WiFi Mesh with IP Telephony Integrated (Like MeshPotato MP01)

102 views
Skip to first unread message

Saifuddin Mahmud

unread,
Jan 21, 2015, 6:11:17 AM1/21/15
to village-...@googlegroups.com
Hi anybody know about any product like "MeshPotato MP01". ATCOM saying that they do not do that any more!

Steve Song

unread,
Jan 21, 2015, 10:17:47 AM1/21/15
to Village Telco Dev
Hi,

The next-generation MP2 is available at the link below in a few different models.


Unfortunately we do not yet have an outdoor version of it available.  

Regards... Steve

On Wed, Jan 21, 2015 at 7:11 AM, Saifuddin Mahmud <saifuddi...@gmail.com> wrote:
Hi anybody know about any product like "MeshPotato MP01". ATCOM saying that they do not do that any more!

--
You received this message because you are subscribed to the Google Groups "Village Telco Development Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to village-telco-...@googlegroups.com.
To post to this group, send email to village-...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/village-telco-dev/5aa446e2-cf00-488e-ae3e-3a010d127049%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.



--

Eliot Osorio

unread,
Jan 21, 2015, 5:11:22 PM1/21/15
to village-...@googlegroups.com
Hi, i will start my project with a mesh potato, but i have a question
how many clients support each access point, and how many access point 
can be connected

Best regards.. 

Eliot


Date: Wed, 21 Jan 2015 11:17:45 -0400
Subject: Re: [vt-dev] WiFi Mesh with IP Telephony Integrated (Like MeshPotato MP01)
From: st...@villagetelco.org
To: village-...@googlegroups.com

T Gillett

unread,
Jan 21, 2015, 6:06:23 PM1/21/15
to village-telco-dev

Hi Eliot

I have used MP02 devices in classrooms with up to 35 student laptops connected to the Access Point wifi interface without issues.

By comparison, we found that entry level commodity routers maxed out at around 15 concurrent connections running either proprietary operating system or OpenWrt. Beyond that level, connections would be randomly dropped.

Regards
Terry

Steve Song

unread,
Jan 22, 2015, 9:33:27 AM1/22/15
to Village Telco Dev
And it terms of how many access points can be connected, the significant limiting factor is the bandwidth loss through each "hop" on the mesh.  An MP loses about 50% of throughput with each hop away from the upstream gateway.  Thus you want to design your mesh so that no MP is more than 3 maybe 4 hops away from the gateway.  While this does limit scale, you can overcome this by creating a wireless backbone, ideally on 5GHz, that can underpin the growth of the mesh.  Think of the backbone links as arteries and the MPs as arterioles/capillaries.


Cheers... Steve


For more options, visit https://groups.google.com/d/optout.

Breno Jacinto

unread,
Jan 22, 2015, 10:49:39 AM1/22/15
to village-...@googlegroups.com
Hi Steve and others,

I have been thinking about that for some time and since the
release of 802.11ac (http://wiki.openwrt.org/toh/tp-link/tl-wdr7500)
I've been wondering if I could get even higher throughput over the
mesh as well as increase the density of the network while not losing
performance. A few points for discussions:

1) How is 5Ghz support for SECN/OpenWRT? After reading this thread:
https://forum.openwrt.org/viewtopic.php?id=53703, I was concerned that
support for Ath10k driver is still immature. Is this the case for
WDR4300 as well?

2) How are those larger mesh networks handling the scale today? Is
anyone aware of how Freifunk, for example, are handling scale?

3) Any plans to support 802.11ac on MPs?

cheers,
---> Breno Jacinto
---> Instituto Federal de Educação, Ciência e Tecnologia de Alagoas (IFAL)
-----> http://www.ifal.edu.br
---> Life is Choice. You can choose to be a victim, or anything else
you want to be. (Sócrates - Peaceful Warrior) <--
> https://groups.google.com/d/msgid/village-telco-dev/CAD_CWO2m%2Br%3DWnUC_fDHx-MwJiPsFqfU3%3DfRi_Ae_rk04ogPKPw%40mail.gmail.com.

T Gillett

unread,
Jan 22, 2015, 6:12:31 PM1/22/15
to village-telco-dev
Further to Steve's note...

For setting up wireless backbone you may find the SECN-Duo firmware useful. 

It allows the use of USB wifi devices to provide a second radio for the MP2.  The USB radio can operate on a different channel and with a specialised antenna.

For details see the wiki page:



T Gillett

unread,
Jan 22, 2015, 7:23:47 PM1/22/15
to village-telco-dev
Hi Breno

Thanks for the link. That is an interesting post.

I am not sure how relevant the specific results will be to real world mesh based networks since it is a test from a router to a laptop over a short distance. 

I suspect that when these devices are used in real world mesh scenarios there will be many other factors that will limit the data throughput to a much greater extent.

The fact that the UDP bandwidth is maintained under OpenWrt is a positive.

Some time ago I did some simple end to end tests with a pair of WDR4300 devices running SECN/Openwrt-AA using scp to transfer large files between laptops connected to the routers. 
This gives a realistic measure of how the system will perform in practice.
 For details see the wiki page:

With the mesh operating on the 5GHz band, I got end to end transfer rates of around 100Mbps.
By comparison, replacing the mesh link with an Ethernet cable gave 200Mbps. 

A mesh that can deliver around 100Mbps effective throughput is probably fairly useful as a backbone for a small campus network such as a remote school or community, which is the primary purpose of the SECN firmware.

In terms of adding 5GHz band support to MP2, an option that may be useful would be to use a USB based 5GHz wifi device as a second radio. This currently works well with 2.4GHz devices. 

Regards
Terry

Breno Jacinto

unread,
Jan 29, 2015, 8:50:51 AM1/29/15
to village-...@googlegroups.com

Hi Terry,

  Thanks for your link. I could see some variable results on the 5Ghz band. Do you think that it could be driver related?

Cheers,

T Gillett

unread,
Jan 29, 2015, 12:46:29 PM1/29/15
to village-telco-dev

Hi Breno

It is a bit hard to know what causes the short term variability in results.

I would expect that a driver problem would display a longer term result.

The only thing that would obviously explain the short term variations is environmental noise.

Those tests were done in a pretty busy wifi environment.

When time permits I plan to do more tests around the USB devices particularly, and they will be done in a quiet open field environment.

BTW  I have posted SECN 3 Duo Alpha 10 firmware with support for 5GHz USB devices if you want to test.

Regards
Terry

Song, Stephen

unread,
Jan 29, 2015, 12:59:43 PM1/29/15
to Village Telco Dev
Reply all
Reply to author
Forward
0 new messages