One reason why WiFi can be variable
 

Many of us have reported significantly differing experiences with WiFi reliability and our tablets. This has been bothering me for some time because, as an analyst, I don't like noise and look for causation.

I think I've stumbled upon one answer to some of the reliability problems with WiFi.

The current issue of Scientific American contains a fascinating article on using light within rooms to provide "last foot" broadband. Within that article is a sidebar explaining pros and cons of radio vs infrared approaches. I was stunned at one of the radio drawbacks and how aptly it may apply to some users.

Radio signals can directly penetrate walls, which is a security risk, but they can also as most know bounce off of some (such as highly reflective) surfaces. This depends on signal strength, angle of emission, etc. The result is a multitude of waveforms permeating the WiFi environment. Lightbound signal receivers can handle this. However, in the case of radio, waveforms can be distorted. Stray signals can result that are up to 180 degrees out of cycle with the original true signal. In this worst case scenario the data completely cancel each other! Even in lesser destructive scenarios, the datastream is distorted and the end result can still be packet loss.

With this in mind, design of your wifi environment takes on new importance-- especially the placement of your router and/or access points. I recommend locating the router in the room in which you'll normally use your internet tablet, and relying on copper for other connections if you can. Repeaters or access points may become a necessity in a multiroom or mixed (indoor/outdoor) environment.

I hope this helps those of you experiencing wifi problems that others don't. I wish this had occurred to me earlier!

Current article: http://www.sciam.com/article.cfm?cha...7B407F9FB74A7B

Older article: http://www.sciam.com/article.cfm?art...EDF&sc=I100322

Re: One reason why WiFi can be variable
 

Also note that the make up of the wall is also important. Walls made of lath and plaster have chicken wire in them to help with rigidity and strucural support. This kills wireless signal. Newer drywall is much better.

Another impedence of wireless signal could be the old style 'BX' electrical wires. They have a metal jacketing that in large amounts can affect signal quality.

I actually went out and bought a second router, subnetted my network and run two wireless networks because of signal issues in the house i am in. Lath and plaster walls dissapate the signal. I subnetted because i dont want to run WEP encryption, and that is the only encryption supported in bridge mode for the hardware i have.

Re: One reason why WiFi can be variable
 

Great post, Texrat.

Here are some additional tips:
Radio signals are radio signals. The higher the antennae, the better.
I have had problems with all my devices on my desktop. Why?
- Six bluetooth devices
- Three wifi devices
- One cell phone

I also have NINE neighbors with WiFi hotspots in their home - all can be seen by my tablet. That's part of the reason I upgraded to my beefy WRT350N.

When possible, move the tablet and/or access point physically away from a cluster of radio devices.

Re: One reason why WiFi can be variable
 

And check which channels your neighbours are on.. select one for yourself that is at least 2 channels away (that is, leave a gap).

Re: One reason why WiFi can be variable
 

These are all good points. I had just realized the stray signal issue may not have been covered here (I hadn't seen it yet anyway).

Re: One reason why WiFi can be variable
 

Actually I'm surprised how far the connection from my Linksys router to the N800 lasts. My router is not optimally placed: just next to it is a big , heavy old HP laserjet full of reflecting metal. All is located in the first upper floor. But even then, I had reception over a distance of almost 100 meters outside on the road.

No problem with a few walls between the devices and down 2 floors to the cellar through cement walls.

Re: One reason why WiFi can be variable
 

Your mileage may vary. ;)

Re: One reason why WiFi can be variable
 

As I understand it, this is the purpose of MIMO in wireless access points, as the multiple antennae provide path diversity.

Re: One reason why WiFi can be variable
 

Mirrors, metal studs, lots of wires in the wall, old fashioned plaster walls, ducting. Pretty much anything that will heat up in the microwave, or you shouldn't put in the microwave, is bad for wifi. If I think of anything else I'll add it.

Re: One reason why WiFi can be variable
 

Hell, microwave ovens can be bad for wifi! My mom's knocks out dad's xbox link... at very inopportune times. :D

Re: One reason why WiFi can be variable
 

Check this out. Watch the video.

http://web.mac.com/barrywoods/iWeb/S...nna%20Mod.html

Re: One reason why WiFi can be variable
 

Just made a test in the kitchen near the running microwave oven and couldn't see a difference.

But I didn't place the N800 inside the oven...:D

Re: One reason why WiFi can be variable
 

Some microwaves are leakier than others. If it's bad enough to make your devices drop of the network, you might want to replace it.

Re: One reason why WiFi can be variable
 

Wow, THAT is noisy.

I'm gonna stick to firewood and lighter fluid...

Re: One reason why WiFi can be variable
 

That's a new microwave, it doesn't do anything to the network. You should see one that actually leaks! My cordless phones look pretty neat too.

