This is based primarly on the proposed antenna 
technologies, as it is the antennas that determine 
the frequency range of each concept.  Not surprisingly,
this places a premium on antenna technologies that are
capable of being pushed to operate in the 30-40 GHz 
region.  All current SKA concepts, except the European
array tiles, could probably be made to operate at 8 GHz 
(X band).  The basic problem with 8 GHz compared with 
32 GHz (Ka band) is the narrow bandwidth allocated 
for space-to-ground data transmission at 8 GHz.  The 
allocated downlink bandwith is ten times larger at 
Ka band than at X band.  Given that the primary science
driver for SKA support of spacecraft tracking is to 
greatly increase the downlink data rates obtainable 
from future missions, this is an important consideration. 

Source:  Dayton Jones, chair of WG9