Hardware Reference
In-Depth Information
9.4.2
PERIOD ADAPTATION
There are several real-time applications in which timing constraints are not rigid, but
depend on the system state. For example, in a flight control system, the sampling
rate of the altimeters is a function of the current altitude of the aircraft: the lower the
altitude, the higher the sampling frequency. A similar need arises in mobile robots
operating in unknown environments, where trajectories are planned based on the cur-
rent sensory information. If a robot is equipped with proximity sensors, to maintain
a desired performance, the acquisition rate of the sensors must increase whenever the
robot is approaching an obstacle.
The possibility of varying tasks' rates also increases the flexibility of the system in
handling overload conditions, providing a more general admission control mechanism.
For example, whenever a new task cannot be guaranteed, instead of rejecting the task,
the system can reduce the utilizations of the other tasks (by increasing their periods in
a controlled fashion) to decrease the total load and accommodate the new request.
In the real-time literature, several approaches exist for dealing with an overload through
a period adaptation. For example, Kuo and Mok [KM91] propose a load scaling tech-
nique to gracefully degrade the workload of a system by adjusting the periods of pro-
cesses. In this work, tasks are assumed to be equally important and the objective
is to minimize the number of fundamental frequencies to improve schedulability un-
der static priority assignments. Nakajima and Tezuka [NT94] show how a real-time
system can be used to support an adaptive application: whenever a deadline miss is
detected, the period of the failed task is increased. Seto et al. [SLSS96] change tasks'
periods within a specified range to minimize a performance index defined over the
task set. This approach is effective at a design stage to optimize the performance of a
discrete control system, but cannot be used for online load adjustment. Lee, Rajkumar
and Mercer [LRM96] propose a number of policies to dynamically adjust the tasks'
rates in overload conditions. Abdelzaher, Atkins, and Shin [AAS97] present a model
for QoS negotiation to meet both predictability and graceful degradation requirements
during overloads. In this model, the QoS is specified as a set of negotiation options in
terms of rewards and rejection penalties. Nakajima [Nak98] shows how a multimedia
activity can adapt its requirements during transient overloads by scaling down its rate
or its computational demand. However, it is not clear how the QoS can be increased
when the system is underloaded. Beccari et al. [BCRZ99] propose several policies for
handling overload through period adjustment. The authors, however, do not address
the problem of increasing the task rates when the processor is not fully utilized.
Search WWH ::
Custom Search