Lyle Johnson wrote:
I discussed some of these ideas with Eric, N7CL, at the TAPR meeting. Eric is very good at antenna system design, did S band stuff for us at MMS and does this sort of thing now for DoD.
Anyway, he suggested we consider Vivaldi's rather than patches. We used Vivaldi's on our trucks at S Band and they work well, are trivial to fabricate (a piece of PC board - I know, radiation, but it is the concept I am wanting us to consider). They are well behaved and have very little coupling to adjacent antennas. I'm not an antenna guy, but you might want to consider the Vivaldi and/or communicate with Eric.
Eric also assisted me with the design of the Microsat 2M rcvrs -- which bore a strange similarity to an MMS design.
AFAIK, the Vivaldi and antipodal antennas are linearly polarized. Multiple elements require crossed antennas, like Fig.9 in http://www.ansoft.com/news/articles/04.09a_MWJ.pdf#search=%22vivaldi%20anten... (note that the radiation is "endfire").
A PhD thesis for a 2-8 GHz describing this antenna used for a Snow radar is seen at http://profusion.ittc.ku.edu/research/thesis/documents/ravi_prakash_rajarama.... Also, to see what an array of these devices might look like, see http://nemes.colorado.edu/microwave/papers/2000/APS_JPWbn_00.pdf#search=%22v... (in this case, the SKA is a concept design for an SKA=square-kilometer array for radio astronomy. Many different approaches are being studied ranging from a field full of tens of thousands of patch antennas, each equipped with signal processing "smarts" to arrays of a hundreds of dishes. See http://www.skatelescope.org/ for more info)
73, Tom