As the C and Ku bands begin to become congested,
significant interest has been generated in Ka band (30/20
GHz) for commercial satellite communications applications. Early research was conducted
from the 1970s with satellites from Japan, US, Italy and
the European Space Agency (ESA), followed most recently by NASA’s
Advanced Communications Technology Satellite (ACTS) in 1993. While development of
Ka-band satellites have been slow with only several in orbit
in 2003 (Optus C1 was launched in June 2003), several
hundred applications have been filed with ITU for Ka-band satellites.
Iridium uses Ka-band frequencies for inter-satellite links as well as
for links to Earth station gateways. The principal interest in
the band is for the delivery of interactive multimedia services.
Ka-band
systems have a number of advantages. Perhaps the most significant
is the increased bandwidth, in a band that does not
have to be shared with terrestrial systems; Ka band
offers 2-3 GHz of bandwidth, approximately twice that available in
Ku band and some five times more than C band.
Ka-band components are much smaller due to the smaller wavelength
allowing more devices and smaller antennas on the same-sized platform.
Footprints for Ka-band spot beams can also be made smaller,
often from the same antenna, facilitating frequency re-use.
The difficulty in
Ka band is that rain attenuation becomes so severe
that link margins of the order of 14 dB are
required. These large losses can be accommodated during periods of
heavy rain by increased satellite transmit power (possible because of
the exclusive access to the band); Ka-band transmit powers may
be 3-4 dB higher than X-band or C-band equivalents. Antennas
can also be made larger; since gain of an antenna
increases with the square of the frequency, the same size
antenna as in C or Ku band will naturally have
a higher gain (although this will be necessary to offset
the free-space loss over the path, which will be higher
by the same degree). Ka-band antennas must also be formed
to much higher precision than for lower frequency bands to
reduce loss. In any case, Ka-band systems are currently more
expensive, although the full benefits may be realised over time
as the price of components drops. Ka-band LNAs can achieve
noise figures of 3-4 dB and TWTs are available to
deliver 45-90 W.
Reliable prediction of propagation impairments becomes critical in
Ka band, as do mechanisms for overcoming them. Since
increased antenna size is already necessary to compensate for additional
free-space loss, beam shaping can be used to increase gain
in those directions that are experiencing fading (although this is
at the expense of gain in other directions). Up-link power
control can be used at the Earth station which, based
on measurement of signal strength, can increase and decrease the
uplink power to accommodate fading. Additionally, site diversity can be
used to overcome localised fading: provided sites are separated by
more than about 10 km, the fading experienced in each
location will be uncorrelated. Site diversity is expensive, however, and
is really only an option in large cooperative networks where
some traffic sharing arrangement can be made.
Other topics in our resources on Satellite Communications related to Ka Band include: