 Hi, my name is Dona Wunama, I am here to present the VillageTelco project. First question is who am I? First I'm a software engineer, I'm an expert in wireless systems, telecom systems and embedded software. Also the founder and city of wireless dialogue, I'll come to that later. The main focus of this presentation is to talk about the VillageTelco project. First question is what is the VillageTelco? Effectively, the VillageTelco is a local community telephone network. What that means is that, I'm not sure if you guys are aware of the wireless mesh networking. With wireless mesh networking, effectively every node becomes a gateway and a router as well as an access point. What the VillageTelco does is take that concept and integrate in an analog telephone adapter so you can effectively plug in an analog device and an analog telephone and make phone calls. The key point of the project is it sets up in minutes, so you don't have to be a bash script wizard to set one up. And also one of the key questions was why do you need a VillageTelco? The main problem in developing countries is that for access you need mobile phone towers. In Europe, we're blessed with the fact that there are copper networks and actually fiber networks. In emerging countries, the government would not actually invest in copper networks, so it's left to private companies to invest. Now, as obviously, looking at the bottom line, they're more interested in getting the most users onto the network as possible so they're only focusing on urban areas. And I've mentioned before it's a wireless mesh technology. I have a sample for those who want to have a look at the unit. At the moment, it's around $100 US dollars per unit, but that price will go down over time. The question is why? I'm not sure if you guys are aware of the mobile telephone business effectively in developing countries, place like India and Sub-Saharan Africa. Effectively, calls are very expensive. The rule part of the world has basically just been ignored, because the substance farmers, they cannot actually buy a mobile phone. Or even if they do get a mobile phone, they cannot afford the tariff prices. So effectively, you say, well, how do we then enable these guys to become connected? And obviously, we can not only use it for developing networks, developing countries, we can also use it for communist wireless, like the FreeFunk in Berlin. The features of it, basically, it's locally owned. No spectrum license, the same as a wireless mesh, it uses 2.4 gigahertz or the 5 gigahertz spectrum. Everything about the whole architecture is actually open sourced, and I'll come to that later. And obviously, the focus on rapid deployment. And this is the construction part. We basically have a billing server, and within the box itself, there's embedded asterisks, embedded inside OpenWRT. And if you guys are familiar with OpenWRT, obviously, this would be a no-hose bar for you to do. And we've actually developed a truly open source mesh management mapping tool. So you can actually be able to configure and actually administer your network. This grab dot shows the amount of internet users in Africa. So wireless dialogue is basically trying to encourage internet usage. As you can see here, 70%, 80% of the population are connected, and we're talking about less than 10% of the population. So our business model is to take the McDonald's type model. So effectively, we get local entrepreneurs to actually buy in to the business model. So we give them the servers, we give them training, and we also provide them with support. And then we could then set up a locally-owned wireless center for the villages. Now, obviously, the cost within that network is free, and we'll go provide good services. One of the main things that the profits go back into getting more equipment, getting the villages connected. Now, the beauty of that is decentralized. The communication network is not owned by the government or by the private companies. And obviously, you all know what happened in Egypt. Effectively, there's a Q-switch. With this system, you can't have a Q-switch. It's owned by the people, and it's controlled by the people. And finally, if you want to go on to the website, villagetelco.org. It's an excellent place for information. And also join the Google group. So if you're a hacker, as most of you are, you can join in. We have the A2 Billion author that monitored the group. We have Steve Song, who's a shuttle world fellow. And we also have Electra. She is the author of The Batman Demon. So if you guys are familiar with OpenWRT, you would know that she's a very, very keen developer on mesh technology. And obviously, you can go to wireless dialogue on the website or email me for my information. And that's it. Any questions? The question is, how do you measure quality on the network? One of the main focus is talking about quality of service. One of the main focus of this was that to be able to use this, obviously, you're sharing a boundary between voice and data, exactly. But if you think about it, although the wireless infrastructure here in Europe is actually quite congested, in most of those regions, there is no wireless network. So effectively, you are going to have a single link. Obviously, again, with the benefits of using mesh technologies, that the more nodes you have, the more routing capabilities you're going to have. So the more infrastructure you build, so 1,500 nodes, the better