[Battlemesh] What hardware still works?

[Valent Turkovic]

On Mon, Mar 14, 2016 at 2:14 PM, fboehm <fboehm at aon.at> wrote:

> Am 14.03.2016 um 14:06 schrieb Paul Gardner-Stephen:
>
>> Unsure. Our humanitarian driven use-cases are somewhat price sensitive.
>> If the cheapest compliant device was around US$100 instead of US$20 -
>> $30, then we would probably be forced to look at making our own devices
>> (although we don't have the money for doing this at the moment).
>>
>
> What is if a 2-times more expensive device has twice the range?
>
> I'm asking because looking back many years, I always had the feeling that
> none of the cheap TP-Link devices had a range comparable to the good old
> WRT54G. Without illegal transmission power tuning of course :)
>
> Franz
>
>
If you have used devices with only one antenna then using devices with two
antennas (2x2 MIMO) could probably give you 2-times the range.

I didn't experiment with raising power on cheap tp-link devices because
from my wifi experience raising power is that it is always the wrong thing
to do. It could maybe even work for one isolated node and make "wifi
better", but if you are building network, having more power usually causes
also too much interference between nodes and range/quality doesn't improve.

Also raising more power doesn't make sense if clients can't also boost
power. Hotspot devices have more power and better antennas then clients
have on their phones and tablets. Clients side is the cause of most issues,
and range from client to ap is the issue. Even if clients hear your
powerful AP from 2-time longer distance they don't have power to return the
signal back (I'm thinking here on tablets and smart phones).

Wifi needs to be power balanced, if one side is much more powerful wifi
doesn't perform "better", but this is how most people think it works.

Summary - you will always have much better wifi network if you reduce range
of your routers, make smaller cells lower the power on them, but have more
of them deployed.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ml.ninux.org/pipermail/battlemesh/attachments/20160314/d054b33e/attachment-0001.html>