Channel parameter estimation in mobile radio environments using theSAGE algorithm

摘要

This study investigates the application potential of the SAGEn(space-alternating generalized expectation-maximization) algorithm tonjointly estimate the relative delay, incidence azimuth, Dopplernfrequency, and complex amplitude of impinging waves in mobile radionenvironments. The performance, i.e., high-resolution ability, accuracy,nand convergence rate of the scheme, is assessed in synthetic and realnmacro- and pico-cellular channels. The results indicate that the schemenovercomes the resolution limitation inherent to classical techniquesnlike the Fourier or beam-forming methods. In particular, it is shownnthat waves which exhibit an arbitrarily small difference in azimuth cannbe easily separated as long as their delays or Doppler frequenciesndiffer by a fraction of the intrinsic resolution of the measurementnequipment. Two waves are claimed to be separated when the mean-squarednestimation errors (MSEEs) of the estimates of their parameters are closento the corresponding Cramer-Rao lower bounds (CRLBs) derived in anscenario where only a single wave is impinging. The adverb easily meansnthat the MSEEs rapidly approach the CLRBs, i.e., within less than 20niteration cycles. Convergence of the log-likelihood sequence is achievednafter approximately ten iteration cycles when the scheme is applied innreal channels. In this use, the estimated dominant waves can be relatednto a scatterer/reflector in the propagation environment. Theninvestigations demonstrate that the SAGE algorithm is a powerfulnhigh-resolution tool that can be successfully applied for parameternextraction from extensive channel measurement data, especially for thenpurpose of channel modeling