We propose a dynamic, decentralized scheduling strategy which leads to load balancing in a large distributed environment. The main activities are agent partitioning and agent migration performed according to local, diffusion-based rules. The strategy is applied to the agent-based CAE mesh generator, and the solution is tested on a network of PC desktops. We discuss the total computation time, network coverage, and different scheduling properties. © Springer-Verlag Berlin Heidelberg 2004.
CITATION STYLE
Grochowski, M., Schaefer, R., & Uhruski, P. (2004). Diffusion based scheduling in the agent-oriented computing system. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3019, 97–104. https://doi.org/10.1007/978-3-540-24669-5_13
Mendeley helps you to discover research relevant for your work.