<html><head></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">Hi Josh,<div><br></div><div>Yes, the nodes were access points and mesh links. </div><div><br></div><div>It may have been that having pico stations on the ground (not on the roof) at no more than 200 meters apart avoided some of the wireless problems that can happen when a device is speaking to devices close by and other routers further away. Or, it could be that the number of connections to the access points were relatively small (because they were covering a small area).</div><div><br></div><div>I think there needs to be more testing of this phenomena (maybe someone knows of some research or would do some testing?). </div><div><br></div><div>Best,</div><div>Ryan</div><div><br></div><div><br><div><div>On Mar 7, 2015, at 3:29 AM, Josh Harle wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"><div dir="ltr"><div class="gmail_default" style="font-family:arial,helvetica,sans-serif;color:rgb(7,55,99)">Hi Ryan,</div><div class="gmail_default" style="font-family:arial,helvetica,sans-serif;color:rgb(7,55,99)"><br></div><div class="gmail_default" style="font-family:arial,helvetica,sans-serif;color:rgb(7,55,99)">Just confused what your set-up was for <a href="https://commotionwireless.net/blog/2013/10/30/building-popup-mesh-networks/">https://commotionwireless.net/blog/2013/10/30/building-popup-mesh-networks/</a>, since I am using the same hardware and presumably similar firmware version, and your numbers seem good. I'm imagining that a number of those nodes were APs too, since you had clients connecting and testing?</div></div><div class="gmail_extra"><br clear="all"><div><div class="gmail_signature"><div dir="ltr"><font face="verdana, sans-serif" color="#073763">Kind Regards,<br><br>Dr Josh Harle<br>____________________<br></font><div><a href="http://joshharle.com/" target="_blank"><font face="verdana, sans-serif" color="#073763">http://joshharle.com</font></a><div><a href="http://tacticalspace.org/" style="font-family:verdana,sans-serif" target="_blank">http://tacticalspace.org</a><font face="verdana, sans-serif" color="#073763"><br></font></div><div><font face="verdana, sans-serif" color="#073763">ph: +61 (0)409 771 163</font></div></div></div></div></div>
<br><div class="gmail_quote">On 4 March 2015 at 01:25, Ryan Gerety <span dir="ltr"><<a href="mailto:gerety@opentechinstitute.org" target="_blank">gerety@opentechinstitute.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
We experienced the same issue while setting up a 10 node network recently. We turned off the access points on all the rooftop/mesh nodes, and connected a cheap access point router (running standard firmware). The Commotion node’s LAN port is plugged into the WAN port of the Access Point. In many cases this is best anyways -- since you either want good internal coverage (dragging an ethernet cable inside) or you want to provide an access point for many people outside (so you might want multiple access points). Our wireless engineer is currently writing up documentation to recommend that people dont use a router as both an access point and mesh.<br>
<br>
I agree with Adam that this is just an unfortunate limitation of wireless.<br>
<br>
Another option is a dual radio is something like this: <a href="https://commotionwireless.net/blog/2014/11/05/do-it-yourself-antennas-for-community-networks/" target="_blank">https://commotionwireless.net/blog/2014/11/05/do-it-yourself-antennas-for-community-networks/</a><br>
<br>
Best,<br>
Ryan<br>
<br>
<br>
_____________________________________<br>
Senior Field Analyst, Open Technology Institute<br>
New America Foundation<br>
1899 L St., N.W., Suite 400<br>
Washington, DC 20036<br>
<a href="tel:%2B1%20202%20492%208841" value="+12024928841">+1 202 492 8841</a> (c)<br>
<div class="HOEnZb"><div class="h5"><br>
<br>
<br>
On Mar 3, 2015, at 9:02 AM, Josh Harle wrote:<br>
<br>
> Hi All,<br>
><br>
> I've been installing 20 picostation M2HPs in a town to try to set up a mesh network, partially based on the "Building Pop-up Mesh Networks" article.<br>
><br>
> I've found that when I have a handful of mesh nodes, if more than just one of them has an access point interface enabled, the ETX goes from about 1 to somewhere between 5 and 7!<br>
><br>
> I've been talking to Adam Longwill about this, who has found similar behaviour. I don't understand how to mitigate this. Isn't it expected behaviour to have all nodes as APs too? How do I get area coverage with WiFi client access?<br>
><br>
> It seems to impact performance so much to have more than one AP turned on in the network that the only thing I can think of is to have one Picostation for mesh, connected to one souly for AP on another channel.<br>
><br>
> I would be hugely greatful for advice on this: I'm in a remote Aussie town for just a couple of days trying to troubleshoot this.<br>
><br>
> Kind Regards,<br>
><br>
> Dr Josh Harle<br>
> ____________________<br>
> <a href="http://joshharle.com/" target="_blank">http://joshharle.com</a><br>
> <a href="http://tacticalspace.org/" target="_blank">http://tacticalspace.org</a><br>
> ph: <a href="tel:%2B61%20%280%29491%20155%20985" value="+61491155985">+61 (0)491 155 985</a><br>
</div></div><div class="HOEnZb"><div class="h5">> _______________________________________________<br>
> Commotion-discuss mailing list<br>
> <a href="mailto:Commotion-discuss@lists.chambana.net">Commotion-discuss@lists.chambana.net</a><br>
> <a href="https://lists.chambana.net/mailman/listinfo/commotion-discuss" target="_blank">https://lists.chambana.net/mailman/listinfo/commotion-discuss</a><br>
<br>
<br>
<br>
</div></div></blockquote></div><br></div>
</blockquote></div><br></div></body></html>