 Okay, let's begin with the actual content where we have a mystery talk now. There happened something incredible. Our next speaker, Peter, is working at Telekom and he was here on time. Imagine that. The technician is not just informed. He is here too. And he will give the information to you. I hope that... Where does it come from that I only have an edge? Yeah, I hope that... He could tell me why I always only had an edge. But no, we are talking today about the things that we don't see in the phone. We see two or three reception bars and he will... can explain us why these bars are not that relevant. Thanks for this introduction and for the invitation. My talk is about mobile telephony network, cell phone network, especially about the interface through the air and about the physical layer of this air and especially the uplink. That is the way from the mobile unit to the buy station. I work in this range for about 30 years. I know about everything except the A-net from back then. And so I've seen a lot and while the uplink is a little hidden genius, but what's the uplink at all? Uplink is the way from the mobile station to the buy station. We also have the downlink. The downlink is from the buy station to the mobile unit, to your mobile phone exemption, for example. And we have the indicator on the phone. It tells us how good the downlink is. It's these bars, two, three or four, even more bars. But now your device needs to send something back to the buy station. This is called the uplink. And the frequency range is used in Europe. Usually these are different frequencies. So the buy station sends in an upper frequency range. In this example, we see GSM 900. This is the upper frequency range. You see the buy station. And in the lower frequency range, we have a lower band and an upper band. It's also called uplink and downlink these days. They're your device, the device the phone sends back. With GSM, we have corresponding channels. They belong together. And other techniques with other technologies, they might be different, they might be not that fixed. In GSM 900, channel one, the red channels, they belong together. They are linked. The channels with the same colors here are linked together. We have even more frequency ranges, of course. You can find them at the Bundesnetzagentur's website, or at the Future Frequencies for 5G, or not yet be found there, but all currently valid frequencies. These are quite a lot. 700 megahertz, 800 megahertz. That's the LTE frequency range, 4G frequency range. The 900 megahertz range, where the legacy GSM is running on it. We have band 32 at 1.5 gigahertz. This has now uplink. 1.5 is downlink-only band. The reason for this is not because the network operators like to just do that. But imagine a smartphone. It does not only telephony and data. It also can receive GPS and other things. And if I would send in this 1.5 gigahertz in such a small device, this would interfere with the integrated GPS receiver. This is about 1.4. So it makes no sense for your device to send something there. Right. And so this frequency is only used if we have another connection together on another frequency band. And the downlink rates can be increased by using this band. We have 1.8 gigahertz, 2 gigahertz. 2 gigahertz is primarily used for UMTS, for the 3G. And we have 2.6 and 3.5 gigahertz. Everyone who is interested, look at the website of Bundesnet Sackentour, the Federal Communication Agency. We have a bit of a problem. Mobile cell phone networks are not broadcast. We cannot send the same thing to everyone. It's not like radio or TV. We need to take care that we only send as much power, so that the mobile device may also send back. We can't just send 5 kilowatts, because then we wouldn't never hear the mobile phones again. For instance, in GSM, you can calculate very well. A channel usually sends about 10 watts. We have an attenuation of the cables. And the device receives about minus 100 dBm. But it cannot send back with 10 watts. Maybe it sends about 1 watt. And this difference needs to be compensated somehow. I need to make that symmetric again. And we can do this with an increased receiver sensitivity. I think not everybody knows what DBM is about. But we technicians, we talk about decibel. We like to talk in DBM, because we don't only deal with easily understandable powers, like 10 watts or 100 watts, but also with extremely low powers in the receiver. And the receiver receives only very low power. And if we start doing femto watts or atto watts, then nobody really knows about the dimensions anymore. And that's why we really like to talk about DBM. The M means we have a power that is relative to 1 milliwatt. So I need a laser pointer here. 1 milliwatt is 0 dBm. 10 times is 10 dBm. And 10 milliwatts 100 times that is 100 milliwatts and 1,000 times. And the 10,000 times, 100,000 times of 1 milliwatt. And that also works into the negative numbers, a 10th and 1,000 and so on. And the DBM is very important, is a power. And if I have 200 watts on a resistor, that really creates some heat. And if I really use just one femto watt, it gets a little, tiny little bit of warmth. But it is