Re: One reason why WiFi can be variable
 

One more thing: Those famous 2.4GHz cordless phones are often very good WiFi signal killers! I have one that I can not use at all... if it is connected and someone calls it both my N800 and desktop lose connection right away...

I assume most newer 2.4GHz phones are more "WiFi friendly". Or use 5.xGHz phones? Currently I use older 900MHz phones at home and they do not interfere at all... :cool:

Then, I suspect some of my neighbors have bad 2.4GHz cordless phones since sometimes I experience times when my WiFi keeps dropping connection all the time... the signal strength is good when I check that in my desktop WiFi applet. After some time (Guessing how long the phone call lasts...) the connection restores and is rock solid after that!

Re: One reason why WiFi can be variable
 

I would stick to the 900mhz phones if you can. They go farther and aren't stopped by the dame materials as 2.4 or 5Ghz.

Re: One reason why WiFi can be variable
 

A few other things about 2.4 GHz.

Indeed reflections are a problem, called multipath in the litterature. But normally relocating the router or computer 10cm away should solve things. And indeed some routers have multiple antennas to address that problem.

Armed concrete stops radio signals. Bricks and stones attenuates them. In the US, most houses are built out of wood, but old houses in Europe are different. The 20cm thick brick walls at my place eat 20dB per wall. The 40 cm stone walls at my grand parents place eat at least 40dB...

Interference is a problem in most cities. Bluetooth is relatively well behaved, modern microwave ovens should not be a problem (and if they are exchange them!), 2.4 GHz phones are uncommon in Europe (which mostly uses DECT at 900 MHz). OTOH, the 13 "channels " of wifi arent really independent, theyre is a frequency overlap between adjacent channels.


Last but not least, if you know where the clients will be, a directive antenna on the router may be a good solution (if you know what you are doing). The idea is to "light" only the space where the clients will be (say: your house). Advantage: the router will only "see" in the direction it sends signals, so it will not be disturbed by interfering signals coming from the other direction (and it will interfere less with them as well).

You can't use a directional antenna on the N800, unfortunately...

Re: One reason why WiFi can be variable
 

This thread turned out more interesting than I'd imagined.

I see a possibility here: Nokia should gather this info and provide a robust wifi educational pamphlet with the tablets.

Re: One reason why WiFi can be variable
 

This did give me a idea. I had ´metal´ Iron Maiden CD box next to my Wifi router and its antenna.. Could big metallic object cause problems?

I dont know.. But I did remove it to elsewere. Anyway wifi signal strength that N800 did show was max even before.

Re: One reason why WiFi can be variable
 

It wasn't the metal case, but the metal music your router doesn't like :p

just kidding, I love OLD iron maiden albums too

Re: One reason why WiFi can be variable
 

It probably wasn't helping things. Both of my wifi routers are as high as I can get them with nothing around them. Both are on top of shelving units, one in the library, the other in the garage.

Re: One reason why WiFi can be variable
 

Oh, definitely.

Re: One reason why WiFi can be variable
 

Well, good to see some correct points being made here, others I'll chalk up to urban legend ;) .

I was involved with the IEEE 802.11 working group that develop the early incarnations of wi-fi so its near and dear to me. Of course that was the 2 Mbps version using Frequency Hopping rather then Spread Spectrum that 802.11 b/g is based. While I could get into all the gee-whiz gut wrenching physics regarding radio signals I will spare everyone :D, but I will say there are some common mistakes I see people make with wi-fi.

1) Most people tend to align the antenna incorrectly. A lot of those dipole antennas have whats called a fresnal zone right on top of them, which means none to very little radio signal is propagates out of the top of the antenna but instead propagates vertically from the sides. This usually is a problem for people who place their wi-fi routers in the basement and then point the antennas completely up and down at a vertical angle. If you're on the 1st or 2nd floor your wi-fi signal will be very weak, so if your wi-fi router is in the basement, adjust your antenna to be at a 45 degree angle and if you happen to be right above your wi-fi router then adjust your antenna to be horizontal rather then vertical.

2) Most people accept the default channel on the wi-fi router, unfortunately, as someone else mention in this thread, that increases you're noise floor with competing routers. So pick a different channel, but pick one that is 3 steps + or - from a competing channel. Unfortunately wi-fi channels overlap into other channels slightly with the non-over lapping channels being 3, 6 & 9. Being 3 steps away from another wi-fi will not cause RF interference.

Re: One reason why WiFi can be variable
 

Good points bac522.

Another little trick: people have a natural tendency to go "up the dial" when changing channels. You improve your odds of reducing interference by going down from the default.

Re: One reason why WiFi can be variable
 

Oh no, now you've gone and told everybody! I was happy with the ample room at the bottom... :)

Re: One reason why WiFi can be variable
 

ummm... d'oh?