Real Time Tasks Scheduling Optimization Using Quantum Inspired Genetic Algorithms

Abstract

Real Time Scheduling (RTS) optimization is a key step in Real Time Embedded Systems design flow. Since RTS is a hard problem especially on multiprocessors systems, researchers have adopted metaheuristics to find near optimal solutions. On the other hand, a new class of genetic algorithms inspired from quantum mechanics appeared and proved its efficiency with regard to conventional genetic algorithms. The objective of this work is to show how we can use quantum inspired genetic algorithm to resolve the RTS problem on embedded multicores architecture. Our proposed algorithm tries to minimize the tasks response times mean and the number of tasks missing their deadlines while balancing between processors cores usage ratios. Experimental results show a big improvement in research time with regard to conventional genetic algorithms.