Efficient FRW Capacitance Extraction in Stratified Media via Multilayer Poisson-Kernel Expansions and Neumann-Trace Flux Evaluation
We develop a floating random walk (FRW) method for repeated parasitic capacitance extraction in stratified dielectric media. The method combines modal expansions of multilayer Poisson kernels in rectangular transition domains with a Neumann-trace formulation for finite Gaussian-surface patches. This formulation rewrites the flux integral as a weighted boundary integral of the potential over the enclosing box. The weights are computed by integrating the Neumann trace of the multilayer Poisson kernel over each patch, while boundary potentials at quadrature points are estimated by FRW paths. The construction of multilayer Poisson kernels requires two sets of vertical modal components. The transmission modes are obtained deterministically by enforcing interface continuity conditions. The stratified Sturm–Liouville eigenmodes are computed by pseudo-arclength homotopy continuation. This provides robust and accurate high-order modes for truncated kernel expansions in complex multilayer stacks. The computational framework separates medium-dependent kernel construction from layout-dependent charge evaluation. For a fixed stratified medium, reusable transition samples are generated offline. In the online stage, the boundary-integral weights are computed for the quadrature points on the local enclosing boxes, and the corresponding boundary potentials are estimated by FRW paths. Numerical experiments demonstrate the accuracy and efficiency of the proposed method for repeated extraction problems.
Authors
- Zhenya Zhou (ORCID: https://orcid.org/0000-0002-2645-6386)
- Chunxiong Zheng
- Jiaqi Li
- Tao Cui
Institutions
- University of Science and Technology of China (CN)
- Chinese Academy of Sciences (CN)
- National Center for Mathematics and Interdisciplinary Sciences (CN)
- Academy of Mathematics and Systems Science (CN)
- University of Chinese Academy of Sciences (CN)
- Tsinghua University (CN)
Publication Details
- Journal
- Communications in Computational Physics
- Published
- 2026-11-20
- DOI
- https://doi.org/10.4208/cicp.oa-2026-0160
- Primary Topic
- Electromagnetic Scattering and Analysis
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Chinese Academy of Sciences