0 00:00:00,000 --> 00:00:30,000 Dear viewer, these subtitles were generated by a machine via the service Trint and therefore are (very) buggy. If you are capable, please help us to create good quality subtitles: https://c3subtitles.de/talk/524 Thanks! 1 00:00:09,350 --> 00:00:10,350 Thanks for the intro. 2 00:00:12,440 --> 00:00:14,659 First, let me say, when 3 00:00:14,660 --> 00:00:16,879 I submitted a proposal 4 00:00:16,880 --> 00:00:19,189 for to talk here, I didn't expect 5 00:00:19,190 --> 00:00:21,439 to be put into a room with 6 00:00:21,440 --> 00:00:23,359 with such a big audience in front of me. 7 00:00:24,630 --> 00:00:26,179 Uh, yeah. 8 00:00:26,180 --> 00:00:27,439 Anyway, now I'm here. 9 00:00:27,440 --> 00:00:29,030 And first of all, 10 00:00:31,190 --> 00:00:33,709 this is not a talk address to 11 00:00:33,710 --> 00:00:36,019 open up to developers. 12 00:00:36,020 --> 00:00:37,490 So the technical level is 13 00:00:38,750 --> 00:00:40,969 it's really on a more abstract 14 00:00:40,970 --> 00:00:43,039 and almost sometimes philosophical 15 00:00:43,040 --> 00:00:44,040 level here. 16 00:00:44,840 --> 00:00:47,179 But I thought about it and 17 00:00:47,180 --> 00:00:49,399 I really I 18 00:00:49,400 --> 00:00:50,929 made the decision that this is the this 19 00:00:50,930 --> 00:00:52,460 is the most important point I want to 20 00:00:53,750 --> 00:00:56,119 get across here to you 21 00:00:56,120 --> 00:00:57,120 and. 22 00:00:58,970 --> 00:00:59,970 And. 23 00:01:01,930 --> 00:01:04,329 I think I'm preaching to the choir here 24 00:01:04,330 --> 00:01:06,609 in when I say that 25 00:01:06,610 --> 00:01:09,159 we are all interested to 26 00:01:09,160 --> 00:01:11,259 keep developing technologies, 27 00:01:11,260 --> 00:01:12,760 communication technologies. 28 00:01:14,070 --> 00:01:16,349 That allow us to communicate more 29 00:01:16,350 --> 00:01:18,360 directly and more authentically. 30 00:01:20,140 --> 00:01:22,869 So we need more of 31 00:01:22,870 --> 00:01:24,969 the things on the right side and 32 00:01:24,970 --> 00:01:26,919 less dependency on the thing on the 33 00:01:26,920 --> 00:01:29,499 infrastructure. It looks like on the 34 00:01:29,500 --> 00:01:30,500 left side here. 35 00:01:32,060 --> 00:01:34,259 Problem is, for example, 36 00:01:34,260 --> 00:01:36,679 wireless mesh networks where this is very 37 00:01:36,680 --> 00:01:38,689 easy to do or we will see 38 00:01:39,890 --> 00:01:41,989 it's an interference limit 39 00:01:41,990 --> 00:01:44,989 problem that we run into quite fast. 40 00:01:44,990 --> 00:01:45,990 So it's not 41 00:01:47,150 --> 00:01:49,759 the performance of the network is not 42 00:01:49,760 --> 00:01:50,760 dependent on. 43 00:01:54,860 --> 00:01:56,959 Creating distances that are small enough 44 00:01:56,960 --> 00:01:59,269 to get at least some 45 00:01:59,270 --> 00:02:01,489 signal to the receiver, it's more or less 46 00:02:01,490 --> 00:02:04,129 the interdependencies of 47 00:02:04,130 --> 00:02:07,099 multiple transmissions in a shared 48 00:02:07,100 --> 00:02:08,100 common medium. 49 00:02:10,759 --> 00:02:13,009 And actually, every 50 00:02:13,010 --> 00:02:14,089 wireless network. 