real-time
control of task execution,
4.Analyze the model in more complex environments
such as those wherein tasks communicate vis memory and thus have preemption
constraints.
Current progress of
research
My research plan is as
follows:
1. I will finish studying the model of sporadic tasks in 2
weeks.
2. Study the LBAP model for one week.
3. Alter the formal model to
account for the requirement in 3 weeks.
4. Analyze the model in more complex
environments
Now I am under the study of the model of sporadic tasks.And may
finish it this week and go on with the LBAP model study.
I will try to
finish all the plans on time and get the conclusion till the end of
July.