Prediction performance method for dynamic task scheduling, case study: the OLLAF Architecture

Abstract

Actual dynamic applications, executed on real-time systems, have the tendency to be built on dynamically reconfigurable hardware devices. These applications require high performance and flexibility towards user and environment needs. To perform these application requirements, efficient mechanisms to manage hardware device must exist. In this paper we target OLLAF as a dynamically reconfigurable architecture which is designed to support complex and flexible applications. In order to deal with all of the dynamic aspects of such systems, we describe a predictive scheduling allowing an early estimation of our application dynamicity. A vision system of a mobile robot and an application of 3D synthesis images were served to validate the presented scheduling approach.