LL Logo



Spatial Signal Processing
for Time-Dispersive
Digital Communication

M. Stojanovic and J.G. Proakis
Northeastern University
Department of Electrical and Computer Engineering
360 Huntington Avenue
Boston, MA 02115
tel: (781) 373-4429
fax: (781) 373-8970

This paper treats the problem of signal demodulation in a digital communication system that employs a large receiver array in the transmission of data through a time-dispersive channel, such as an underwater acoustic channel. In general, severely time-spread communication channels require long equalizers to combat intersymbol interference. With spatial diversity, the receiver performance may be limited by its complexity due to factors such as noise enhancement, slower tracking of time variations in the channel characteristics, and increased sensitivity to numerical errors. To alleviate these problems, a beamforming interpretation of the optimal diversity combining method is exploited. The resulting receiver consists of a many-to-few coherent spatial pre-combiner followed by a reduced complexity multichannel equalizer. Its performance is studied on experimental underwater acoustic data demonstrating capability to fully exploit spatial variability of multipath channels while keeping the complexity at minimum, thus allowing efficient use of large receiver arrays.



LL Logo Disclaimer

Direct comments and questions to: webmaster@ll.mit.edu

MIT Lincoln Laboratory. All rights reserved.