An error analysis of feedback correlation beamforming for the IS-95 reverse link

Abstract

We investigate the application of an existing beamforming weight estimation method to the IS-95 reverse link and propose an extension to the algorithm which yields dramatically improved accuracy. An error analysis presents the statistical parameters of the various estimators. This includes an eigenvector perturbation analysis and the derivation of the Cramer-Rao lower bound for this situation. Sample results are used to demonstrate the close agreement between observed and predicted values, thereby justifying the use of the analytical derivations to reduce simulation complexity in cell capacity evaluation studies.