51 00:02:15,950 --> 00:02:18,169 Eventually becomes an interference 52 00:02:18,170 --> 00:02:19,170 limited network 53 00:02:21,290 --> 00:02:22,819 with a growing node density. 54 00:02:28,420 --> 00:02:30,519 It's even analogous to society itself, 55 00:02:30,520 --> 00:02:32,529 because right now we are starting to live 56 00:02:32,530 --> 00:02:35,050 in a society that is attention limited, 57 00:02:36,930 --> 00:02:39,009 the relation 58 00:02:39,010 --> 00:02:40,330 attention interference is 59 00:02:41,710 --> 00:02:43,060 pretty strong if you look at it, 60 00:02:44,080 --> 00:02:46,059 obviously, the technology that we like to 61 00:02:46,060 --> 00:02:47,349 use to build measures. 62 00:02:48,410 --> 00:02:50,929 We refer to it as I 63 00:02:50,930 --> 00:02:53,299 believe it's 11 W land, or most 64 00:02:53,300 --> 00:02:54,949 people nowadays, globally, say WI, 65 00:02:54,950 --> 00:02:57,199 although it's a trademark term from 66 00:02:57,200 --> 00:02:58,200 the wi fi alliance. 67 00:02:59,670 --> 00:03:02,399 And the link speeds 68 00:03:02,400 --> 00:03:04,889 that these protocol families 69 00:03:04,890 --> 00:03:07,199 offer are sort of growing 70 00:03:08,700 --> 00:03:09,700 like. 71 00:03:10,610 --> 00:03:13,519 Monotonously, it's really 72 00:03:13,520 --> 00:03:15,619 strange to look at the throughput figures 73 00:03:15,620 --> 00:03:17,419 that some vendors now give in the range 74 00:03:17,420 --> 00:03:19,459 of, I don't know, seven gigabits per 75 00:03:19,460 --> 00:03:21,799 second or something like that, although 76 00:03:21,800 --> 00:03:24,079 the hardware itself starts to look really 77 00:03:24,080 --> 00:03:25,080 ridiculous. 78 00:03:27,470 --> 00:03:29,599 But anyway, this is 79 00:03:29,600 --> 00:03:31,819 the most suitable technology that we now 80 00:03:31,820 --> 00:03:32,959 have. 81 00:03:32,960 --> 00:03:35,299 It's it's becoming, thanks 82 00:03:35,300 --> 00:03:37,099 to Moore's Law, extremely cheap, 83 00:03:37,100 --> 00:03:38,659 extremely powerful. 84 00:03:38,660 --> 00:03:40,759 But it needs one thing to work, which is. 85 00:03:43,130 --> 00:03:44,130 Free channels. 86 00:03:45,580 --> 00:03:48,609 And especially when a gear 87 00:03:48,610 --> 00:03:50,739 is directed toward five gigahertz 88 00:03:50,740 --> 00:03:53,019 usage and if you look, for example, 89 00:03:53,020 --> 00:03:55,119 in North America in 90 00:03:55,120 --> 00:03:57,549 order to get these gigabit speeds, 91 00:03:57,550 --> 00:04:00,039 we must use channel with 92 00:04:00,040 --> 00:04:02,439 in the order of 80 or 160 megahertz 93 00:04:02,440 --> 00:04:03,579 even. 94 00:04:03,580 --> 00:04:05,319 And then we have the same situation that 95 00:04:05,320 --> 00:04:07,719 we know from years ago, indicus 96 00:04:07,720 --> 00:04:09,340 and 2.5 gigahertz band. 97 00:04:12,250 --> 00:04:14,499 So I'm a fan of wireless networks. 98 00:04:16,779 --> 00:04:18,939 In the previous talk, we heard 99 00:04:18,940 --> 00:04:21,159 what great things we can do with 100 00:04:21,160 --> 00:04:23,350 meshes and. 101 00:04:25,760 --> 00:04:27,980 This this person. 