Conference Paper
Compact reduced order modeling for multipleport interconnects
Dept. of Electr. Eng., California Univ., Riverside, CA, USA;
DOI: 10.1109/ISQED.2006.35 Conference: Quality Electronic Design, 2006. ISQED '06. 7th International Symposium on Source: DBLP

Article: Efficient ReducedOrder Macromodels of Massively Coupled Interconnect Structures via Clustering
[Show abstract] [Hide abstract]
ABSTRACT: In this paper, a novel algorithm for creating efficient reducedorder macromodels from massively coupled interconnect structures is described. The new algorithm addresses the difficulty associated with the reduction of networks with a large number of input/output terminals, that often results in large and dense reducedorder models. Application of the proposed reduction algorithm leads to reducedorder models that are sparse and blockdiagonal in nature. An additional advantage of the proposed algorithm is that it does not assume any correlation between the responses at ports and thereby overcomes the accuracy degradation that is normally associated with the existing singular value decomposition based terminal reduction techniques. Also, the presented algorithm is highly suited for multithreading implementation and thus facilitates parallel transient simulation. Validity and efficiency of the proposed algorithm are demonstrated through computational results.IEEE Transactions on Components, Packaging, and Manufacturing Technology 01/2013; 3(5):826840. · 1.26 Impact Factor 
Conference Paper: WaveletBased Passivity Preserving Model Order Reduction for Wideband Interconnect Characterization.
[Show abstract] [Hide abstract]
ABSTRACT: Model order reduction plays a key role in determining VLSI system performance and the optimization of intercon nects. In this paper, we develop an accurate and provably passive method for model order reduction using adaptive waveletbased frequency selective projection. The wavelet based approach provides an automated means to generate low order models that are accurate in a particular range of frequencies. Theresults indicatethatourapproachprovides more accurate reduced order models than the spectral zero method with uniform interpolation points and the zeroshift and multishift Block Arnoldibased techniques. In this paper, we develop an adaptive, passivity preserv ing methodology for the selection of interpolation points in spectral zerobased model order reduction. We dynamically select expansion points by applying Haar wavelets to de tect complex changes in the frequency points spanned by the spectral zeros of the system and select the dominant interpolation points. The adaptive scheme provides a low order realization with optimized matching of the system re sponse for a given range of frequencies. The preservation of passivity is guaranteed by selecting interpolation points as a subset of the system's spectral zeros. The approximate loworder model can then be directly constructed from the projection matrices derived from these interpolants. In or der to demonstrate the efficiency of the approach, we ap ply our technique to an RLC network representing an in terconnect wire. The results indicate that the waveletbased method provides higher accuracy approximate models than techniques based on moment matching and uniform inter polation point selection.8th International Symposium on Quality of Electronic Design (ISQED 2007), 2628 March 2007, San Jose, CA, USA; 01/2007  [Show abstract] [Hide abstract]
ABSTRACT: An efficient clustering scheme for macromodeling of massively coupled interconnect structures is presented. The proposed method addresses the issue of dense reduced models through a novel reduction algorithm that leads to sparse blockdiagonal system matrices. It also overcomes the accuracy degradation, normally associated with the existing portreduction techniques.01/2010;
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.