I met Sanjit Biswas, the founder of Meraki, when he was a student at MIT and I was visiting Nicholas Negroponte. This was only last year which shows, among other things, a great side of USA and that is how fast somebody can go from being a super talented student to being CEO of a Sequoia funded company.

While at MIT, Nicholas recommended that FON use Sanjit’s open source software and we seriously looked into it. The problem that we have seen so far is that FON has chosen to collaborate with the ISPs and Telcos rather than fighting them, and we have come up with a formula that is seen as fair by them, namely that if you already pay for broadband you can use it elsewhere. What Meraki does, however, by meshing is a problem for us on two fronts. One is signal degradation, and the second is that meshing could be viewed as abusive by ISPs and Telcos as it extends their connections way beyond their points of sale.

At FON, we have come up with a formula that is simple: you share a little extra bandwidth at home and you roam the world for free. It’s a one to one ratio and FON promotes bandwidth growth as Foneros are more likely to buy broadband even if they travel a lot (indeed because they travel a lot). With meshing it’s hard to promote a balanced formula like this. Still, I see one situation in which FON could use Meraki and that is when a Fonero has their DSL or cable end away from a window and cannot project his/her signal to other Foneros on the street. In this case Meraki would act as a range extender.

So while FON would not promote that people just get a box like Meraki without paying for broadband and extend the signal of their neighbor as that would be seen as too antagonistic to ISPs and Telcos, we could promote Meraki as a FON signal extender for the Fonero and we are looking into that. Indeed, one of the challenges of FON right now is that some Foneros have their La Foneras inside their homes and they don’t reach the street. This is a problem with Linus Foneros only, as Bill Foneros want to reach the street to make money and they are more likely to place their La Foneras by the window. But some people try but fail to reach the window and the Meraki outdoor model that could send signal far into the street looks very appealing.

Also the people who could really use Meraki are the Muni Wifi projects. FON is from the home to the street and Muni WiFi is from the street to the home. Both are highly complementary. Meraki could have a nice spot in the middle of this. Bottom line, I very much like Meraki, and I think that we can cooperate and make Sequoia, our mutual investor, happier. In general, FON’s idea is to cooperate with anyone who helps propagate WiFi, but so far FON believes that the best way to extend signal without the degradation that meshing entails is to use more powerful antennas. That’s why next month we are launching the Fontenna, and we will send it for free to all of those who have consistently connected roaming Linuses or Aliens.

PD. A Bill at FON cannot roam but makes money selling access to the Community. A Linus roams the world for free but does not make money. An Alien does not build the Community, but pays for and uses WiFi.

Follow Martin Varsavsky on Twitter: twitter.com/martinvars

No Comments

John on March 20, 2007  · 

Martin, I am a Fonero, an investor in a major muni-wifi infrastructure supplier and a fan of Meraki so I, like you, have given a lot of thought as to how these 3 wifi “solutions” aide / hinder each others growth.

As you say, “FON has chosen to collaborate with the ISPs and Telcos”.

Why has FON decided to adopt this business model? And why do the ISP’s think its fair that “if you already pay for broadband you can use it elsewhere” ?

It doesn’t sound like the ISP’s and Telco’s I know !

It would be understandable if we are talking about a muni-wifi model where you pay once for broadband access ( ie wifi) and can then use it elsewhere. In a fixed broadband environment however it seems a bit illogical, as conceptually it costs them more for each physical DSL / Cable connection that has to be delivered to a house, apartment etc.

I think they just see the writing on the wall and are prepared to turn a blind eye to anything which enables them to hang on to their existing business a little longer – and that’s fair enough !

But is it fair enough for FON to collaborate with them and become a promoter (even reseller one day perhaps?) of fixed broadband when it is always promoting the benefits of community wifi ?

As a corporation, FON plays on the community aspects of being a Fonero and as part of this image(subtly) suggests that most Fonero’s are Linus’s. I signed up on this basis – contribute to the community, disruptive business model, it’s my connection why can’t I let other people use it ? etc. – but my conclusion is that if most users of FON access points are Bills and most providers of FON access points are Linus,’ then the real winners here are FON (you keep all the money from the Bills) and your ISP and Telco partners (because you push more fixed broadband connections for them) – I’m sure that being the smart guys you are that you will already have negotiated some sort of kickback from these partners –this effectively makes you a reseller.

From a community point of view I would be very interested to see FON’s stats on the number of Fonero’s that are actually using FON access points. Most people don’t travel the world like you and would only need broadband at home and at the office (I also travel a lot but I have never used a FON access point as they have never been convenient.).

My bet is that most people who use Fon access points do so as Aliens because they don’t have broadband at home. (Or possibly they are “pseudo Fonero’s” that have managed to “steal” usernames and password from genuine Fonero’s.)

This is good for FON’s business (so that’s how you’ve just got another round of investment !) but these Aliens may well be the sort of people that can’t afford broadband access at home and would actually benefit most from Meraki or Muni-wireless solutions. In this respect FON potentially hinders the growth of broadband “access” by assisting the incumbents to gain more fixed broadband customers.

My long term view ? – the wi-fi winners will be the ones who help the customer the most and I can’t see that being the fixed broadband ISP’s – I’d be looking for an alternative business model in which FON can provide a truly beneficial service to the community rather than just profiting from it.

I have a couple of ideas …..

3.0 rating

Paul Maunders on March 20, 2007  · 

Why not make meshing an optional feature that is disabled by default. Then people who want to build a mesh network and have a compatible ISP can enable it. I know that my ISP (AAISP) in the UK would certainly allow it.