some power, and the dB is always relative. So for example, I have a built-in amplifier that has 3 dB amplification that makes 1 watt into 2 watts or 10 watts to 20 watts if it's still capable of that. And it also works in the negative way for attenuation. And so it may attenuate by half. So there are bigger, bigger jumps. There are 50 decibels or 100 decibels. And we will use those values when we look about the radiation field. And it gets a little bit more complicated here. We have the base station that has 10 watts at the end of the cabinet. And then there's attenuation in the wire. It depends if it's remote or if it's somewhere else. We think about two decibels maybe. And then we have antenna gain. The antenna gain means I have then an effective radiated power, so ERP. That is higher than the base station power. But the power isn't really there. It's as if you use a torch light and you focus the beam. And then it seems to be very bright. It's just an equivalent value. So an antenna gain, for example, it looks for the receiver, looks like 400 watts, as if it were an isotropic radiator, which it isn't. But the antenna really has two decibels wire attenuation. And there's less than 10 watts real power. Then I have an attenuation on the way. There's maybe 156 decibels on the way through the air. And that's maybe 1,000 kilometers. It's not possible with GSM, because after 30 kilometers there we have time delays. And the mobile device sends back with 156 decibels attenuation on the way with the same antenna gain, also through the same cable to the base station. And it ends up there with maybe 1 tenth of a power. If it sends with 1 tenth of a power, then the signal is also about a factor 10 smaller on the base station. The base station can receive that, because the base station has electric power and we can build more complicated receivers than in a smartphone. And the smartphone receiver has to be small and energy efficient, and we don't have that restriction on the base station. And on the base station, we have since GSM, we've always had two antennas at least. And then we have RX diversity. So we receive the mobile device with two antennas. And so I can compensate for different running times and through mixers and SDRs and post-processing. And what do I do with the two signals that I have received from the mobile device that I have received with my two antennas? There are, roughly speaking, there are two ways. I can either, in the post-processing, I can have a switch that switches to the path where the mobile device can be heard, can be received with the best quality, and someone has to switch that toggle switch via software. And the advantage is, if one of the paths has a high noise, for example, has some other interference, then I can switch it off. It has a high bit error rate. And I can see that. And I can switch to the other. But if I use only one, then the total reception sensitivity is lower. And there's the other variant. So I can add the two antenna signals and use them together. So the total sensitivity is higher. But if one of the antennas has a problem or has some interference and the other is clean, then the entire cell doesn't work anymore. So it's a philosophical decision. One technician does it like that. And at other places, it's done the other way. So the uplink has some other details. Some tricks, if I may say so. So managing the uplink is easier than the downlink. And in the uplink, it's easy. The base station says I have data. And then the mobile station says I'm here. And then it has to tell the device where to send it to. But in the uplink, there's always the problem that the mobile device wants to send data, wants to transmit data, wants to initiate a call. And the base station has to see where our resource is free in the uplink and has to select them and has to tell the resources to the terminal device and says the mobile device use these resources, these resources, these frequencies for these durations. And I tell you at which times. And if the device is finished earlier or it has more data, it has to tell the base station. So the handling of the data in the uplink resource, handling of the uplink resources, is much more complicated than for the downlink. Good. OK, for a change now, before we come to interference, some antennas from the inside, let's see it from the outside. There are different antenna types that are used in mobile radio. Some of these broadcast antennas, omnidirectional antennas, there were two reception antennas. And there was no place on the top for the antenna or these omnidirectional antennas, you can't find them anymore these days because they have not a very good coverage. And so you use field antennas these days. And there in this one of the case, there are at least two antennas in the casing that can be used for different frequencies. And these two antenna types or these two antennas, they have a crossover polarization and you have diversity not with antennas at different positions, but you use diversity for the polarization. If there's something on one polarization level, then you don't need to listen on the other one. And from the inside is