 Thank you very much Well, I've the first slide actually you can see a bequat antenna and Yeah, you can see the scale It's actually quite large Because we are testing Wi-Fi mesh networking in Berlin into 62 centimeter band Who understands wave length? Please raise your arms Okay, well, it's quite quite a good audience excellent Well in front you can see one of the prototypes that I've built But let's go into politics first What you can see here on the slides shows the 2.4 gigahertz band It's a trash band or a garbage band as they call it We have three non-overlapping channels if we use Wi-Fi according to the 802.11 B standard, which is now a legacy But because We had so many devices that were following the US regulations of the FCC Actually many administrators of their networks. They only use channels 1 to 11 Because that then FCC gear can be switched to channel 13, which is legal in Europe under the ETSI But these days that's a proposal I want to make before we start if you Happen to configure a wireless network in a 2.4 gigahertz band. You should use channel 1 channel 5 Channel 9 and channel 13 Because these are channels that don't overlap if you a turn if you use a 22 to 11 n or g so Thank you. Thank you because this way we can have 33 percent More benefit that we can all share. I think that's a decent proposal to make But even though we have all these limitations and all these legacies It's a very important infrastructure that we use every day Like any reasonable media use today. It doesn't work with either ethernet or Wi-Fi You can only have frugal use at least in Germany where wireless broadband is very expensive and always capped Yeah without Wi-Fi, we would be I don't know where we would be and these days The big mobile operators they do something they call offloading For example, if you're a customer in Germany and you're subscribed to a German tennis com you have a DSL line You might actually experience that there is an open access point called telecom Which is used in use by telecom users. They're sharing your internet bandwidth Luckily, you get more than you pay for so in order to cater for that and By using this unlicensed band that is already overcrowded the EU has estimated that is that is mobile offloading saves 84 Billion euros each year. That's the figure for 2015, which is of course an estimation But something has happened You might have noticed that television in Europe and in many other countries is digital now Here I have a picture. I took that from the new American Foundation. They're talking about spectral efficiency in the in the EU we have we used to have the analog PAL Television color standard which uses eight megahertz wide channels these days thanks to the spectral efficiency that has increased I'm because we have Digitalized the television and we all paid for that because we had to replace our receivers We can accommodate six channels in one PAL channel. So in eight megahertz of bandwidth. We can accommodate Six television channels or four television channels in a couple of radio channels so a large Share of the spectrum has not been in use for a couple of years What you can see here in this slide the dark gray areas There is Four bands here in the US marked as used for broadcast television Actually, the situation in Europe is there is free because we don't have this gap where it says wireless medical telemetry at least not to my knowledge and So today we only need one sixth of the bandwidth to have the same amount of programs So we can reuse that for something that fits better into our time Instead of simple broadcast what I've just shown you our is the spectrum the whole spectrum below 1 gigahertz and This spectrum has very interesting properties if we look at the very high frequency spectrum VHF and UHF The wave length is in DC meters And wave length it matters. That's why I have this slide with this big wave What will happen if this big really big wave hits the surfer? Not much. I guess it's mildly affected But what happens if a wave of the size of a glass of water Hits the surfer. Well, the wave is cancelled out mostly and the same effect we have in radio physics If you have a wave that is several meters long and it hits an object of this size, it's not affected much If a wavelength of this size is hit or is hitting an oil tanker For example, then it's cancelled out or reflected or broken so if we have longer wavelength We can communicate better in case there are objects That are smaller than the wave length Unfortunately in the 2.4 gigahertz band our wavelength is only 12 centimeters and in the 5 gigahertz band It's even 5 centimeters. So very tiny. So any object which is sizable in relationship to the wavelength is an obstacle that Breaks the wave front. I have to explain that a little maybe to some of you Wave length it's the length of the wave that it travels through space While it goes through one cycle Here on the slide There is a wave that cycles Twice in one second. So the wavelength is 150,000 kilometers give or take because in the vacuum and In-air electromagnetic waves Travel at light speed, which is 300,000 kilometers per second Well, our waves that we use for Wi-Fi day go 2.4 billion cycles per second So they only travel 12 centimeters while they go through a full circle and hence we have the wavelength I hope this is quite a clear explanation Well, maybe I shouldn't dive too much into that