generally when you prep files for radio you assume there's an element of heavy compression (either wideband of multiband), phase rotation, stereo enhancement, EQing, high frequency limiting and/or some form of peak limiting. That's the usual nature of the processors used to transmit radio signals.
Any given radio station may receive such a wide variety of material that you kind of end up with the need to hypercompress them all to achieve a constant level for broadcast. The processors are set to work over a variety of different media, so you have to keep that in mind.
The station does what it can for the mix, but they aren't mastering engineers. You sort of have to set up the mix in a way that it delivers the goods first.
Qualities of a radio ready mix may be things like having your centered/most important (like vocals and leads) information boosted higher than you would for a home playback system and/or narrowing the width of your mix.
The important thing is that the message translates clearly on radio. It's not so much about a perfect sonic blend. It's about getting the message across.
Editing will be another factor. Make sure you ask how long the track can be and that the content is morally acceptable for that particular station.
Also, remember that since the track will be so compressed, it'll bring out your background stuff (reverb, FX, you name it). So you may mix a little dry for that situation.
In terms of mastering, the general rule of thumb is to avoid the extra stage of compression as well as any additional sweeteners that you may normally add for a mobile media kind of production.
Just listen for a track you like on the radio and compare that to the original album itself.
