Hmmm...Well, I just wanted to notify the site, that this might set off a chain of events - that could get interesting...
In the years I've worked with MOSFET - unless there are "EKL" and their counterparts used, the issue of self-biased MOSFET is a rare realm of effort.
Bipolar can do it thru their "PN junction presence" at the Base to connect itself to the other two in the substrate, Collector and Emitter. RF gets rectified there - but remember too, the CLASS the thing works in, Class C and Class D for it's biasing and power - the reason why it works the way it does is by the Collector - using the BASE to flow both to the Emitter.
It's when you ADD in a method to turn the transistor on using a smaller level of signal - using BIAS - does the post you made - worth letting others know of the approach - else we just might not be seeing a step they did - and we missed it because they did this further up a chain or feeder in the TX line.
A diode of some sort, or a biasing method - was used in/and still is used, just getting the MOSFET to turn on in the presence of RF. Now, I'm - not saying it's impossible, just a lot of energy is needed in the RF signal to make the GATE turn on - rendering much of the signal that can be used, lost to the effort of just getting the thing turned on - there is not much left to reproduce.
So double check to see if they "regulated" the Bias using a feeder from - say, the 8V+ rail or TX line - so they could bias the Gate with just enough to turn it on when it was in TX mode.
