FC - Refer you to
this post...
One of the first debates over stealth concerned whether it was indeed possible to build a stealth air-to-air combat platform, because at that point all fighters had one thing in common, to wit, a bloody great RF searchlight in the nose.
The solution for the ATF/F-22 was more reliance on passive RF, low-probability of intercept radar, IRST, sensor fusion and offboard (AWACS), to get first-look, first-shot, first-kill with AMRAAM. Basically this is where Team F-35 still hangs its hat in the BVR regime.
This has been chipped away at, in subsequent years. MAWS, better EW and agility say "you may get the first look but you're going to have to get close for a high-Pk shot." A bit of RAM and much, much better passive and active EW conspire to raise the LPI bar a lot higher - trying to detect, much less track, a fighter without giving your own presence and location away gets much harder.
Dealing with leakers in WVR was another issue. Some said "Stealth rules, win in BVR and that's all" - look at the YF-23. The AF wanted belt and braces, so the F-22 ended up with a complex and heavy AIM-9 installation that gave wide-field-of-regard LOBL. The T-50 and J-20 have basically the same thing. It wouldn't fit on JSF.
However, the idea with JSF is to track everyone all the time in ACM with EO-DAS, and blast through the fight in a straight line while launching LOAL AIM-9X at the threat. But as mentioned, it can't carry AIM-9 in stealth mode - and the task of validating that EO-DAS works as advertised and will stitch its six sensors together while trying to assemble a 3D picture from 2D data sounds... interesting.