Antenna pattern graphs or positioning tips

[kevinkar]

Hi all,

I remember a long time ago I was able to find some sites that showed typical antenna signal patterns of various routers as well as providing some advice on how to position them and arrange the antennas for best signal transmission.

Newer routers and access points apparently seem to be dropping external antennas and may or may not include stands or mounting holes such that you can lay them flat, mount them on a wall, or stand them up on a desk. There do not seem to be any major recommendations as to which way to place them nor are any distinctions being made as to a particular user's situation.

The end result is that you just put them where they are and hope for the best but that has to be less than optimal, right? Laying a router flat with the antennas in plane has to be throwing out a different pattern than standing it up on its side, right?

So is there any way to see "published" antenna patterns on the more modern crop of internal antenna devices and position them for best reception or is it really not a problem anymore?

I have a Linksys E4200 in my office upstairs laying flat. Directly below that on the entertainment unit is a D-Link DAP-1522 also flat. I'm getting around 33Mbps between the two which isn't great and, incorrect configuration notwithstanding, I'm wondering if the antenna patterns are just out of sync with the two being two parallel planes and not really intersecting. Would it be better to have them perpendicular or in-plane?

With little to no details about antenna patterns, I'm not sure how to best position them without lots of experimentation. If there were graphs I could check, that might make it easier.

Any thoughts on this?

I should add that the two are connected via N-only and the channels set to auto (20/40). I've also got them connected via WPA2 AES/TKIP which I should probably change to AES-only to help througput as that may be my biggest bottleneck.

Thanks,

Kevin

 

[stevech]

I doubt you'll find antenna patterns for internal antennas. The FCC Part 15 filing for the product is public information on line, if you can determine who the OEM was for brand x, and search for that.

But as an RF guy, I'll make an educated guess that orientation H/V won't make a significant difference, especially if there are some walls/floors in the path. In line of sight paths, cross-polarization (one end H, other end V) can make about 6dB difference. But non-line-of-sight, with scattering and multipath, tends to depolarize all the reflected signals as they sum up at the receiving device. To test this, create a good line of sight path that's as long as possible - no doorways in the path. Like in the back yard or some such. Then observe the received signal in dBm (not bars or percent) with nothing moving and see what the variation is, get some sort of average. Repeat for the rotated alignment. Rough measurement, but I suspect you'll see a few dB, this being tiny compared to the path loss which even in line of sight will be many tens of dB.



The truly MIMO 11n product probably have 3 antennas and combinatorial diversity (summing best two signals), whereas older/cheaper 11n and all pre-11n, and 11a have, instead, switched diversity. In such, a few bits in the header in WiFi frames are examined for signal/noise ratio and the best is used for future bits/frames. How often this is done is not in the standard and is a source of sloppily designed product low level firmware.

 

[thiggins]

Not sure what you mean by "between the two". The 33 Mbps is throughput via a wireless bridge between the two?

If you are on the 2.4 GHz band, use 20 MHz mode. If in 5 GHz, you can try auto 20/40. But generally, 40MHz mode helps only if you have medium to strong signal strength.

You must use WPA2/AES to get link rates > 54 Mbps.

As stevech points out, 802.11n uses multipath and reflections to its advantage. So antenna orientation is less sensitive. But you can run
some simple experiments reorienting one or both of the units and see if you can improve throughput.

 

[kevinkar]

Not sure what you mean by "between the two". The 33 Mbps is throughput via a wireless bridge between the two?

Yes, that's what I mean. Using a large file transfer (large DVD ISO files) from the NAS box upstairs connected to the router to a small media player downstairs connected to the DAP-1522, the average throughput for the file transfer calculates out to about 33Mbps which is corroborated by the two device's GUI interfaces. That's using TeraCopy, not Windows Explorer as TeraCopy copies faster.

It's a Patriot Javelin S4 with Seagate 3TB Barracudas inside connected to a Linksys E4200 (N-Only 5GHz 20/40 WPA2-TKIP/AES) wireless to the D-Link DAP-1522 connected to a Mvix Ultio Pro (which only has a 10/100 connection) with a 2TB Seagate Barracuda inside. So the bottleneck would be the Ultio but I'd expect closer to 70-80Mbps overall through this chain.

So my guess is that it's not the antenna pattern, though I will experiment with positioning, and it's more likely the TKIP/AES and I should first change to AES only and try again. Of course large files tend to transfer differently than smaller files like MP3 files which seem to transfer quicker.

Thanks for the input so far,

Kevin