Subtopic Deep Dive

← Electromagnetic Scattering and Analysis

Domain Decomposition Methods for EM Scattering
Research Guide

What is Domain Decomposition Methods for EM Scattering?

Domain decomposition methods (DDM) partition large electromagnetic scattering problems into smaller subdomains solved iteratively in parallel using non-overlapping or overlapping techniques like Schwarz and FETI solvers with optimized transmission conditions.

DDM enables efficient parallel computation for massive EM scattering systems from aircraft and ships. Key approaches include integral equation-based DDM (Peng et al., 2011, 220 citations) and adaptive integral methods (Bleszyński et al., 1996, 892 citations). Over 10 papers in the list address parallelism and preconditioning for these solvers.

Curated Papers

Key Challenges

Why It Matters

DDM supports real-time radar cross-section analysis for aircraft by distributing MoM matrices across thousands of processors (Bleszyński et al., 1996). It scales to billion-unknown problems in ship hull scattering via parallel oct-tree methods (Warren and Salmon, 1993). Structured preconditioners accelerate FETI convergence for dielectric interfaces (Bai, 2005). These enable design optimization in stealth technology and antenna arrays (Matekovits et al., 2007).

Key Research Challenges

Ill-conditioned system matrices

Large EM scattering leads to dense, ill-conditioned matrices from integral equations. Iterative solvers like GMRES diverge without preconditioning (Peng et al., 2011). Domain decomposition must optimize transmission conditions to restore conditioning.

Transmission condition optimization

Schwarz and FETI methods require Robin or optimized conditions for fast convergence across dielectric boundaries. Poor choices cause slow iteration counts in heterogeneous media. Balancing accuracy and parallelism remains critical (Bai, 2005).

Scalability to massive unknowns

Problems exceed 10^9 unknowns for full aircraft models demand extreme parallelism. Load balancing across subdomains fails on unstructured geometries (Warren and Salmon, 1993). Memory per processor limits fast multipole integration (Darve, 2000).

Essential Papers

AIM: Adaptive integral method for solving large‐scale electromagnetic scattering and radiation problems

E. Bleszyński, M. Bleszyński, T. Jaroszewicz · 1996 · Radio Science · 892 citations

We describe basic elements and implementation of the adaptive integral method (AIM): a fast iterative integral‐equation solver applicable to large‐scale electromagnetic scattering and radiation pro...

A parallel hashed Oct-Tree N-body algorithm

Michael S. Warren, John K. Salmon · 1993 · 457 citations

Article Free Access Share on A parallel hashed Oct-Tree N-body algorithm Authors: M. S. Warren Theoretical Astrophysics, Mail Stop B288, Los Alamos National Laboratory, Los Alamos, NM Theoretical A...

The Fast Multipole Method: Numerical Implementation

Eric Darve · 2000 · Journal of Computational Physics · 367 citations

Analysis of Large Complex Structures With the Synthetic-Functions Approach

Ladislau Matekovits, Valeriu Adrian Laza, G. Vecchi · 2007 · IEEE Transactions on Antennas and Propagation · 268 citations

An innovative procedure is presented that allows the method of moments (MoM) analysis of large and complex antenna and scattering problems at a reduced memory and CPU cost, bounded within the resou...

Structured preconditioners for nonsingular matrices of block two-by-two structures

Zhong‐Zhi Bai · 2005 · Mathematics of Computation · 249 citations

For the large sparse block two-by-two real nonsingular matrices, we establish a general framework of practical and efficient structured preconditioners through matrix transformation and matrix appr...

Convergence and stability of a discontinuous Galerkin time-domain method for the 3D heterogeneous Maxwell equations on unstructured meshes

Loula Fézoui, Stéphane Lanteri, Stéphanie Lohrengel et al. · 2005 · ESAIM Mathematical Modelling and Numerical Analysis · 249 citations

A Discontinuous Galerkin method is used for to the numerical solution of the time-domain Maxwell equations on unstructured meshes. The method relies on the choice of local basis functions, a center...

Closed-form Green's functions for general sources and stratified media

G. Dural, M.I. Aksun · 1995 · IEEE Transactions on Microwave Theory and Techniques · 239 citations

The closed-form Green's functions of the vector and scalar potentials in the spatial domain are presented for the sources of horizontal electric, magnetic, and vertical electric, magnetic dipoles e...