102 00:04:30,100 --> 00:04:32,679 Cause me to look into ways how to improve 103 00:04:32,680 --> 00:04:34,809 the overall performance of meshes of 104 00:04:34,810 --> 00:04:35,810 true measures. 105 00:04:37,430 --> 00:04:39,889 And there are many spots 106 00:04:39,890 --> 00:04:41,720 you can look at in proving. 107 00:04:42,930 --> 00:04:45,689 This is just one way to 108 00:04:45,690 --> 00:04:47,969 categorize, so 109 00:04:47,970 --> 00:04:50,309 I looked at, right, controlling 110 00:04:50,310 --> 00:04:51,310 and 111 00:04:52,560 --> 00:04:54,239 power controlling. 112 00:04:54,240 --> 00:04:55,769 So these are things that are not really 113 00:04:55,770 --> 00:04:56,669 standardized. 114 00:04:56,670 --> 00:04:58,679 So there is room for improvement, room 115 00:04:58,680 --> 00:05:01,439 for optimization for mesh networks. 116 00:05:01,440 --> 00:05:03,629 I also looked at rather 117 00:05:03,630 --> 00:05:05,729 sort of fancy things like network coding 118 00:05:05,730 --> 00:05:06,730 for some time. 119 00:05:08,800 --> 00:05:10,329 And I also had the chance to look at 120 00:05:10,330 --> 00:05:13,179 congressional control because I 121 00:05:13,180 --> 00:05:14,559 worked for a project that 122 00:05:15,760 --> 00:05:17,949 defines to distribute congressional 123 00:05:17,950 --> 00:05:20,919 control in the vehicular communication 124 00:05:20,920 --> 00:05:21,920 field. So like. 125 00:05:23,130 --> 00:05:25,140 It is tough if you'd like to call this. 126 00:05:27,220 --> 00:05:28,540 And when I did all this. 127 00:05:30,470 --> 00:05:32,419 Over the years, he came to me that there 128 00:05:32,420 --> 00:05:33,470 is one strange thing 129 00:05:34,490 --> 00:05:36,739 missing here, actually, which 130 00:05:36,740 --> 00:05:39,019 is we know that these 131 00:05:39,020 --> 00:05:40,519 networks are interference, limited 132 00:05:40,520 --> 00:05:43,099 mostly, or they all become interference 133 00:05:43,100 --> 00:05:44,100 limited in the end. 134 00:05:45,180 --> 00:05:47,069 But where exactly can I measure this 135 00:05:47,070 --> 00:05:49,019 interference anywhere when I'm 136 00:05:49,020 --> 00:05:51,479 participating in such a network? 137 00:05:51,480 --> 00:05:53,639 So I'm actually making trying to 138 00:05:53,640 --> 00:05:55,739 make the point here that in 139 00:05:55,740 --> 00:05:58,229 order to make the distributed 140 00:05:58,230 --> 00:06:00,869 system, um, robust 141 00:06:00,870 --> 00:06:02,399 enough or. 142 00:06:03,540 --> 00:06:05,789 At least more robust, 143 00:06:05,790 --> 00:06:08,249 um, we need 144 00:06:08,250 --> 00:06:10,319 to implement some 145 00:06:10,320 --> 00:06:11,880 way that lets us. 146 00:06:14,820 --> 00:06:17,039 Directly observe what are 147 00:06:17,040 --> 00:06:19,319 the resources that are 148 00:06:19,320 --> 00:06:21,509 available currently in the channel in 149 00:06:21,510 --> 00:06:23,729 that room, like right now 150 00:06:23,730 --> 00:06:25,889 in here, there are probably lots of 151 00:06:25,890 --> 00:06:28,109 there's the 152 00:06:28,110 --> 00:06:30,299 2.5 gigahertz band is buzzing 153 00:06:30,300 --> 00:06:31,859 because of probe requests. 