Wireless channel advice

[truoc]

I'm forced (for right now) to use wireless for my desktop because of the location of the router and my computer so I'm asking for some advice on which wireless channel I should use. I have attached a screenshot of inSSIDer showing my network (KennHome) and the surrounding networks and channels. Looking at the info which channel would be best for me to choose? Thanks for any help!

 

[thiggins]

Based on signal strength, I would choose Channel 1. But the real decider is the amount of activity in other networks, which inSSIDer and other channel scanning tools don't show, unfortunately.

Have you considered using HomePlug (powerline) instead?

 

[truoc]

Based on signal strength, I would choose Channel 1. But the real decider is the amount of activity in other networks, which inSSIDer and other channel scanning tools don't show, unfortunately.

Have you considered using HomePlug (powerline) instead?

Thanks for such a quick response, I appreciate that! I will try Channel 1 per your suggestion. As for trying powerline, yes I have considered that as a next step. I had been using MOCA adapters and they were working just fine for awhile, but I had an issue crop up where my download speed would drop tremendously in the evening. I had a tech come and check everything out and he noticed a lot of noise in the line when I had MOCA hooked up. He tried a high pass filter (I think that is what it was called, can't remember exactly) that he had in the truck, but it didn't reduce the noise and he said he couldn't get another one so I guess I'm just out of luck on that. That is why I decided to go back to wireless for the time being. Powerline might be my next step!

 

[bobpow]

A question of INssider

What are the most important factors when looking at Insidder data ?

The Highest score the overlapping Chs or the Co-channels..

I tried Channel 3 and it gives me the highest score. but high over lapping channels.

Thanks for your help

 

[thiggins]

Do not use channels other than 1, 6 and 11. Due to channel overlap, your activity on Channel 3 will look like noise to users on Channels 1 and 6 and vice versa. Higher noise will mean poorer performance.

Number of users is probably the most important factor. However, it is only an indirect indicator of amount of activity, which is the most important factor and something that inSSIDer can't tell you.

The more simultaneous users of a channel, the less bandwidth for each user.

 

[bobpow]

Thank you ... I will stick to the lowest number of co-channels on 1,6,or 11
actually they are all about the same

 

[azazel1024]

Wouldn't 2, 5/7 and 10 also work? It just would leave no guard channel between (IE they'd abut, instead of having a 5MHz seperation) yours and anyone on the adjascent channels...which would mean the possibility of a small amount of interference, but likely very small.

Beyond the cochannel multi user issues, you also have the chance of environmental interference, which is more likely to be more spread spectrum, but its also possible that it might be more isolated to specific channels.

So really, 1 looks like a good place to start, but I think your best bet is to start there for a week or two, see how it goes and then try 6 for a week or two and then try 11 for a week or two. You may find one better or worse during times of the day or find that someone has a particularly busy network a lot of the time.

 

[L&LD]

What I try is sort by signal strength and pick the channel with the weakest signal for a quick and dirty setup.

The best way is to pick a primary channel as suggested and try each one to see which has the highest throughput and most responsiveness.

 

[RogerSC]

I'm forced (for right now) to use wireless for my desktop because of the location of the router and my computer so I'm asking for some advice on which wireless channel I should use. I have attached a screenshot of inSSIDer showing my network (KennHome) and the surrounding networks and channels. Looking at the info which channel would be best for me to choose? Thanks for any help!

Any chance that you could use an external USB wireless adapter on your desktop so that you could use 5GHz., assuming that your wireless router would do that? Given how busy the 2.4GHz. band usage in your area looks (from what I can see from the teensy weensy little print on your inSSIDer attachment), something to consider. I use 5GHz. between my router and laptop using a USB wireless adapter, and it works really well for me.

 

[bodean]

What I try is sort by signal strength and pick the channel with the weakest signal for a quick and dirty setup.

The best way is to pick a primary channel as suggested and try each one to see which has the highest throughput and most responsiveness.

By weakest signal, you mean a higher - number? (ie. -80 signal, using Wifi Scanner on my macbook air)

Also, do you have links to your recommended settings for RMerlin's firmware?

 

[azazel1024]

Yes, the bigger the number after the minus, the weaker the signal. Since that's a negative sign there, its actually the smaller number, the weaker the signal.

 

[RogerSC]

For example, a signal strength of -40 dBm is quite a lot stronger than a signal with strength of -80 dBm. dBm is a logrithmic scale, so the difference between -80 dBm and -40 dBm is a lot more than 2 times *smile*.

In fact, every time you double your signal strength you only get a 3 dBm increase.

 

[azazel1024]

Yup.

10dBi is 10x stronger than 0dBi.

-20dBi is 100x stronger than -40dBi. -40dBi is 2x stronger than -43dBi and so on.

This is in part why wireless is impressive, if incredibly difficult.

The fact that my laptop sees no slow down in tranfers at all between the -28dBi sitting 6ft from my router and -58dBi sitting a couple of rooms over, yet represents a signal strength 1/1000th of the signal I was getting sitting 6ft away with no obstructions, to me, is impressive.