Reading Guide

Foundational Papers

Start with Bleszyński et al. (1996) for AIM baseline scaling, then Peng et al. (2011) for core IE-DDM, followed by Bai (2005) for block preconditioners essential to FETI convergence.

Recent Advances

Śmigaj et al. (2015) for BEM++ implementations; Matekovits et al. (2007) for synthetic functions in complex structures.

Core Methods

Non-overlapping FETI-DP with primal constraints; overlapping Schwarz with Robin transmission; fast multipole for subdomain fills (Darve, 2000); hashed oct-trees for parallelism (Warren and Salmon, 1993).

How PapersFlow Helps You Research Domain Decomposition Methods for EM Scattering

Discover & Search

Research Agent uses searchPapers('domain decomposition EM scattering FETI') to find Peng et al. (2011), then citationGraph reveals 220 citing works on transmission conditions, and findSimilarPapers uncovers parallel DDM extensions from Bleszyński et al. (1996). exaSearch queries 'integral equation domain decomposition preconditioners' for 50+ related scalings.

Analyze & Verify

Analysis Agent applies readPaperContent on Peng et al. (2011) to extract convergence rates, verifyResponse with CoVe cross-checks condition numbers against Bai (2005), and runPythonAnalysis reproduces GMRES iteration plots using NumPy eigenvalue decomposition. GRADE assigns A evidence to scalability claims via citation analysis.

Synthesize & Write

Synthesis Agent detects gaps in dielectric transmission conditions across Peng (2011) and Darve (2000), flags contradictions in parallelism overheads, then Writing Agent uses latexEditText for equations, latexSyncCitations with 10 papers, and latexCompile for IEEE-formatted review. exportMermaid diagrams subdomain partitioning and FETI coupling.

Use Cases

"Extract convergence data from DDM papers and plot iteration counts vs subdomain size"

Research Agent → searchPapers → Analysis Agent → readPaperContent(Peng 2011) → runPythonAnalysis(NumPy pandas matplotlib plot GMRES curves from 5 papers) → researcher gets convergence comparison CSV with statistical fits.

"Write LaTeX section on FETI-DP for EM scattering with citations"

Synthesis Agent → gap detection → Writing Agent → latexEditText(FETI equations) → latexSyncCitations(Bai 2005, Peng 2011) → latexCompile → researcher gets compiled PDF section with numbered equations and bibliography.

"Find GitHub codes for parallel domain decomposition in EM solvers"

Research Agent → searchPapers(DDM EM) → paperExtractUrls → paperFindGithubRepo → githubRepoInspect → researcher gets 3 active repos with BEM++ implementations (Śmigaj et al., 2015) including convergence benchmarks.

Automated Workflows

Deep Research workflow scans 50+ DDM papers via searchPapers → citationGraph → structured report with convergence tables from Peng (2011). DeepScan applies 7-step CoVe to verify scalability claims in Warren (1993) with Python eigenvalue checks. Theorizer generates novel transmission conditions by synthesizing gaps between Bai (2005) preconditioners and Darve (2000) FMM.

Try Doxa for Domain Decomposition Methods for EM Scattering Research

Frequently Asked Questions

What defines domain decomposition methods for EM scattering?

DDM splits scattering surfaces into subdomains solved iteratively via Schwarz or FETI with transmission conditions (Peng et al., 2011). Non-overlapping uses FETI-DP; overlapping employs optimized Robin terms.

What are main methods in DDM for EM?

Integral equation DDM (Peng et al., 2011), adaptive integral methods (Bleszyński et al., 1996), and block preconditioners (Bai, 2005). Parallel oct-tree accelerates matrix-vector products (Warren and Salmon, 1993).

What are key papers on DDM EM scattering?

Peng et al. (2011, 220 citations) introduces IE-DDM; Bleszyński et al. (1996, 892 citations) develops AIM for large-scale; Bai (2005, 249 citations) provides structured preconditioners.

What open problems exist in DDM for EM?

Optimal transmission for multi-dielectric media; hybrid DDM-FMM scaling beyond 10^9 unknowns; GPU acceleration of FETI iterations without preconditioner redesign.