a network becomes. And obviously, by building on a single routing, rather than congesting the space with multiple routers, every router is actually utilizing the network properly. And obviously, there's actually a new routing algorithm made by Electra called Batman Advanced. Now, that's a layer 2 routing algorithm. So it effectively accesses the switch. So it means you can have multiple servers on top of that. Now, basically, it means that you have a huge wireless LAN network that's totally invisible to the user. Thank you. The hiding nodes. The question is, how do you resolve the problem with the hiding nodes? I've never come across that problem myself. I think what you find is that it depends on the routing algorithm that you use. If you're going to build a fixed, because most of the time, we're going to be the fixed wireless. So if you look at OLSR and if you look at Batman, you have to actually look at what you're building, how you're going to design your network. So it actually comes down to how you design your wireless network initially. So you say, well, if I'm going to have nodes moving around constantly, now, which routing algorithm do I use? And which one is my preference? So I always find that the weakest part of the point is which algorithm I write an algorithm to use initially. And that mostly solves your problem. The question is, do I have relations with the server-backphone project? Actually, the server-backphone project is a fork of this village telco project. Paul Gardner Stevens, he actually approached the group and said that he wants to explore and extend it. Obviously, being opposed to what it is, it was welcome. He's actually putting, for those that I don't know, the backphone project actually utilizes embedded access inside an Android phone. So effectively, an Android phone acts as its own telco router and access point. So you can make calls within the village telco network without incurring any cost to the operator. That's just for those that I don't know. So if you repeat the question, the question is, I'm trying to rephrase that into a short sentence. The question is, in terms of, in reality, how do you ensure that do we actually build decentralized and totally unconnected networks which are not connected to the rest of the world? In our experience, we find that most people want to connect to the world. And the biggest problem, which is obviously we can't solve with the village telco project, is that the ISP licenses restricts you from actually gateway and outside. So governments have actually licensed the spectrum specifically for the big telcos. So as a small local IP, you have to basically pay maybe $35,000 to $40,000 per year in license costs. Now, if you approach the Belgium operators in terms of actually having an ISP, it's virtually at no cost. So you can see the imbalance there. So that's a major hurdle for us at the moment. Hopefully, with the coming of Firebar Broadband across the west coast of Africa, main one, and a few others, we should be able to alleviate that problem. But again, this is not part of this talk, but we're trying to actually build locally owned and locally provided services. So locally designed websites, local servers, everything locally derived for the local languages. So hopefully, that problem should be alleviated. Yes, please? The question is, how do I configure IP address to the mesh node? It's actually very, very basic. The key point is that within five clicks, you should be able to configure a mesh potato. So there's actually a guide. If you go to the VillageStalker.org, I don't want to go into too much detail here. Effectively, there's a special facility. You can actually FTP-ing into using a crossover cable. Sorry? No, no, at the moment it's manually, yes. Yes, please? The question is, what do our franchise charge the customers? Now, this is where we usually call the tiered model. So we provide services to the enterprises, to the huge companies, and then in turn, subsidize the price of the equipment going in to the small businesses and to the small entrepreneurs. So obviously, it's done on a local basis. You're not going to charge the same price as Africa that you would charge somewhere in Niger, for instance, because the income difference are just too great. So we do it by case-by-case basis. And obviously, if big enterprises that actually want to locate within that region, if they take our services, part of the agreement is that within that mesh network that they create for themselves, they could be able to connect and use a downlink and then obviously reduce the cost to the small village entrepreneur. Yes, please? The question is, have we designed the hardware for that? For basically the limitations that you have of wireless networking. The answer is the only closed-property hardware inside the village circle is actually an atherostrip. Unfortunately, it costs about a quarter of a million dollars to actually get the code within that to actually then modify the blobs, as you say, to actually make the route more efficient. Now, obviously, we're moving on to software-defined radios. I've just saw a talk yesterday about FPGAs. Hopefully, once that barrier has been opened up, because at the moment, it's just literally impossible for us to design your own radio chip, but everything else is open-sourced. But the problem is now, to answer your question, is we still have the limitation. Yes. Any more questions? Okay. Thank you.