A six-sided box made of 0.5mm titanium (or indeed
any metal) will be an excellent Faraday cage - almost
no RF signal of any frequency will make it through.
Cellphones are mostly not metal (injection-molded
plastic, perhaps coated with a conductive material
in areas which are intended to be shielded) - and
in any case, the antenna is outside the shielded
part of the cellphone.
The recent debacle
with the Apple iPhone was due to their clever solution
to this problem - they placed the antenna on the
outer rim, where it could be contacted by the user's
hand, and if held so that the hand bridged one of
the gaps, signal dropped. Still, it was in some
ways a good solution to the dilemma of not enclosing
your antenna with a shield, when you box your product.
Antenna design for portable electronics is often
a very difficult field.
The antenna must
be outside the metallic enclosure, somehow. If you
cut a window in the metallic box, of the right size,
you will get some signal - but the antenna and box
combination will almost certainly be very directional.
"Directional" means "almost deaf in some directions".
How do I know that you can't put the antenna
inside? The "skin depth" is how far an electromagnetic
field ("radio wave") can penetrate into a conductor
(like metal). The formula is d=sqrt(f*pi*mu*sigma),
where f is the frequency in Hertz, mu is the magnetic
permeability of the material, and sigma is the conductivity
of the material. I don't have the numbers for titanium
at my fingertips, but for copper at 900 MHz the
depth is 2.2 microns. If the wave can't make it
through the metal, it doesn't reach the inside.
So what can you do? "Slot antennas" are a possibility,
as is the Apple iPhone idea.