what it looks like. Looks like on the outside you normally only see the case and there's a lot of stuff in it, antenna elements and cables on the backside. And this is a multi-band, a dual-band antenna. And these boxes that you can see are phase shifters and because such an antenna isn't supposed only to cover the field, but as an operator, I maybe want to lower the power, depending on the demand, such as if I shift the headlights on the car and move them up and down. And for example, if I take my car up to the roof and if it's dark and I want to illuminate it and light is much easier to focus, but it's quite similar. So this is an antenna from the inside. There's a lot of stuff in it. And the most modern antennas that we have have lots of connectors for all the different frequency bands and you really want to, on the mast, you really want to have as little as possible mass on the mast. So it's best to have an antenna with multiple frequency bands. Well, if now we look at the cell sizes, what we see on the mobile device is the downlink cell size. You have a different signal strength. You see them with the reception bars. The cells, mobile cells, they tell the device if the received signal is lower than a certain value. You see the values written here for different technologies. Then please leave the cell. So this is basically the end of the downlink. This may also be earlier when the device doesn't have this sensitivity or if nothing can be decoded anymore. The cell size of the uplink needs too much of this because it must not be smaller. If it would be smaller, then it doesn't make sense to have a reception channel because I cannot send nothing back. And while the antenna, the cell phone tower, it may receive different noise and, et cetera, different interferences, which maybe my mobile phone does not see because it's not that high. And we can calculate the value of the downlink versus this uplink. And usually it's not exactly the same. So what happens if I want to make a call? I want to establish a call. First, I make a mobile-originated call. The mobile device accesses the base station. It tells I'd like to make a call. But somehow the receiver is deaf. It does not receive anything. Then the mobile device tries again, et cetera. And then if there is still no response, mobile devices, they switch another cell. As a customer, you usually don't see this. It takes a bit longer to establish a call. But you don't see that. It's all done automatically. So there is different rescue possibilities for outgoing connections, for incoming connections. If my mobile device is in a certain cell, then it's known to the network that my device is in this certain cell. And the network does a paging. It tells, while there's a call for you, and then the mobile device can say, yes, I'm here. And if now this does not arrive, then the device station thinks, oh, that device, I couldn't reach it. But the customer, the phone, does not ring. That's the worst case. The client has, well, if you want to establish a call, if you want to make an outgoing call, then you can go to the window, et cetera. But when you don't receive a call, that's basically the worst case, because you don't even see that there was a problem. Broken antenna can also have a negative effect. Then the uplink and the downlink cell, the range, is not exactly the same. For example, at this here, the customer has a good RSSI, but at the left side, you have no signal indicator, but the base station would hear you very well. And while this may be a broken cable at the antenna, which can produce such effects, there are influences in the uplink. Man-made or somewhere from the radio field, we divide between external sources. As a network operator, I can't really do anything about that, except looking for them. And we also have internal sources of interferences. It's in the base station itself. Maybe something is broken. Let's begin with the external. As a network operator, how can I detect it? We have different. Yeah, it's difficult, because the customers also generate traffic and generate radio waves. So we need some triggers. There's, for instance, if on one antenna path, there is more amplitude than on the other one, then it's an anomaly you can detect. They should have more or less the same amplitude at all the time. Or we can do performance tests. One can do performance tests. Are there many call setup errors? Is the, how about the modulation schemes, are they more or less do they match together? And it's not that easy. It's quite difficult. But I don't have much. But I don't have much other possibilities to see if there's a, yeah, never. So if I have an interference, there's a trigger for the field services dispatch to look at the problem. And they can do that with a spectrum analyzer or some other tools, which I will show in a moment. And they can then remove the problem or you can task the Federal Network Agency with that. They are going to look for the interference or you're going to hunt the interference yourself. External interference are, for example, broken decked phones. I'll show you some examples in a moment. You can measure at the site. And if