3.0 rating

freechelmi on March 20, 2007  · 

Why do you think Meshing with degrade the signal ? You even talked about Meshing one year ago , But did not see any Mesh thing in your firmware, I guess if you healp standardise meshing techno by collaborating with openwrt and freifunk, it could be good for the community.

I don’t feel like you software is anything but a non friendly fork , but I could be wrong

3.0 rating

Martin Varsavsky on March 20, 2007  · 

Good idea. We will look into it. Thanks Paul.

3.0 rating

Martin Varsavsky on March 20, 2007  · 

John, thanks for your ideas!

3.0 rating

Martin Varsavsky on March 20, 2007  · 

Freechelmi,

I am not the only one who thinks meshing degrades signal. Indeed all WiFi degrades signal say if you compare to a direct cable connection. What I am saying is that Fon believes that one router with a better antenna maybe better than two routers but I can see why i may be wrong in some cases.

3.0 rating

juergen on March 20, 2007  · 

Hi Martin –
in my eyes Meraki is really cool and it would fit perfect to FON. But as the biggest Wifi community in the world – FON – should use the potential of Meraki especially in areas where FON is not so present. If you want to make money out of it by milking you Wifi – Meraki has a better payment method. It’s up to you how much you’re going to charge the users… and because of this reason it makes it somewhat rival to FON.
Some people want to make out some money by sharing the Wifi connection – and some of them don’t.

3.0 rating

Martin Varsavsky on March 20, 2007  · 

Mmemeteau,

we love meshing but we believe that only in a very dense and localized scenario the Fonero will take advantage of meshing. On the other hand meshing algorithmic are costly on CPU and data traffic, if they are used on environments where it is not adequate it will definitely affect the AP overall performance. But again, we are considering to use meshing at some point, maybe for specific deployments.

3.0 rating

Martin Varsavsky on March 20, 2007  · 

Juergen,

We will do our best to work with Meraki.

3.0 rating

Alex Mayorga Adame on March 21, 2007  · 

Martin,

Just in case you don’t check the FON English blog.

Here’s one idea, not meshing exactly, but I’d like you to consider it.

Have you thought of doing the following:

– Make La Fonera work as a repeater
* one SSID plugs to my current private WLAN
* second SSID does the FON_ thing
* the kicker, let me plug my xbox/xbmc/slingbox/htpc and connect to my private WLAN trough the Ethernet port.

I believe this is an everybody win approach, because of the following:
– If placed in my living room, chances are La Fonera would be closer to a window/patio/[front|back]yard and therefore a public area than if it’s located in a basement/attic and therefore linus/aliens get better reception.
– The fonero gets to plug non Wi-Fi enabled gadgets in the living room.
– La Fonera has less chances of getting bricked when following this guide http://www.dd-wrt.com/wiki/index.php/La_Fonera_Flashing

Let me know if you plan to implement this or if this is readily doable with plain Fonera firmware so I can refrain from hacking my way around the coming weekend.

My 2¢

3.0 rating

Martin Varsavsky on March 21, 2007  · 

Alex, I totally agree on your comment. This is a feature that is on our roadmap. We still need to get around some technical issues for this implementation to work on la Fonera firmware but we will definatelly add this to our Firmware.

3.0 rating

otilie on March 23, 2007  · 

Thanks for the inspiration. Nice and useful site.

3.0 rating

Robert Keyes on April 14, 2007  · 

Well, I’ve come a bit late to the discussion, but better late than never. I was going to suggest some mesh connectivity to Martin but it seems like he’s already talk to Sanjit and knows of Meraki. I know Sanit and some of the other Meraki people as well, as we helped each other out from time to time when the Meraki people were still students at MIT (they are currently listed as ‘on sabbatical’). I’ve taken the Roofnet software that MIT developed and modified it, as has Meraki, but we’ve diverged in our beliefs in how a mesh network should operate. Meraki believes in having a large number of less powerful mesh nodes, whereas XA believes in having a smaller number of more powerful units. Presently, Meraki and XAMesh (My company, XA Networks, calls our units XAMesh) can interoperate, but the message that Meraki has presented is that they may change their protocols without notification or explanation, and no longer be compatible with the roofnet protocols. Oh well. We too have changed the protocols, but we are committed to backwards compatibility as far as is practical.

It’s true that the mesh nodes on the fringe of the network, far from any gateways to high speed networks (notice I didn’t say ‘cable modems or DSL’) will suffer some speed limitations. There are various causes for this: some come from the natural wifi signal degradation as it moves through empty space, or is absorbed and reflected by objects. This causes the signal to be weaker, sometimes meaning packets get lost and have to be retransmitted, which causes higher latency and possibly a reduction in throughput (“bandwidth”). Another source of latency is that caused within the mesh node itself, as it does computation and routing. The way that various protocols suffer from bandwidth restriction, latency, and packet loss differs from protocol to protocol. We deal with each protocol separately, in order to present the best user experience (another way we differ from Meraki). I don’t want to get too technical here, but rather to give an introduction how meshing problems can be minimized.

Mesh networks still represent the most cost-effective way of providing wifi Internet coverage on a large area. If the proper mesh technologies can be coupled onto FON, the network would be greatly benefited.

Martin, please contact me in private to discuss some of the specifics. I’d like to work with you and FON. I am already a FON Fan and a ‘Linus’.

Regards,
Robert Keyes

3.0 rating

Martin Varsavsky on April 16, 2007  · 

Very interesting Robert. We will contact you!

3.0 rating

josh on May 23, 2007  · 

so ….would fon add mesh function in the furture?

3.0 rating

Martin Varsavsky on May 29, 2007  · 

Yes Josh, we will add it.

3.0 rating

Leave a Comment

Español / English


Subscribe to e-mail bulletin:
Recent Tweets