Technology Innovation - Beam Multiplexing
This research thrust focuses on the exploration of phased array technology merging with novel
signal processing techniques to develop optimal scanning strategies.
An agile beam phased array
radar has the potential not only to increase the scanning rate, but also to measure meteorological
variables not currently available and to enhance data quality.
For a mechanically rotating radar a long dwell time is required to obtain a sufficient number
of independent samples for accurate spectral moment estimation. For example, a dwell time of
approximate 50 ms is needed to achieve the precision of 1 dB in reflectivity and 1 m/s in velocity
estimates for WSR-88D
(Smith, 1995;
Bringi and Chandrasekar, 2001).
This fundamental limitation
of a mechanical scanning radar is due to the fact that signals from continuous beam directions are
highly correlated (hence no new information content). In other words, fast scanning will be achieved
at the expense of degradation in data quality because fewer independent samples are obtained and
used in the spectral moment estimation.
On the other hand, a phased array radar can steer its beam
to other directions and return to its original direction after signals become uncorrelated, therefore
no wasteful time is spent. Thus, the use of radar resources can be optimized to produce a maximum
update rate for a specified precision of reflectivity and velocity
(Smith et al., 1974).
This scanning
scheme is defined as
Beam Multiplexing (BM). A minimum number of two pulses is transmitted
in each beam direction in order to estimate all three spectral moments using the autocovariance
method (the pulse pair processing)
(Doviak and Zrnic, 1993).
Various scanning strategies based on
BM will be developed and will be used adaptively for varying weather phenomena.
Torres and Zrnic (2003) have shown that
higher scanning rate for NEXRAD is feasible by applying whitening
technique on range oversampled data. The whitening technique makes use of oversampled signals in
range to increase the number of independent samples, while BM operates in sample time to achieve
the same goal. The two techniques are complementary and can be implemented simultaneously. It
is hypothesized that the scanning rate of BM can be doubled if oversampling in range with a factor
of 4.