you have received a trigger from as a field service technician, I go there and some manufacturer have this spectrogram. And with directly at the station, I can look at the uplink spectrum. And you can see the white curve. The white is bigger than the green. So some of the antennas is perhaps broken there. And you can look at it in more detail and change the power and turn off the power and turn it back on. And you can see that this is modulation problem. In the end there, there's a little 800 megahertz headphones, which doesn't belong there. You can see things like that. And if I can't see anything like that, then I have to get access to this interface. And if the device has a test port and you can connect a spectrum analyzer there, and usually I have to really climb up the tower with a spectrum analyzer and there's no power. And if it rains, you get wet and so on. And if I really have a device without such a test port, you can put this device into the antenna cable and it requires external power and things like that. So and look for interference for three days. That's it's really not very comfortable there. And another method is for technology, for supported technologies. They have this RF over fiber. And I have between the remote head of the radio and the system module on the ground, there's a fiber optic cable. And on this fiber, there's not IP, but there's a special protocol of a frequency band. And it is transformed with a fast-food transformation from the spatial domain to the frequency domain. And if I connect to the fiber optic cable and then there's an optic splitter there, and you can get some information of the information that goes from the ground up to the tower and a measuring device and an instrument can show me the things, that's what it looks like. And I can only see the frequency which I have selected from this data stream. And that's really the most comfortable thing that you can do. But because you have to put the splitter in there, you have to disable that cell for a moment. It's not nice for the customer, but afterwards it will run fine again. And there are some more sources from the real life. There's a GSM 900 frequency. It's a waterfall diagram that you can see here. And this waterfall diagram will show a spectrum analysis on the bottom. And it shows it as a history plot. Blue is deep water and low noise and yellow and red. There's a high level, and you can see very well here. And this kind of analysis, without waterfall diagram, it's very hard to diagnose an interference. Very few look exactly like this as a continuous carrier with a continuous video signal. But this was a video baby phone built for the American market, really. And they have very powerful transmitters. They have a range of 30 kilometers. And my argument for turning that off to the customer is everyone who has a device like that can also really see what happens. Everybody can see what happens in the bedroom there with an infrared light. So it's a good idea not to use that, really. Another interference are active antennas, active DVBT antennas. And if they're really in the sun and they get hot, they create a very high-power carrier walking through the frequency band. And depending on where the sun is in the mobile radio frequency, then there are some decked base stations. Many people have decked base stations and decked handhelds and use different base stations and use the decked base station only as a charging station. And sometimes they think they are in a different region, in a different European region where they were produced. And sometimes they transmit them in the UMTS frequency range. And Vodafone had a lot of trouble with that. And we, not so much, but you can see them really well. They have a very characteristic fingerprint. And if you look at this, you can see there's a burst every 50 or 100 milliseconds. And you can really see that what this is. You don't have to go up to the tower. You can just search for the decked phone. They have really high power. And you can easily find them. But what's more complicated is Wi-Fi access points. And they think they are in a different country, for example. And they don't make such a beautiful signal. But you can see, for example, you can see pulses. And you can see that in the Vodafone diagram it looks like a staircase. And you can see that doesn't really belong here. And if you really look at the, you change the band and set it to zero. And you can see that every 50 or 100 milliseconds it makes a pulse. And then you can see that it's probably a Wi-Fi device. Or what's another thing is a forgotten terrestrial TV antenna amplifier. They make a big carrier signal with a lot of noise in it. And because they just have a free-floating oscillator, they walk through the frequency. And maybe they are on this frequency today and tomorrow on another frequency. But there are quite a lot of those people. I'm really surprised, if I tell them, I have to go up there and say, oh, yeah, that old thing we forgot about that. And the last thing that we have is wireless headphones. They are just above the LTE 800. And if they go wrong, they make a very nasty spectrum which