Technical Program

Paper Detail

Paper:SAM-P1.11
Session:Array Processing
Time:Tuesday, May 18, 13:00 - 15:00
Presentation: Poster
Topic: Sensor Array and Multichannel Signal Processing: Signal detection and estimation
Title: ROBUST ITERATIVE FITTING OF MULTILINEAR MODELS BASED ON LINEAR PROGRAMMING
Authors: Sergiy Vorobyov; University of Duisburg-Essen 
 Yue Rong; University of Duisburg-Essen 
 Nicholas Sidiropoulos; Technical University of Crete 
 Alex Gershman; University of Duisburg-Essen 
Abstract: PARAllel FACtor (PARAFAC) analysis is an extension of low-rankmatrix decomposition to higher-way arrays. It decomposes a givenarray in a sum of multilinear terms. PARAFAC analysis generalizesand unifies common array processing models (like jointdiagonalization and ESPRIT); it has found numerous applicationsfrom blind multiuser detection and multi-dimensional harmonicretrieval, to clustering and nuclear magnetic resonance. Theprevailing fitting algorithm in all these applications is based onalternating least squares (ALS) optimization, which is matched toGaussian noise. In many cases, however, measurement errors are farfrom being Gaussian. In this paper, we develop an iterativealgorithm for least absolute error fitting of general multilinearmodels, based on efficient interior point methods for LinearProgramming (LP). We also benchmark its performance in Laplacian,Cauchy, and Gaussian noise environments, versus the respectiveCRBs and the commonly used ALS algorithm.
 
           Back


Home -||- Organizing Committee -||- Technical Committee -||- Technical Program -||- Plenaries
Paper Submission -||- Special Sessions -||- ITT -||- Paper Review -||- Exhibits -||- Tutorials
Information -||- Registration -||- Travel Insurance -||- Housing -||- Workshops

©2015 Conference Management Services, Inc. -||- email: webmaster@icassp2004.org -||- Last updated Wednesday, April 07, 2004