Task mapping and priority assignment for soft real-time applications under deadline miss ratio constraints

Abstract

Both analysis and design optimisation of real-time systems has predominantly concentrated on considering hard real-time constraints. For a large class of applications, however, this is both unrealistic and leads to unnecessarily expensive implementations. This paper addresses the problem of task priority assignment and task mapping in the context of multiprocessor applications with stochastic execution times and in the presence of constraints on the percentage of missed deadlines. We propose a design space exploration strategy together with a fast method for system performance analysis. Experiments emphasize the efficiency of the proposed analysis method and optimisation heuristic in generating high-quality implementations of soft real-time systems with stochastic task execution times and constraints on deadline miss ratios. Manolache, S., Eles, P., and Peng, Z. 2008. Task mapping and priority assignment for soft realtime applications under deadline miss ratio constraints.

Scheduling Access to Shared Space in Multi-robot Systems

“…Some of them modified existing deterministic algorithms to support variable execution times, e.g., [10]. Others have implemented heuristic approaches to schedule tasks with stochastic execution times on multi-processors [9]. All these works, similarly to ours, have in common the use of probability distributions to characterize the stochastic execution time of the tasks.…”

A hierarchical run-time adaptive resource allocation framework for large-scale MPSoC systems

“…To overcome the shortcomings of pure static [4,11,15] and dynamic [21,36] task mapping algorithms, hybrid (semi-static) approaches have become increasingly popular in recent years. Usually, in this kind of approaches, multiple mapping solutions are found at design time and applied at run time based on the current state of the system.…”

Robust Scheduling of Task Graphs under Execution Time Uncertainty

“…FPS scheduler: The considered FPS method is a variant of the one presented in [44], based on Tabu-search and performing both mapping and scheduling. Since the focus of this paper is on pure scheduling, we force the mapping from Section V-B and use tabu-search only for priority assignment.…”

“…Since the focus of this paper is on pure scheduling, we force the mapping from Section V-B and use tabu-search only for priority assignment. Moreover, we have extended the approach to deal with non-unary resources 12 : in [44] there is no such issue since only single-thread processors are taken into account. Finally, we replaced per-resource priority lists in [44] with a single global one; this is necessary to avoid conflicting priority assignments, since tasks in our model can have multiple resource requirements 13 .…”

“…Estimating the completion time with FPS requires more complex techniques compared to our approach. In [44] a clever approximation method is used to compute end time distributions for each task; the computation is fast enough to be embedded in the Tabu-search optimization, although (by far) slower than the simple proxy we adopt in this work; on the other hand, the accuracy of the estimate is much higher 14 . An advantage of FPS scheduling over our method is the ability to always start the highest priority task as soon as it is possible.…”