Improve WiFi Reception with Phased Array Antenna

[russ_watters]

I moved into a new condo last week. It has concrete block walls and concrete floors and ceilings. Needless to say, the wireless reception is terrible.

I bought two of http://www.hawkingtech.com/prodSpec.php?ProdID=183 - 90 degree directional antennas. They helped a little, but disappointingly little.

I've hacked into my router's software to increase the output as shown http://ihome.ust.hk/~westland/new_page_7.htm (I'll probably install a cooling fan so I can max out the wattage without killing it). It helped a little more, but its still a little slow and unstable.

So next up - phased array antenna. I was brainstorming on my way back from work and remembering how the Navy's Aegis system's radar works (I was in the Navy) - its a 3d phased array with hundreds of antennas and computer beam steering. I figured with a little effort, I can build a 2d, static phased array. I found a little bit of info http://www.wifi-forum.com/wf/showthread.php?p=1245#post1245 . It may be enough to build a 6-antenna array: I have the frequency, so its a piece of cake to mount 6 antennas at intervals of half the wavelength, but I'd like to investigate a little more.

My physics is a little thin in this area, so does anyone have any suggestions/resources for optomizing the array. Its planar, but how planar? Any way to make it more planar? How do I find beam width and strength? I was thinking of just making some graphical representations - concentric circles and tracing where they intesect: would that suffice? Any way to calculate/graph it instead (I have Maple and Matlab).

How about software: I'd like to do some testing - Net stumbler records/graphs signal strength but doesn't work with my card. Anyone know any similar software?

My condo is a split-level. For now, the router is by the front door. Walk straight back from the front door about 20 feet and you get stairs that go up 5 feet to the upper floor and down 5 feet to the bottom floor. A beam that's 10 degrees vertical and horizontal ought to be good.

Also, people are making Pringles can directional antennas - I'll look into them some more too.

Could be a fun, if geeky hobby for a while...

 

[GENIERE]

New condo! I’m surprised it isn’t wired so to speak. Does it have cable TV outlets on each floor? If so, try sticking a ¼ wavelength piece of wire in the outlets.

 

[russ_watters]

New condo! I’m surprised it isn’t wired so to speak. Does it have cable TV outlets on each floor? If so, try sticking a ¼ wavelength piece of wire in the outlets.

Heh, no - its new to me. The condo is 40 years old. It has some strange looking coax outlets (not like regular tv), but they don't seem to be connected to each other. I haven't investigated that much yet.

I mounted a fan on my router, kicked up the wattage, and installed two of those directional antennas. The signal is pretty good, but I want more...

 

[wimms]

With wifi, upping wattage is usually not what you'd want. If at all, it might help on the remote side if signal attenuation is high due to distance or obstacles. But main issue is with reflections and sensitivity of receiver.

Fancy antennas help sensitivity, but together with narrowing directionality. From your post I couldn't get what your intended application is. If its link to remote site, then it could benefit from single good directional antenna and leaving other antenna outlet empty, if its for inhouse lan, then double-omnis should be good for your access-point and its the end equipment that needs directional antennas, perhaps those you purchased.

I may be wrong, but its my understanding that wifi devices use 2 outlets not for getting directionality of phased array, but to distinguish reflections and noise signals from the main signal. Not sure here, but they might not even use both for transmitting, only for receiving. DSP's inside wifi units make lots of assumptions when operating in 2-antenna mode, and attaching directional antennas or phase arrays there confuses them.

Use of phase arrays in wifi sounds like attempt of frequency reuse rather than increasing range or performance of individual directions. To make good use of them you'd need to change DSPs in your equipment. Sure, rejection of other users signals has indirectly good impact on range, but that's most probably needed for ISP access-point with quite many users.

Regarding reception issues with wifi, I recall that main trouble is with reflections from nearby walls, especially those on opposite side from signal source. Directional antennas that have parasitic lobes behind their direction can have awful results when mounted on the wall, basically because DSPs are unable to distinguish reflections from signal. In that sense, 6-antenna array could be worst thing to try.

Very good highly directional flat panel antennas can avoid most reflection issues, and that's why they are often used with single antenna outlet, leaving DSP processing unused to cancel reflections, gaining even little more sensitivity for that.

 

[russ_watters]

Yeah, I need to look into the router antenenna thing: it wouldn't make sense for it to have directionality straight out of the box.

I have it working reasonably well right now, but its still a little unstable. I may end up just moving the router, but I need to do some coax wiring.

 

[GENIERE]

Re: The old coax outlets, the only thing I can think of is that they may be for 300 ohm cable.