electromagnetic waves have an electric and a magnetic field and As they travel through space at the speed of light. Well, all those free Properties are shown in this little slide but Let's not get confused by radio physics now at the first place The consequences that our experiment in Berlin uses 62 second centimeter long waves means that also our Antennas tend to be quite large In order to show it I have a little Demonstration here This is a 2.4 gigahertz only directional dipole antenna from a D-link access point It looks very tiny That's because this antenna doesn't contain much air because these days Longer seems to be better So the manufacturers they have longer antennas, but basically there's just air inside if you're lucky The radiating part is a little bit higher in the plastic So it's a little bit higher about the above the access point So maybe you have a little bit better propagation properties On the other hand, this is now omni for the 62 centimeter band that we use for our tests in Berlin so Quite a difference. I must say Here in the picture you see a eight element Co-linear dipole antenna that in theory gives a gain of 8 dbi I'm not sure if that actually is accurate, but it works pretty well So we have a resource That is now unused at least at the moment to a large part but unfortunately this resource is quickly auctioned off and Again, I have to say unfortunately this happened already for two large parts of the band in Germany on the upper in the upper graph You can see the spectrum that was auctioned first in 2010 to the big mobile telephone operators telephonica telecom and Vodafone and At a very high price they paid billions of euros for that and our politicians were celebrating that they generated so much revenue for the state but That's a really silly claim to make because Probably we're going to try to use this services So we will repay them for the investment that they made and they want a quick return of their investment So we have to pay for that. It's like an extra tax. We pay on wireless broadband and the lower picture is the auction that just happened in In June 19th For five billion euros it was auctioned off to the three big mobile operators in Germany for a considerable amount of time so Are we now all laughing all the way to the bank in Germany? Well, if we are bright no We are not going to do that even though Yeah, some people think we are like sheep and Tell us really silly things Yeah, just if you're if you're not from Germany, you're probably not that much interested in those figures I'm not going to read it to you, but just to give you a quick overview. What has been paid for? So what is wrong with that? Now only commercial providers they get access to that resource It's privatized But it used to be public if you consider Analog television broadcast a public good and we're promised that the market would handle would handle our demands but as we have seen in All those community networks There is a big gap between the promise and reality and it's very expensive if you're a user so I Have a different proposal to make But that's not only me. That's also several others. We're actually Submitted the paper to the EU Commission suggesting that we should all have a Dedicated license exempt Wi-Fi band in the UFH range for everyone Even the mobile operators They can also do their mobile offloading there Because it's free and open can be used by everyone. They can use it as well and they don't have to pay for So we would all win there is already TV white space communication going on Using a Wi-Fi like protocol or a substandard of Wi-Fi But it deals with databases Before your device is allowed to go online. It has to request a database So in order to get online you need internet to get started first and The problem is in many areas the database will simply tell you that you are not allowed to use that device So actually the situation in the United States is that people fake The geographic location of their TV white space devices in order to use them illegally The FCC is now demanding that or has proposed that the devices must have GPS built-in So the user is no longer able to fake the geolocation They say since the devices cost $1,000 each The GPS doesn't add much to the cost Of course thousand dollar Sounds a little expensive to me, but the devices that are on the market are mostly prototypes and They're mostly software-defined radios because no manufacturer is making a chip dedicated for that band yet And since the situation is very unclear. I don't know when it actually will come to us Another problem with this use that TV white space devices share a resource is That you could actually disturb the reception of other people that try to watch digital television or analog television for that matter and If this proposal would be successful and we got at least a couple of megahertz We would at least keep something of that private resource for the public. So There is not much on the market and it's very pricey and Since I like to eat my own dog food. I decided to build prototypes for that band You can see the very first batch in this picture. I did it in a way by Frequency converting. It's something that many hams are using for example when the 13 centimeter band was Introduced into a ham radio people used so-called transverters to use their existing gear from the 70 centimeter band or from any other band in order to