Lots of missions and mission phases don't require bigger antennas. Look at http://eyes.nasa.gov/dsn/dsn.html for current usage of the Deep Space Network. As I type this, 3 smaller dishes in Goldstone are handling 4 Mars missions, all solar powered, Odyssey at 14.22 kb/s. At Canberra, the only active dish is a small one, receiving data at 159 b/s from Voyager 2 (!). Which, I grant, doesn't have a lot to say right now.
3 sites, 1 big dish and 3 smaller ones at each, except for a 4th smaller one at Goldstone, that's a serious investment and I gather enough for now, especially since satellite technology now allows storing massive amounts of data (8 GB for New Horizons) for download as the mission and ground stations allow. Note also that after New Horizons finishes downloading the data from its flyby in November 2016, it won't have a lot to say unless and until its Kuiper belt object flyby.
Fair enough. I had not realized that NH could send data at a faster rate when bandwidth was available. I wondered about the point of having 8 gb storage when you had such low speeds - it would take years to download.
Eventually, a lot of spacecraft communications will switch to optical systems instead of using radio. The major gating factor there, at least in regards to solar system exploration, is the creation of off-Earth infrastructure to facilitate it. Which likely would be just a small fleet of relatively modest sized space telescopes designed for 2 way communications. Having the infrastructure off Earth would mean that it would be insensitive to weather and would also be easier to maintain contact with spacecraft nearly anywhere in the sky at any given time. Building that infrastructure probably won't happen until launch costs fall but realistically it's something that could easily happen in the next decade or so.
