LL Logo

THIRD ANNUAL
ASAP '95 WORKSHOP

____

On the Use of Non-Causal
Filters for Adaptive Sensor
Array Processing

Pauline C. E. Bennett
Cornell University
Ithaca, NY 14853
email: pauline@ee.cornell.edu

Abstract
Many recent papers tout the advantages of multisensor adaptive antenna arrays for use in personal communication systems because of their demonstrated ability to suppress cochannel interference (CCI) and intersymbol interference (ISI). This interest is spurred by the expectation that use of multi-sensor adaptive antenna arrays will result in an increase in the amount of channel capacity available in such systems. This increase in the channel capacity is obtained as the outputs of the multisensor arrays are optimally combined resulting in suppression of the undesirable sources of interference.

Unfortunately, the computational cost of multisensor systems which use optimal combination seems to be prohibitively high for use in low-power portable systems. Currently most work assumes that the narrowband assumption holds for the signals which impinge upon the array. While illuminating, this work avoids the fact that the time-varying channels available for mobile communications are inherently dispersive in nature. Because of this, the narrow-bandwidth assumption may not always hold.

In this presentation we restrict ourselves to the case where the signal is corrupted only by co-channel interference. Analytical data is presented which suggests that an extension of Van Trees' work1 on non-causal filter-and-delay structures to a multi-sensor environment offers significant performance improvements over the use of purely causal filters of identical length. In deriving these results we use a mathematically elegant method of showing how the autocorrelation matrix of the received data depends explicitly upon the filter structures, thus simplifying the analysis. The performance improvements were shown to be valid for signals with fractional bandwidths as low as 0.5%.

1 H. L. Van Trees. Detection, Estimation, and Modulation Theory. Part 1, New York: Wiley, 1968.

____

 


LL Logo Disclaimer

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

MIT Lincoln Laboratory. All rights reserved.