SIXTH ANNUAL
ASAP '98 WORKSHOP

Mutual Coherence
Parameter Estimation

 

 

Tim Barton
MIT Lincoln Laboratory
244 Wood Street, Room C-370
Lexington, MA 02173-9108
tel: (781) 981-3278
email: barton@ll.mit.edu

Abstract Combining data from multiple radar sensors with different operational bandwidths through ultra-wideband (UWB) signal processing has the potential to provide increased information about a target under track over the information provided by the radar sensors individually. All-pole models have been used to model the frequency domain radar return for multiple-scatterer targets for these radars with different operational bandwidths. This has lead to all-pole models, which not only model the frequencies for which observations have been made, but also provides a method for interpolation between, and extrapolation beyond, the observation frequency bands. This model then provides an UWB frequency response of the target scatterers which can be useful for many applications such as scatter characterization, UWB range profiling, UWB imaging, and other discrimination and intelligence applications.

Combining data in different frequency bands from different radar sensors for use in UWB signal processing assumes that the data is mutually coherent with respect to phase and time delay (or range offset) for each target scatterer in each radar. However, data from multiple sensors is not, in general, mutually coherent due to such characteristics as hardware phase and time delay "synchronization" differences between the sensors, sensor deployment differences, target motion, observation time differences, and range offsets due to transmit waveform range-Doppler ambiguities. This presentation will show that mutually coherent data is essential for successful UWB signal processing through system analysis, simulation, and real data analysis. Specifically, the required mutual coherence is examined under a variety of different sensor-target scenarios. Given these considerations, data models are developed which are parameterized by the mutual coherence parameters as well as the other UWB signal processing parameters. Based on these data models, maximum likelihood methods for estimating the mutual coherence parameters are then derived. The process of maximum likelihood mutual coherence parameter estimation is then examined through simulation and data analysis. It will be shown that these techniques can potentially provide a robust method for mutually cohering data from multiple sensors and lead to successful UWB signal processing.


 


Disclaimer

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

MIT Lincoln Laboratory. All rights reserved.