 

[L&LD]

dBi is not dBm.

https://answers.yahoo.com/question/index?qid=20071112034240AAhk1S6

 

[stevech]

dBi is not dBm.

https://answers.yahoo.com/question/index?qid=20071112034240AAhk1S6

Short version:

dBm... RF Power on the log scale simplifies calculating power vs. range (ignoring obstacles). From physics, signal strength will halve with doubling the distance. Thankfully this is true, which is why we can communicate with, for example, the Voyager spacecraft. It's also why cell phones are not a hazard to human tissue, especially CDMA (Verizon/Sprint and now, LTE).

dBi ... is a *relative* measure of antenna gain. 0dBi = gain of an ideal spherical antenna pattern. An antenna with directionality (non-spherical) has some gain, stated in dBi. A common 1/4 wavelength vertical is about 2-3dBi gain.

dBm is the common measure of absolute power in watts, relative to a standard of 1 milliwatt (0.001 watt). 0dBm == 1mW. 10dBm == 10mW. 20dBm == 100mW (0.1W). In RF transmitters, you need to clarify the claimed power in real world power-bandwidth, i.e., how wide of an RF channel is needed to encompass x percent of the transmitted power. There are government (in the US, it's FCC) regulations on this, to prevent spilling power out of the unlicensed 802.11 band. And there are WiFi standards for in-band adjacent channels.

dB, alone, is dimensionless. But 3dB is twice-again, no matter if it pertains to RF or sound.

 

[RogerSC]

Short version:
dBm... RF Power on the log scale simplifies calculating power vs. range (ignoring obstacles). From physics, signal strength will halve with doubling the distance. Thankfully this is true, which is why we can communicate with, for example, the Voyager spacecraft. It's also why cell phones are not a hazard to human tissue, especially CDMA (Verizon/Sprint and now, LTE).

Hmmm...I would have thought that signal strength would go down by a factor of 4 (in a vacuum) if you double the distance, following the inverse square law. Been a while since I've had a physics class, though *smile*.

 

[stevech]

dBi is not dBm.

https://answers.yahoo.com/question/index?qid=20071112034240AAhk1S6

Hmmm...I would have thought that signal strength would go down by a factor of 4 (in a vacuum) if you double the distance, following the inverse square law. Been a while since I've had a physics class, though *smile*.

You're right... it differs if the measure is field strength vs. power.

 

[azazel1024]

Short version:

dBm... RF Power on the log scale simplifies calculating power vs. range (ignoring obstacles). From physics, signal strength will halve with doubling the distance. Thankfully this is true, which is why we can communicate with, for example, the Voyager spacecraft. It's also why cell phones are not a hazard to human tissue, especially CDMA (Verizon/Sprint and now, LTE).

dBi ... is a *relative* measure of antenna gain. 0dBi = gain of an ideal spherical antenna pattern. An antenna with directionality (non-spherical) has some gain, stated in dBi. A common 1/4 wavelength vertical is about 2-3dBi gain.

dBm is the common measure of absolute power in watts, relative to a standard of 1 milliwatt (0.001 watt). 0dBm == 1mW. 10dBm == 10mW. 20dBm == 100mW (0.1W). In RF transmitters, you need to clarify the claimed power in real world power-bandwidth, i.e., how wide of an RF channel is needed to encompass x percent of the transmitted power. There are government (in the US, it's FCC) regulations on this, to prevent spilling power out of the unlicensed 802.11 band. And there are WiFi standards for in-band adjacent channels.

dB, alone, is dimensionless. But 3dB is twice-again, no matter if it pertains to RF or sound.

I always mess up the unit designation.

As for gain, ideal quarter wave length IIRC is around 1.8dBi, half wave length is 2.15dBi.

Signal strength or power is reduced with the square of the distance. Field strength reduces linearly with distance.

Also to talk with voyager we use massive radio telescopes. When Galileos main antenna failed to unfold and we had to communicate using the backup, it requires some of the most powerful radio telescopes to talk to it at a very, very low signal rate (around 1kbps with compression, 160bps native IIRC). It was an isotropic antenna, for the low gain, and has a radio power of 15-20 watts. RSSI on Earth was -170dBm, or 10 zeptowatts (1x10^-21). The deep space network was used to communicate with it...using things like 35m radio telescopes. Those telescopes have a signal gain of 80dBi and a pointing of roughly 1/10,000,000th of the sky (or around 1/1th of an arc second).

Wireless is hard. By comparison if the failure had occured in Lunar orbit the radio power would have been around -138dBm and likely would have been able to be picked up with just a lowly 1-2m parabolic dish or so.

Voyager is significantly further away than Galileo was, but Voyager also has its high gain antenna pointed at us, at also is communicating at, I think, an even lower data rate (a few dozen bps at this point). It had a max, at Jupiter, com rate of 112.5Kbps. It also has a 12ft HGA. I am not sure the exact signal strength gain on it, but I'd guess at least in the 40-50dBi range...which means it was MASSIVELY better able to punch a signal to Earth compared to Galileo (and despite Voyagers current >>>10x distance greater from Voyager to Earth than Galileo to Earth).

Apollo used something like ~100w radios and resonably large sized parabolic (2m?) VHF/UHF antennas for radio and data links to Earth just from Lunar orbit and the surface of the Moon.

So, anyway, yeah I like using the wrong units.

dBi is antenna gain. dBd is also antenna gain, but normalized to a half wave length dipole. So 2.15dBi = 0dBd. The i is for Isotropic and the d is for Dipole...basically what the gain is normalized to (a perfect dipole or a perfect isotropic antenna. There can also be a negative gain on an antenna if the design is crap, its outside the radiation pattern, its high impedence or too low impedence, etc).

dBm...that is inverse square law. So doubling the distance would mean a reduction of 6dB (1/4th the received power). So -50dB at 30ft would be -56dB at 60ft and -62dB at 120ft. At ~1,000ft it would be -80dB. Hence long distances needing very high gain antennas.

Of course at 1,000ft if you added in a 8dBi antenna on each end over just a 0dBi (isotropic antenna) on each end, that -80dB becomes -64dB. Make it a pair of 12dBi antennas and it is -56dB...a pretty fair signal strength.

 

[bodean]

L&LD:
do you have links to your recommended settings for RMerlin's firmware?

 

[L&LD]

http://www.smallnetbuilder.com/forums/showthread.php?p=98452

I would recommend testing each channel on both bands separately though to find the best channel in your environment.