I got this article from Jon, for my consideration, before he posted it on the LocustWorld site.

I liked it so much, I slightly edited it, and ran it as a contributed article.

I've actually been using the LocustWorld London Mesh as my VoIP gateway to the world, calling people in London. But I could equally well call people through other LocustWorld meshes. There's one in Germany, for example; another in America on the West Coast, and so on.

Local calls, in each case...

Is a two radio or multi radio mesh router necessary to scale? I can't tell, even from news releases from Firetide because they never quote anyone other than the CEO (maybe a control freak?), so you never hear anything from their VP of Engineering about scalability and what not.

Just thinking about that hidden node problem with wireless, and the single radio products out there cannot be good for voice or video, especially after a couple of hops.

Any opinions on this?

I find it extremely hard to believe that single radio backhaul (and there are two radio systems that a single radio backhauls) can perform well in a dense citywide deployment. The reasons are explained under: http://www.meshdynamics.com/voip-results.html and in more details in a slide presentation made in 2005: http://www.meshdynamics.com/documents/VOIPCityWide_10_19_05.pdf

I am told that The USAF sponsored a study of different mesh architectures and their ability to support VOIP and Video in a battlefield environment. I know they tested our equipment and the test results as in the links above.

Francis daCosta
www.meshdynamics.com