Home // ICAS 2015, The Eleventh International Conference on Autonomic and Autonomous Systems // View article

Autonomic Metaheuristic Optimization with Application to Run-Time Software Adaptation

Authors:
John M. Ewing
Daniel A. Menascé

Keywords: Intelligent systems; Autonomous agents; Evolutionary computation; Genetic algorithms

Abstract:
This paper presents a general meta-optimization approach for improving self-optimization in autonomic systems. This approach can improve optimization performance and lower costs by reducing human effort needed to tune optimization algorithms. We apply our meta-optimization approach to Self-Architecting Software Systems (SASSY). A genetic algorithm is used to meta-optimize both the architecture search module and the service selection search module in SASSY. Four different heuristic search algorithms (hill-climbing, beam search, evolutionary programming, and simulated annealing) are made available to be meta-optimized in both the architecture search module and the service selection search module. This meta-optimization process generated twelve new heuristic search algorithm pairs for solving SASSY optimization problems. In a large set of simulation experiments, two of the generated heuristic search algorithm pairs provided superior performance to the control (which was the previously best heuristic search algorithm pair known in SASSY).

Pages: 63 to 72

Copyright: Copyright (c) IARIA, 2015

Publication date: May 24, 2015

Published in: conference

ISSN: 2308-3913

ISBN: 978-1-61208-405-3

Location: Rome, Italy

Dates: from May 24, 2015 to May 29, 2015