Optimal sensing matrix for sparse linear models

S. PAZOS; M. HURTADO; C. MURAVCHIK; A. NEHORAI

San Juan, Puerto Rico

Workshop; IEEE Int. Workshop on Computational Advances in Multi-Sensor Adaptive Processing; 2011

In this paper, we propose a method for designing theoptimal sensing of measurements which can be characterized bya sparse linear model. The aim of the sensing operation is notonly to reduce the amount of data to be processed but also toreject undesired signals (interferences). As a result, we reducethe computation time and the error for estimating the unknownparameters of the model, with respect to the uncompressed data.Using synthetic data, we analyze the performance of the proposedalgorithm. Additionally, we use real radar data to show anapplication of the method.