I have not done it on the Axon2, but have done if for Basic Atom Pros.
There are a few different approaches:
1) To start off you might try the function: pin_pulseIn
You would do this for each IO pin connected to your receiver. This is the simplest way, but could be real slow, especially if you order your pulsein's in the wrong order. That is suppose that you order it in the logical order: 1, 2, 3, 4, 5, 6 and suppose the pulses are also generated in that order. This will probably cause real slow code. As after your pulsein for 1 completes and it enters into the code for pin 2, the transition may(probably) already happened and you have to wait for full RC cycle time until it can process it (20ms typically). So you would be much better off, reading the pulses in some other order like: (1, 3, 5, 2, 4, 6), this way you can get all 8 in the time of 2 to 3 RC cycles.
2) You could write a function that looks at all 6 IO pins at once and it looks for any of them to transition to high, grab time, wait for it to transition low and calc the delta time and again check to see if any of the other IO lines changed... Do this until you get a reading for all channels. I did this in H8 assembly language for BAPs, worked fine for Hitec and I believe Spectrum, but ran into issues with Futaba, as some of the channels overlap significantly. I wrote up an outline of code to handle this in basic, but never completed it as I have a Hitec...
3) You could try using the Pin Change interrupt and use it to capture the pulse widths in the background.
4) You may be able to Hack the RC receiver and solder a wire onto the clock pin of one of the chips and get the signals for all X channels on one pin. There are several places up on the web that talk about this. I did this on a Hitec 6 receiver... More details about this on the thread: http://www.lynxmotion.net/viewtopic.php?f=21&t=6629.
Hope that helps