High performance computing on scattering from extremely large targets by MLFMA

Xiao Min Pan; Xin Qing Sheng

doi:10.3969/j.issn.1001-506X.2011.08.02

High performance computing on scattering from extremely large targets by MLFMA

Xiao Min Pan^*, Xin Qing Sheng

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

5 Citations (Scopus)

Abstract

Difficulties associated with extremely large targets, including construction of the geometry tree in the algorithm, reduction of the number of far-field patterns and fast computation of radar cross section, are pointed out. Based on the architecture of popular parallel platforms, distributed memory parallel systems, and properties of targets with extremely large sizes, relative solutions are proposed. The efficiency of the proposed parallel multilevel fast multipole algorithm (MLFMA) scheme is validated by successful computations of scattering from targets with over five hundred million unknowns and several thousand wavelengths in size.

Original language	English
Pages (from-to)	1690-1693
Number of pages	4
Journal	Xi Tong Gong Cheng Yu Dian Zi Ji Shu/Systems Engineering and Electronics
Volume	33
Issue number	8
DOIs	https://doi.org/10.3969/j.issn.1001-506X.2011.08.02
Publication status	Published - Aug 2011

Keywords

High performance
Multilevel fast multipole algorithm (MLFMA)
Parallel
Radar cross section (RCS)
Scattering

Access to Document

10.3969/j.issn.1001-506X.2011.08.02

Cite this

@article{f9ebccad413f47d2983de6ac8127a1e4,

title = "High performance computing on scattering from extremely large targets by MLFMA",

abstract = "Difficulties associated with extremely large targets, including construction of the geometry tree in the algorithm, reduction of the number of far-field patterns and fast computation of radar cross section, are pointed out. Based on the architecture of popular parallel platforms, distributed memory parallel systems, and properties of targets with extremely large sizes, relative solutions are proposed. The efficiency of the proposed parallel multilevel fast multipole algorithm (MLFMA) scheme is validated by successful computations of scattering from targets with over five hundred million unknowns and several thousand wavelengths in size.",

keywords = "High performance, Multilevel fast multipole algorithm (MLFMA), Parallel, Radar cross section (RCS), Scattering",

author = "Pan, {Xiao Min} and Sheng, {Xin Qing}",

year = "2011",

month = aug,

doi = "10.3969/j.issn.1001-506X.2011.08.02",

language = "English",

volume = "33",

pages = "1690--1693",

journal = "Xi Tong Gong Cheng Yu Dian Zi Ji Shu/Systems Engineering and Electronics",

issn = "1001-506X",

publisher = "Chinese Institute of Electronics",

number = "8",

}

TY - JOUR

T1 - High performance computing on scattering from extremely large targets by MLFMA

AU - Pan, Xiao Min

AU - Sheng, Xin Qing

PY - 2011/8

Y1 - 2011/8

N2 - Difficulties associated with extremely large targets, including construction of the geometry tree in the algorithm, reduction of the number of far-field patterns and fast computation of radar cross section, are pointed out. Based on the architecture of popular parallel platforms, distributed memory parallel systems, and properties of targets with extremely large sizes, relative solutions are proposed. The efficiency of the proposed parallel multilevel fast multipole algorithm (MLFMA) scheme is validated by successful computations of scattering from targets with over five hundred million unknowns and several thousand wavelengths in size.

AB - Difficulties associated with extremely large targets, including construction of the geometry tree in the algorithm, reduction of the number of far-field patterns and fast computation of radar cross section, are pointed out. Based on the architecture of popular parallel platforms, distributed memory parallel systems, and properties of targets with extremely large sizes, relative solutions are proposed. The efficiency of the proposed parallel multilevel fast multipole algorithm (MLFMA) scheme is validated by successful computations of scattering from targets with over five hundred million unknowns and several thousand wavelengths in size.

KW - High performance

KW - Multilevel fast multipole algorithm (MLFMA)

KW - Parallel

KW - Radar cross section (RCS)

KW - Scattering

UR - http://www.scopus.com/inward/record.url?scp=80052385412&partnerID=8YFLogxK

U2 - 10.3969/j.issn.1001-506X.2011.08.02

DO - 10.3969/j.issn.1001-506X.2011.08.02

M3 - Article

AN - SCOPUS:80052385412

SN - 1001-506X

VL - 33

SP - 1690

EP - 1693

JO - Xi Tong Gong Cheng Yu Dian Zi Ji Shu/Systems Engineering and Electronics

JF - Xi Tong Gong Cheng Yu Dian Zi Ji Shu/Systems Engineering and Electronics

IS - 8

ER -

High performance computing on scattering from extremely large targets by MLFMA

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this