PhoenixEnigma wrote:Something like a CubeSat is probably the way to go, if only because they're standardized to a good degree and a lot of the work has been done for you. Launch costs are pretty variable, but look to start at $15k per U (10cm cube), and run up closer to $50k/U. Looking over their pricelist of components, I'd expect a basic satellite to run $20k+, though you could easily spend far more. All told, $100 000 would probably be enough to get something in orbit.
This is your best option (launch systems are horribly complex. Only one amateur group has got beyond the Karman line and that was a very light weight sounding rocket that did not get into orbit.
I don't think you need government permission to launch one; I think you are covered by the people who arrange the launch. What you will need to do is to get your cubesat in someone's launch vehicle, and convince THEM that it won't cause any problems; specifically problems for their primary payload. It is important to understand that, from the point of view of the people launching the rocket, your cubesat payload is basically ballast.
Main problem is the complexity of a cubesat. One individual almost certainly can't pull it off without screwing something up. Most cubesats are built by teams of engineering postgraduates, often with hand-holding from industry or national space agencies. If you want to simply be able to point up and say "I've made something that flies in space" perhaps your best bet would be to find an existing team near to you, and offer whatever skills you have to them.
As for radiation; don't sweat it. COTS electronics can survive a while in LEO, even if you fly through the south Atlantic anomaly. You just need to make your system reasonably robust; its got to be able to cope with being shut off by a brown out or a single event upset and be able to reboot itself nicely. Your obit will probably decay over the same timescale as your electronics being permanently damaged.