• home
• current students
• graduate
• events

Presenter: Dr. Y. Rosa Zheng - NSERC Post-Doctoral Fellow, Dept. of Electrical and Computer Engineering, University of Missouri, Columbia, Missouri
Supervisor:

Date: Mon, April 11, 2005
Time: 14:30:00 - 15:30:00
Place: EOW 430

ABSTRACT

Abstract

Multiple-Input Multiple-Output (MIMO) systems have emerged as a new paradigm in wireless communications due to their enormous capacity gains predicted by recent theoretical researches. This seminar presents a practical MIMO transceiver that utilizes adaptive data transmission designed by stochastic waterfilling for the future high data rate and high user mobility wireless communication networks. These future systems often encounter realistic MIMO channels undergoing triply selective fading, that is, spatially correlated, time varying, and frequency selective fading. We propose a new triply selective fading channel model and investigate the three types of correlation matrices associated with it. We demonstrate that, besides spatial and temporal correlation matrices, there exists an ISI intertap correlation due to the delay spread in frequency selective fading channels. This ISI intertap correlation has been largely overlooked but it has significant impact on channel capacity and transceiver design. The three types of correlations are intertwined with each other and a new decomposition property is discovered to tackle the capacity analysis. To approach the capacity potentials in practical systems, the three types of correlation matrices, which are small-sized and slowly time-varying, are used at the transmitter to design adaptive transmission policies. We show that these stochastic waterfilling schemes based on the transmit and receive antenna correlations and/or the ISI intertap correlation can significantly improve the channel capacity and frame error rate, approaching the performances of deterministic waterfilling schemes that require perfect instantaneous channel state information at both transmitter and receiver.