a) Is this physically possible?
b) Is this technologically feasible?
a) Is this physically possible?
Sounds like you need some form of microwave isolator (assuming you’re talking about RF frequencies). However, I’m not sure this is possible in the sense you want, which seems to be a non-reciprocal faraday cage.
I think it’s possible for a limited set of frequencies, but not the whole electro-magnetic spectrum. Or, it’s possible for a limited direction, but not the whole solid angle.
What you’re asking for essentially is a perfect EM-radiation absorber. The limiting factor is that all the energy and entropy received has to go somewhere. I suppose it’s possible to have a perfect absorber over all frequencies and all angles, and then emit the received energy as something other than photons. Imagine a perfect photon-to-matter/anti-matter converter, for example. Even that might have trouble–I’m not sure how the entropy content of matter compares to that of photons.
A black hole comes close, for a certain value of “enclosure”, but the Hawking radiation is still emitted primarily as photons.
The closest you’ll get with current technology is a device that absorbs almost all photons within a specific band and in a specific direction. For example, a camera or a large-aperture radio-frequency antenna. You could make an enclosure by arranging a spherical array of such devices.
Not a passive enclosure, at least, as that would violate the Second Law of Thermodynamics.
Won’t that depend on the radiation being absorbed? My guess would be if black-body radiation is received, no useful work could extracted from it, and so an external power source would be required to convert the radiation into something else. However, if radiation with low entropy like laser light or a Tesla-style RF beam is received, enough useful work could be extracted from the radiation itself to power the conversion process.
I expect that the EM spectrum in most locations near Earth is rather closer to black-body radiation than laser light, so the hypothetical converter would have to be extremely thermodynamically efficient in order to be ambiently powered. But that’s an engineering problem, rather than a physics problem.
If it only needs to work over a certain limited range of frequencies, then yes, it is possible.
Make yourself a handy dandy Faraday cage. Place an antenna outside of the cage, and have the signal from the amplifier fed into the Faraday cage (which necessitates a hole in the Faraday cage and some special handling there). Inside the Faraday cage, have the signal go into a circulator. A circulator is a passive 3 port device. Signals that go in port A come out port B. Signals that go in port B come out port C. Signals that go in port C come out port A. So you attach the antenna to port A, your receiver to port B, and a dummy load to port C. If you try to transmit, all of your energy goes into the dummy load on port C. Anything coming into the antenna goes into your receiver though. You can receive, but not transmit.
I just noticed that the OP specified that EM signals could not be sent. One could argue that black-body radiation is not a signal. If so, that makes the enclosure much simpler, since we no longer have to worry about dumping the received energy and entropy. In fact, the definition of an ideal black body is such a device–it absorbs all frequencies EM radiations and emits only black-body radiation. Any star is a good approximation of a black body.