shift it to the 13 centimeter band and I'm doing the same the same by using multiplicative mixing. I'm going to explain that to you multiplicative mixing means that you have a local oscillator that generates a frequency. It's a transmitter and An input signal is mixed with the local oscillation signal and the output are mixing products and You can see that in the next slide New frequencies are generated from the input frequencies. I Hope it's it's readable I've made an example with some easy to calculate values F1 is one of the frequencies like the oscillator and F2 is the input signal If they are mixed by multiplicative mixing we get two more signals at the output of the multiplicative mixer We get frequency 1 minus frequency 2 and frequency 1 plus frequency 2 That's a very standard technology. Any modern receiver is Working according to that principle because they all use intermediate frequencies For example, if you have a FM radio and you tune that to one hundred point seven megahertz If you have another receiver and you tune that to 90 megahertz You will receive in close proximity to your radio a blank radio signal That's the local oscillator of your FM radio because the intermediate frequency in FM radios is 10.7 megahertz most Commonly because there's filters exactly for that purpose with the bandwidth that matches FM radio so you could actually Receive transmissions from a receiver, which is funny. So in our example that we use in Berlin We do exactly that. I have the figures here. I have this device Which is a standard Wi-Fi device operating on channel 5. That is 2432 megahertz and I mixed its output signal with 1,950 megahertz and as a result I get frequency 1 minus frequency 2 482 megahertz and an unwanted byproduct which is f1 plus f2 But that's in the high 4 gigahertz range and it is filtered out So that's the block diagram of the prototype on the left-hand side. You see the local oscillator The frequency for up and down conversion has to be the same because the input and output channel Needs to match in the receiving and in the transmission path so there's two mixers one for reception path and transmission path and The output is filtered to remove the unwanted frequencies Then in the transmission path. There's a driver again a filter Which also does impotence impotence matching for the PA yet another filter and the RF switch and then the signal goes to an antenna and in the in the receiving path the RF switch of course has to be switched to the receiving path that is done by a GPIO and Then it's filtered amplified filtered and mixed to the 2.4 gigahertz frequency again And in our case it's 482 megahertz that is present at the antenna The circuit that I have designed is very widely configurable. So you can mix it to almost any frequency that is is Sufficient to carry the modulation of the Wi-Fi signal like you could operate it in the 50 megahertz range for example Which is good Because the FCC has come up with new regulations that new devices are only approved by the FCC If you are unable to install new firmware on it or modify the Wi-Fi driver So you always have to be in their specs, but they cannot stop you from doing mixing Like we do it here. Also the approach of the FCC is very anachronistic because well, you got a software defined radio today that can broadcast anywhere between 50 megahertz and 4 gigahertz and You should do some filtering if you amplify it But then you can go to what if you like and the FCC can do nothing about it. Of course, it's not legal Yeah, but well So the actual implementation The blue board that you see at the bottom. It's a Dragino 2 It's a device made by a guy in Shenzhen in small quantities The name Dragino suggests that it's Something like an Arduino and it's quite hackable because it has GPIOs exposed In principle, you can do what I have done with this device also with other wireless routers But then you have to take care for yourself how to get access to the GPIOs, but here they are exposed Via a pin header. So I just designed the PCB that I could just plug in I have a close up so I can explain a little bit more in detail So on the right-hand side You see two antenna cables. That's why I take the TX and RX signal at 2.4 gigahertz from the Dragino The little black stone below that's the mixer chip Very hard to solder for me in the first place. I had to learn quite a lot And then there is the paths that I have explained at rock diagram. I hope I'm not running out of time the result is Well, that's one of the first samples that I've tested This signal that you see here is 802 to 11 B Sending a massive amount of broadcasts. So they are at basic rate which you can see if you're familiar with the modulation scheme at basic rate and Yeah, that's basically the same signal that comes in at the 2.4 gigahertz band It's not perfectly clean because there is some over swing on the right side But I checked it at the 2.4 gigahertz band. The signal looks quite the same So I'm not making messing it up so much You have seen it on the first slide. That's one of the antennas that we use. That's a directional antenna a B quad It's a very nice antenna very easy to make if you want to make a directional antenna for your existing Wi-Fi equipment I would recommend it This is a little bit larger Obviously because the wavelength is longer and I've added this light because Where we assembled it on the church, you can