interferes with the uplink of the base station. And this spectrum really reduces voice quality in LTE. But the uplink, voice quality and your partner then hears you with poor quality. And there are some cameras with some Wi-Fi or other radio connection. And they make lots of noise. And the tower, the base station tower is in line of sight. And it really can see that camera and receive the signal and just disable 20 cameras on a gas station. Then there are some test transmitters from a subcontractor. They can do test measurements. But if you operate them in the uplink area and transmit there, there's also the base station antenna. And then they overload the receiver. And there are several things that happen. And there's now another critical chapter here. And there's unlicensed repeaters or not uncertified repeaters. In Germany, the jurisdiction is like that. So only the network operator may transmit. And the transmission also means amplification in Germany. And if you have a receiver base station and amplified, so in Germany, they are really forbidden. They are good and bad ones. And the bad ones you can buy on eBay. They don't cost a lot. And they are bad because they make lots of noise. And if someone has a repeater like that and operates it and doesn't know what he's doing, then maybe we knock on your door. And it transmits on other bands. And people just align it with the base station and really interfere a lot with that. And there are professional installations of these repeater systems. There are two antennas. And you could really say passive repeaters. There's no problem with that. You can do that. I did some calculation. If I have a relatively good reception outside and want to use it also in the basement, but you can't really go too far away from your antenna in the basement. And you can see here in the appendix, there are some calculations about the field strength. And you can see and calculate what the field strength is. And then add the field strength and you get some of these values. Another nasty topic is our jammers, handy killers. Usually this one, they should jam the downlink signals, such that the devices new to it don't find network anymore. But also these are not that expensive usually. So they are not that well-filtered and they also jam the uplink. And the jammer can be quite powerful. And if you jam in the uplink, then the network operator will see that. This is not nice for the operator of the jammer because it's highly illegal and it costs about 1,500 euros fine and maybe even more prosecution. They are easy to find, easy to buy. Well, take care if you want to buy one or don't buy one at all. Usually they're not allowed. But some of them, it's written that they are not allowed. I absolutely don't recommend to use such a thing. Don't do it. Just don't do it. While it doesn't only hurt you, it also jams everyone around you and that's not good. If you want to try such a thing down in your concrete seller, then it's OK. But don't do it somewhere else. How do we find? How do we hunt and interference? Yeah. I should have more or less an idea where the jammer or the interference could be. It's a good idea to go up at the antenna and look around the area. Me personally, I drive with a non-directional receiver antenna. It's a little one. It receives not only horizontal, it's omnidirectional release from everywhere. And when I find interference source with the small antenna, then I take a directional antenna. And then I will probably find it. The Bundesnetzaggen Tour does this with more professional means. But it also works with my more simpler means. If you want to do something at home, professional spectrum analyzer is expensive. But if you have an SDR with this RTL chip, well, this is the smallest spectrum analyzer that I have at home. It's also not that well quality, but it works between 70 megahertz and 1.4 gigahertz. Because these USB sticks are built for DVBT, which only operates in these frequency range. Also have a look at the talk of Federica. Gee, it's about this today. I use different tools, for example, SDR-SHARP or RAPSMI, the Hocker-RF radio, et cetera. Probably no doubt about it. Another topic is passive intermodulation, PIM. It occupies me since the last 25 years. It's inside the antenna. It happens because these days we almost always only have one antenna which sends and receives inside the same antenna. We have quite a high power downlink. And with the same antenna and the same cable, we receive very weak TX signal. There may be up to 150 decibel weaker. This is a number of 15 zeros. It's quite a lot different. And if there can be corona effects or interferences inside the cables, for example, if the cable is broken, and then it may happen that in the uplink there are interferences from your proper downlink, you can calculate them. And the broad noise is a corona effect, but still lacks somehow the physical explanation. A mixing can be explained quite simply. You have a nonlinear resistor, for example, oxides on the plug, corrugated plug, then it bends this, it modulates new waves, basically, and makes mixing products. And you can calculate them. We have a third