Electromagnetic shielding with metal box and wire inside I want to know if this model will have a good electromagnetic shielding or I mean it will have low radio waves losses. There is a box of aluminium foil let say 0.5x1x0.1m (50 to 100 to 10 cm) and copper wire inside. I marked wire which goes inside by red. Then I connect it to a 1GHz high power source and I assume that this system will have low radio waves radiation losses. Let say less than 10% of energy will go to radiation the rest will just go through the wires as electricity. I assume also that electricity is uniformly distributed inside aluminium box (let say I make something special to make it true). Skin effect I do not consider now. It just will make resistance higher, but, for example, I can calculate skin effect later and take some measures to avoid it.

If this model will have big radiation losses, what can be done to decrease it? I need 1 side of the box to be flat (or almost flat) and that high-frequency current goes through it.
UDPATE1
Another point is that size of the metal box is very big, much bigger than internal wire size. Can it bring any complications? Will it help if I will fill space inside the box with insulator?
 A: This is similar to a coaxial cable, so it should work. One concern is that the impedance may be pretty high, so you may have signal reflection losses unless you match the impedance at the source and destination with transformers, resistors and such. Also possibly you may get additional issues as your impedance approaches the impedance of vacuum or open air (meaning your wire inside is relatively too thin).
Every transmission line has a wave impedance (the inductance divided by capacitance). Normally it does not depend on the length. An inch or a mile of a 50-Ohm coaxial cable is still 50 Ohms. The impedance depends on tbe geometry and is proportional to the ratio of diameters of the shield and central conductor. So it is high in your case. Every time it changes, the signal reflects back and forth. For example, if you have a 75-Ohm input and output connected by a 50-Ohm cable, then the signal would be bouncing inside the cable like light between two semi-transparent mirrors creating ghost signals and a big loss.
A dielectric creates losses and distortions. Better without it if you can. Otherwise use the one with the lowest dielectric constant. The skin effect depends on the surface of the conductors. Ideally they should be polished and protected from oxidation. Beware of the silver oxide myth. To keep the line impedance constant, why not use a pipe with a rod inside? 
You'll need to match your input and output resistance to the transmission impedance. The only way to do so is to observe the reflections on an oscilloscope and vary the input and output resistors (between the signal wire and signal ground) until the reflections are gone. You can do it at low power. You can't really work blind without an oscilloscope in the GHz range.
If your line impedance is much different from input and output, you may want to use transformers instead of resistors to match. Fir example, if the input and output imoedances are 75 Ohm, but the line impedance is 300 Ohm, then you'd need two 1:4 transformers. Make sure they are rated for the power and frequency. You can make them yourself too with a bit of knowledge. In tbe GHz range, every turn of a wire is a coil.
If your line impedance is close to the vacuum impedance (377 Ohms), as it may be based on the description, then I am not sure what the implications are. You may want to ask this as a separate question to get someone's expert opinion. Also, check this out regarding the silver oxidation issue: 
How is silver a better conductor than platinum?