only see it on the backside this is actually the Zwingli church in Frittesheim and Below in the picture you can see the B quad broadcasting north and Since the wavelength is quite long. I didn't bother shooting through the window Which is already occupied by the 2.4 gigahertz gear. So we go straight through the wooden wall Doesn't change much Also our cable is really easy to handle like the coax doesn't have a lot of attenuation attenuation So you can use the cheap RG 58 if you're into radio design, you know that yeah that then into attenuation gets lower as you have lower frequencies and higher wavelength and the result is if You are radio aware and have some experiments, you know about the Fresnel zone That's just a little example of the Fresnel zone if the in the Fresnel zone you have trees and bushes Then they those uf UHF waves They pass through that quite decently. They will still be attenuated But the any into attenuation is far away from the severe attenuation that you experience in a 2.4 gigahertz band and Far far less than the 5 gigahertz band I think you have experienced that before if you try to connect to a 5 gigahertz network outdoors if there is some bushes You don't get far you get much further at 2.4 gigahertz and At 482 megahertz for example in the UHF band the effect is even greater So the range is even better very very nice to for an environment like here to have a broad coverage at a decent bandwidth Giving the same modulation band with the speed of the wireless Network in the 2.4 in the 482 megahertz band is exactly the same That's my Fresnel zone at home On the Half to the right behind all those trees bushes and metal fence there is the church and The signal gets through it gets attenuated is still a great deal But it's far away from the attenuation at 2.4 gigahertz So I like it a lot and we only have the license to use that for a limited amount of time So I hope you could help me lobby a little for that that we get a part of that spectrum for us all Thank you very much. I hope you enjoyed the talk and I hope we have some time for questions Yeah, you have actually two more minutes left So if you have any questions, please line up at the microphones So we can take them to tap them on the stream and Everybody will hear them Shootout, please To what extent do you feel that the the juxtapination at 482? Given that you're proposing an unlicensed band Which is not only going to contain Wi-Fi It's going to end up containing program making for special events radio mics all sorts of stuff That increase sir. Can you say it again because I couldn't quite understand correctly You were talking about a cordless microphones that operate in UHF band. Yes. I What I was talking about was that at 482 you have less sorry at at 482 megahertz. Yes You have less attenuation So you're going to be able to accommodate less users in a given area Yes That's a that's a good point There is a patch It's called ministerial blues Which regulates the transmit power of Wi-Fi? At the moment our Wi-Fi equipment at 2.4 and 5 gigahertz it always broadcasts at the same power level and By default we all use full power There's only a few people that are aware of that and that are limiting the power that they use so Of course, that's a very good suggestion Because the range of the UHF band is far greater It also means that the cellular structure that we use in order to cater for More bandwidth for more people or more devices in a certain area has to be taken care of So my suggestion and we also suggested it in the paper that we submitted to the EU is that Power regulation is mandatory So for example, if you can have a fine bandwidth to another node If you broadcast at a very low transmit level The device will automatically choose that so in order to reduce the interference that you can have with other networks and I also like the idea to use this frequency for Relatively long-range coverage. So not necessarily for very very high bandwidth For example in our prototypes. We can now use 5 megahertz 10 megahertz 20 megahertz and 40 megahertz wide channels and For example, if we decide to use the 5 megahertz wide channels We can have four channels instead of one like we have by default in Wi-Fi and At 25% of the bandwidth, which I think is a fair deal and another very interesting Property of this frequency band is that the polarization of the wave front It doesn't change so much like for example if you have Wi-Fi and you shoot through a tree Then if your antenna is vertically polarized, you can also receive your signal horizontally Polarized because due to the wave front getting broken in the leaves The the polarization is rotating so you don't have a simple linear polarization and That matters because if you have a straight linear polarized vertical Electromagnetic field and you try to receive that with a horizontal antenna, you will have an attenuation of attenuation of 20 decibels and So therefore you could very nicely separate the different wave fronts From each other even though they are in the same frequency so We can actually decide whether we use vertical or horizontal linear polarization or a circular polarization But of course, that's maybe a little too complicated to explain now and we're running out of time. Yeah, sorry But the time is over. We have to shut you up Okay, sorry about that. It was a great talk. Please give a brilliant That's your applause