order, fifth order, et cetera. Then they can also overlap. If our LTE signal is 20 megahertz broad, then these mixing products are also quite broad. This is a case I measured at GSM. It's a very beautiful case. At the right we have a pre-traffic channel. The other one's our broadcast channel. If you want, we don't see this one. It was in the by-station. It wasn't in the filter. And we see here the 9th and the 11th and the 13th intermodulations of these mixings, of these modulations. This works well with GSM and with broadband signals. It's not that easy anymore to see them. The corona effect, this broadband noise, I observed that this happens when I have sharp edges in the copper, in the antenna, or in the plaques, or fire inhibitor in the plaques. Things happen. There seems to happen some kind of ionizations. The electrons go away from the atom and come back. And this creates a non-ossillations. It will never be... It's probably a core resonance. I'm not in this field. Maybe someone can explain me. And this is how it looks like. It can look like we have strong noise, strong broadband noise. Different frequency bands, there are also modulations. For example, in the telecom band, at new 100 MHz, we see this IM7 here. That's okay. It's a bit weaker than IM3. But if I look at the multi-provider antenna, this looks different. We have frequencies from different providers which can mix. And the telecom receives the interferences of other providers. It's a bit of a risk with multi-provider antennas. How can we measure the uplink-downlink symmetry with a smartphone? Maybe do it at home, or for a simple way? I tried it for three different services or things. For GSM, UMTS, LTE. I first need to do a dedicated call. Now with a data connection, we need a call. It makes a proper, more beautiful call. And when I've done that voice call in GSM, for instance, with a monitor in the Samsung S7, I see these two levels in the RX and TAG level. And I can calculate a ratio of these and then I see if it might be okay or not. It also has some other parameters of the network operators. But if most of them use the same parameters, the lower my reception is, then the device needs to send more powerful back. This also works with UMTS. We use the common pilot channel. That's the pilot channel where the signal indicator is based on. And with UMTS, this Coralight 1-2-1, you can't change that. It's a network operator. It also works with LTE. I have some more stuff about LTE. We assume we have a reference signal. We can measure it with LTE. It's about 15 dBm. It's not much, and we have much of them. So the reception is quite low. But if I correlate this with the transmit power, I can see if it works more or less with this table. I can calculate that. If it correlates well or not. So that was my talk about the uplink. If you have problems and interference that might happen, and how we can look at it with a DIY means. Thank you for listening. Time for questions. Thank you very much, Peter. Are there any questions? Thank you, Peter. Are there other questions? Let's begin with one. Let's say I found out about, for example, this app, or this FIST app, or this analysis you just saw. That's not true. How do I get it right? First-level support, I usually want to film you. With the adrenaline rush. Report this. Yeah, of course, they think it's normal customers. SIM card is broken. All the devices are turned on, things like that. Yeah, you're laughing, but yeah. If you tell them enough information, and I really found this and that, they probably pass it on to field service. Did I understand it right that the network tells the phone from one, the quality is not sufficient anymore, and to leave the cell? Yeah, almost. There are several things with mobile radio. But what I told you before is that the cell says, if you get not at least minus 160b, that's all right. If you have minus 107, then please go away and leave the cell. That's a parameter that's fixed. The values are much too low, in my opinion, but it's European recommendations by the system manufacturers. Does it mean that the network says how many bars are shown at the telephone? No, that's the thing the telephone does itself. They create it from the ASU value, and every handset manufacturer makes their own thing. Sometimes five bars are sometimes a really relatively low level of minus 80 dB. Old base stations have transmitted with 50 watts, at least near the borders. And of course, you have to be careful that the devices don't get to the border of the cell. So they change these threshold values up to minus 90 dB, and maybe the customer had two bars, and then it went straight to zero. A question from the internet. Have you seen IMSI catchers, or can they be detected? I haven't found any. No. I've found mobile phone jammers, but no catchers. They make no radiation in the uplink. You can't find them. What operations do you have to shut down the jammers or interferences? I don't have any authority like that. I can ring the bell and ask them politely maybe in your attic there's an amplifier and they can let me enter or not. Most people will let me enter. They are usually very