Starting with one or two basics;

- Keep the losses down
- Channel as much of the radiation as possible in the direction you need it

Losses in this sense more come from coax runs and connectors. Keep the
run short, dont use joiners and if possible mount the WiFi box close to
the antenna. You have to also allow for obstructive losses like trees
and buildings. The other path loss is from errant reflections. Moounting
the antenna high up in the clear resolves many of these issues.

Radiating in the right direction isnt rocket science in theory. The
average "omnidirectional" antenna is actually "all directions" only
parallel to the ground. ie you dont waste radiation skyward or warming
the dirt! In terms of distance covered though an omni of this kind wont
be as good as employing an antenna that "beams" in one direction.
Consider this as a kind of flashlight with reflector and you'll get the
right idea. In terms of time/construction/distance investment you'll do
a lot better with this kind of antenna than an omni, elthough you will
need to point them in the right direction. Which one you choose depends
on its intended use.

You have probably figured out that the antenna gain figure (in dB) is
the measure that is used to determine how much better one antenna is
over another. It is a log comparison of the desired radiation direction
"strength" vs that of an antenna that radiates eqaully in all
directions. An omni however will almost always have a lower gain than a
"beam" type antenna. 9 and 12 dB are some fairly high omni gains but
"beam" gains can go 15-30dB. The higher the number the greater the range
in the desired direction.

So my view.. If you intend one site being fixed and the other movable
then you'll at least need an omni at the base site. If you only want a
point to point link then a "beam" variety at each end is the way to go.
I wouldnt go for "best" but engineer the path/strength requirements to
determine the antenna gain needed. If I was making the antennas I'd
probably use end fire helixes as they seem to be most forgiving in
construction errors vs best gain.

You might want to tell us your end application, distance to cover,
movement of sites and so on so we can suggest something a little more
finite.

Cheers Bob

Penang wrote:


What's your suggestion / opinion on how to construct an excellent
(pair) of MIMO antenna for wifi (802.11 b/g/n) ?

Thank you all !