Cooked ALIVE? Seriously?
Gotta throw the BS flag on this one.
Let's do some simple analysis. Per international safety guidance (*) the controlled area (**) maximum permissible exposure level (PEL) for RF from 3GHz on up is 10mW/cm^2, averaged over six minutes. Note there is a factor of ten safety margin between the controlled area PEL and .any. observable effect. In this frequency range you're talking heating inside the eyeballs, and "effect" means a measurement signal just about noise in a dielectric simulation of a human. Your dangly bits will not tingle (***).
Typical SATCOM uplink frequency ranges are: 5.925 - 6.425GHz; 7.900 - 8.395GHz; 14.000 - 14.800GHz; 17.300 - 17.800GHz; 27.5-31GHz.
Typical SATCOM transmit powers? Hundreds of Watts. Extreme levels? 3.6kW at C band, 2.5kW at Ku band, hundreds of Watts in the mmW bands. See (****)
Antenna size? It matters. Commercial kit can go anywhere from man portable to a 9m monster. Let's assume a 4.8m compact cassegrain - typical of "large" SATCOM installations. This is not kit that you will just have lying around. Typical gain at Ku band is +54.6dBi. Note beamwidth at this gain is about 0.4deg (-3dB), so your attacker needs amazing aim.
What's the effective radiated power? 2.5kW * +54.6dBi = 721MW. Sounds scary. But let's see... power density at range in the main beam is Pt/(4*pi*r^2). At one meter(*****) the density is about 5,740,000 mW/cm^2. At 240m I'm down to 100mW/cm^2 and definitely safe. At 740m I'm down to the controlled area PEL, and legal as well as very safe.
What's the spot size at 240m? Length of arc S = r*theta; For 0.4deg the spot is ~5m across at 740m; at 240m downrange its 1.7m across. Enemy needs damn good aim.
Reality is no member of the general public will be within a inside 200m from the most powerful uplink sites - its called a 'controlled area' PEL for a reason. Most of the giant antennas are moved into position and then mechanically locked down; at these spot sizes you need rigidity. Smaller stuff that's actively pointed whilst going after orbiting vehicles moving around is ... smaller; ERP is lower and I care even less. For most large antennas I've seen I'm not even sure I can physically point them within 10deg or so of the ground; there's no point having the mount articulate that much.
Reality is the only real threat at these sites is to maintenance personnel exposed to open waveguides, feedhorns, leaking waveguide flanges. Slightly burned the meat INSIDE my hand one time on a leaking UHF waveguide flange ... Probably kW/cm^2 ... THAT hit about ten on my 'life sucks' meter. Eleven when the itching set in. But that's not an IT problem.
Life's hard enough without inventing problems.
(*) IEEE standard C95.11.
(**) Uncontrolled area PEL (general public) is a lot lower but I cannot be arsed to look it up at the moment. Gave you the reference. I've memorized only controlled area PEL and the real hazard levels. I don't have general public at my sites because ... its a controlled area.
(***) Except one dumbass sailor I had on a mast directly in front of a very large UHF air search radar - having failed to lockout/tagout same. Turns out he had piercing that had "boldly gone where no man had gone before". Holiday song time! "Jack's nuts roasting over an open fire...."
(****) https://www.cpii.com/product.cfm/4/13.
(*****) Yeah, I know this analysis is sketchy in the near field. Fraunhofer distance for 4.8m reflector at 15GHz is something like 2.3km. But this is a blog post, k?