Parallelization of the Scale-Changing Technique in Grid Computing environment for the electronmagnetic simulation of multi-scale structures

✍ Scribed by F. Khalil; C. J. Barrios-Hernandez; A. Rashid; H. Aubert; Y. Denneulin; F. Coccetti; R. Plana

• Publisher: John Wiley and Sons
• Year: 2010
• Tongue: English
• Weight: 564 KB
• Volume: 24
• Category: Article
• ISSN: 0894-3370
• DOI: 10.1002/jnm.760

✦ Synopsis

A parallel computing approach to run fast and full-wave electromagnetic simulation of complex structures in Grid Computing environment is presented. In this study, we show how Grid Computing improves speed and/or reliability over that provided by a single computer, while typically being much more cost-effective than single computers of comparable speed or reliability. An efficient monolithic (unique) formulation for the electromagnetic modelling of complex (multi-scale) structures, i.e. structures that exhibit multiple metallic patterns whose sizes cover a large range of scales, is used here. This approach, named the Scale-Changing Technique, is based on the cascade of multi-modal Scale-Changing Networks, each network modelling the electromagnetic coupling between two successive scale levels. These networks can be first computed separately, in an adaptive use of Grid Computing architecture nature, and then cascaded for the global electromagnetic simulation. Based on this technique, a fast computer algorithm was developed and tested in the Grid-Computing environment. For illustration purposes, the electromagnetic analysis of multi-scale structures, applied to phase-shifter elements and an example of infinite passive reflectarray, was carried out. The obtained results have confirmed the effectiveness of such an approach compared with sequential computing. This approach shows very good computation performance while keeping the same accuracy. Besides, this method is very promising for optimizing circuit with multiple design parameters to handle and for the global electromagnetic simulation of multi-scale and/ or oer-sized structures.