No, I am not talking here about diversity antennas, where multiple-antenna setup is used to overcome local problems of fading, multipath, and related ills, nor politically correct "diversity".
I am talking about the amazing variety of antenna configurations, designs, shapes, topologies, and physical implementations. To invoke a cliché or two, antennas are both awesome and thought-inspiring: so much true cleverness and creativity, and all so tangible and visible, "Antennas give me a headache". And the antenna itself is a passive component, which adds to the challenge– but is also liberating, in a way.
I was reminded of this when I saw the March 2010 "antenna" issue of QST, the official publication of the ARRL, an organization devoted to amateur (ham) radio (and yes, ham radio is still alive and well). Since the contents are not available online except to subscribers, you'll have to go over to your better local library (remember those? They have new and old material that is not available online, yet still is vital and valuable) or your better-stocked bookstore (such as Barnes & Noble).
Although there are just a few antenna designs in the theoretical sense (dipole, long-wire, to name two) the reality is that the number of both simple and complex actual realizations of these is staggering. Sometimes this is due to performance priorities, or wavelength/frequency issues, or the need for integral antennas such as in handheld devices (I still can't believe that the ones built in cell phones can possibly work at all, but they do). There are also PCB-based antennas, as well as huge dishes doing deep-pace tracking. And of course, the antenna itself is only part of the story, since your receiver preamp or transmitter driver must be matched to the antenna to achieve efficient power transfer.
The irony is that functionally the antenna is a very simple component with a straightforward mission: to capture or radiate RF; and to act as a transducer between electromagnetic energy and wire-confined energy. While today's antenna design is disciplined by complex and sophisticated simulation, unlike the "hey, this might work" approach of the early days, there are still interesting physical designs that must first be "thought up" by designers before the software can analyze them.
There's lots of room for innovation in antennas, and it is one area where "lone-wolf" individuals can still make a difference.♦