friendly. Surprisingly, yes. Thank you very much to all the people I have visited. Most people just don't know what happens there and that they have a thing like that and that it might cause interference. And where I have problems with entering and then I call the federal network agency and usually they'll be there within a few hours, maybe even with the police. How about a self-built antenna? It's quite common to do it with Wi-Fi and Wi-Fi. Does this create interference? No, I haven't heard anything about that. It's usually the hardware. I don't know. I don't know. Great. No, I haven't heard anything about that. It's usually the hardware inside the device that goes crazy and transmits. Wi-Fi is only 100 milliwatts effective radiated power. But there hasn't been any interference by special antennas or something like that. Another short question. We saw pictures about 512 antennas. There might be more antennas in it. How many active connections can you use? How many connections can be done with that? Is one of them used multiple times? So how must I imagine it? Are there thousands of end devices? So how many connections can I do that? A good cell can do up to 100 continuous data connections and very, very many passive terminal devices. And we all sit here and make no traffic. They only have just one dataset, one record there. And UMTS, for example, can do on the 5 MHz channel. In theory, it's about 120 calls. So 105 MHz, 100 calls. And you can imagine on a broader cell, on a wider band, that's more. If you want to do high quality voice, you need a little bit more data and then you have fewer channels. But you can do that adaptively as network operator. If you have many calls on LTE, because data is really the problem, then you can reduce the quality, the voice call quality that goes automatically to make room for the data. But telephony has become quite unimportant and data is much more important. And data is a much bigger problem than if voice goes away. I ask number three about the police radio. Does it have similar difficulties on the technical side? Does it do, for example, police officers have more... So police radio is what... How about the receiver sensitivity? No power in the downlink. That doesn't increase the communication quality. So in the end, some thresholds should be changed or reduced. I have no idea about police radio. They're in a completely different frequency band. There may be some corona effect problems, four or five hundred megahertz, unfortunately. They don't interfere with us. Do we also have a little problem? What's that? That the decline in the cell... Although it had a stronger cell, it would have. Yes, that happens with LTE. If a phone remains in an 1800 cell, although another cell would be available. But we have to tweak the parameters there a little bit for all operators. There are some mechanisms specified in the 3GBP and for the terminal device, which then preferably parks on the 800 megahertz and there's a data connection and goes to the window and then goes up to 1800 megahertz. But that requires a wide area module systems by all operators. And yeah, that is going on. Thank you for this long time. Until now I thought that the complexity of the complexity of the technology is going down. That is, in problems or in business names, Jovai or Ericsson goes by himself and parametriates everything. And I don't give it to the providers personally. Can you say something about it? Field service, look from the top, we change line replaceable units. And many of the operators in Europe, maybe all except the telecom, except Deutsche Telecom have outsourced that service. It's not the best solution. I think Deutsche Telecom hasn't done that so far and won't do that. And we can react more flexible to interference. So I think it's an advantage if you do that in-house and can really react to problems like this. And if you just change the LRU, it's not solved. About these cell phone killers, how fast do you or the Bundesratsier again to react to such thing? Typical technicians answer is it depends. If there's no complaints and there are no interference on the uplink, on the base station, then nobody will notice. But it happens that customers complain and if it's more than one and they say, well, I think it worked yesterday and now it's gone. So I think there's a mobile phone killer, then it's quite fast. In how far was it in your experience, or other communication technologies that shouldn't work, but also work very frequently? Is it the problem with things which aren't radio related at all, for instance, power line communications, et cetera? Is this a problem? Such cable networks make lots of problems, but there's not a lot. It's not that cable Germany has the upper frequencies and we see them in the uplink and we have a big problem with that. If the systems are not built according to regulations, for example, there was a big line of houses that made these problems. They really radiated in 800, 900 megahertz and the installation company always connected the core conductor and not the shielding. Thank you very much. And that concludes the presentation of the hidden side of mobile radio.