228 
In Europe and Canada the tendency seems to consist of replacing 
the usual array processors (AP 120 B) by recent and efficient 
array and/or signal processors. Two typical examples of such 
machines are the ST-100 by STAR Technologies and the T-ASP by 
ESE. 
In the U.S.A, the most recent approach is not known at the‘time 
of: writing this report (March 1984), but will be the subject of 
a JPL presentation during the working group session in Rio de 
Janeiro. 
3.3 Algorithms for Preprocessing 
  
The evolution in the field of algorithms has been threefold: 
- ‘first, the preprocessing algorithms (ice. for image 
production) have been refined either to offer more 
operational flexibility (e.g. autofocus, automatic chirp 
extraction etc.) and/or more efficiency for a dedicated 
mission (e.g. SPECAN for C-band processing) 
- second, the relationship between algorithms and product 
quality has been considered (e.g. refined comparison of 
algorithms including quality considerations) in particular 
in view of new SAR products (e.g. wave spectra) 
- third, the most recent efforts consist of rewriting proven 
algorithms to optimally map them on new array processors. 
The trade-off flexibility (e.g. algorithm for L-, X- and 
C-band) and computer efficiency becomes of particular 
importance. The mapping of the generalised SAR processor 
(GSAR by MDA) on a ST-100 machine (at DFVLR) is a typical 
example of such an evolution. 
3.4 Acquisition/Validation/Calibration/Simulation 
Side issue in 1979, this topic emerges as of peculiar 
importance today, to allow for a proper use of the SAR data 
(see next paragraphs). 
The acquisition of the Convair 580 SAR data offered the first 
opportunity (Frascati meeting) to raise the validation and 
calibration issues, not only for usual scenes (e.g. using 
corner reflectors) but also in case of dynamic scenes (i.e. sea 
surface). The tutorial paper presented in Ottawa did confirm 
that all these issues are just in their infancy. They are 
today limited to using corner reflectors for calibration and to 
developing tools (e.g. at ESA, JPL and in Japan) for assessing 
the quality of actual SAR image products in terms of resolution 
(spatial and radiometric) and geometry (e.g. spatial 
distortions). 
More efficient tools are known to be under consideration for 
calibration (e.g. transponder), and validation, in view of the 
next spaceborne SAR's.