154 00:06:31,860 --> 00:06:33,929 And I don't know, 155 00:06:33,930 --> 00:06:36,179 uh, the normal, 156 00:06:36,180 --> 00:06:38,249 uh, activity with the access points in 157 00:06:38,250 --> 00:06:39,250 here. 158 00:06:40,820 --> 00:06:43,159 Where where, you know, in our devices. 159 00:06:43,160 --> 00:06:44,160 Can we see that? 160 00:06:45,200 --> 00:06:47,359 And most people think, yes, it's 161 00:06:47,360 --> 00:06:49,159 possible because basically you just 162 00:06:49,160 --> 00:06:50,809 looking at the information that is 163 00:06:50,810 --> 00:06:52,909 decoded by your by the 164 00:06:52,910 --> 00:06:55,189 radio front-end, and then you can sort 165 00:06:55,190 --> 00:06:57,349 of assume 166 00:06:57,350 --> 00:06:59,329 that you have a very good correlation 167 00:06:59,330 --> 00:07:01,339 with. Exactly. This is the activity that 168 00:07:01,340 --> 00:07:03,469 is actually 169 00:07:03,470 --> 00:07:05,119 taking place right now. 170 00:07:05,120 --> 00:07:06,769 But the thing is, because it's a 171 00:07:06,770 --> 00:07:09,109 distributed coordination 172 00:07:09,110 --> 00:07:11,239 that is done when accessing the 173 00:07:11,240 --> 00:07:13,669 medium, actually, it's called 174 00:07:13,670 --> 00:07:15,439 distributed coordination function. 175 00:07:15,440 --> 00:07:16,789 So pretty straightforward in the 176 00:07:16,790 --> 00:07:17,790 standard. 177 00:07:19,430 --> 00:07:20,430 Which is. 178 00:07:21,580 --> 00:07:23,380 More or less the equivalent of 179 00:07:25,030 --> 00:07:27,879 so it's mimicking human communications 180 00:07:27,880 --> 00:07:30,309 in a sense that if you put several polite 181 00:07:30,310 --> 00:07:32,319 people in a room and they want to have a 182 00:07:32,320 --> 00:07:34,479 discussion, they do exactly that. 183 00:07:34,480 --> 00:07:37,059 So they listen, if the medium is free, 184 00:07:37,060 --> 00:07:38,949 if it's not free to back off and wait 185 00:07:38,950 --> 00:07:41,079 until the person 186 00:07:41,080 --> 00:07:42,609 who is currently speaking has finished a 187 00:07:42,610 --> 00:07:44,679 sentence or appears to have finished 188 00:07:44,680 --> 00:07:45,680 the sentence. 189 00:07:46,240 --> 00:07:47,499 So. 190 00:07:47,500 --> 00:07:50,230 And since this is distributed over space, 191 00:07:52,390 --> 00:07:54,759 every receiver has a different perception 192 00:07:56,320 --> 00:07:58,539 of exactly what is the medium 193 00:07:58,540 --> 00:08:00,159 now, active or not? 194 00:08:00,160 --> 00:08:01,750 Am I allowed to speak or not? 195 00:08:02,830 --> 00:08:04,899 And eight to 11 works in a 196 00:08:04,900 --> 00:08:07,089 way where it just defines sort 197 00:08:07,090 --> 00:08:09,159 of a sort of a power threshold 198 00:08:09,160 --> 00:08:11,529 or most 99 percent of the devices 199 00:08:11,530 --> 00:08:12,519 work like this. 200 00:08:12,520 --> 00:08:13,660 There's a power threshold, 201 00:08:15,010 --> 00:08:16,839 for example, in a way, for them it's only 202 00:08:16,840 --> 00:08:18,069 a power threshold. 203 00:08:18,070 --> 00:08:20,679 And so there's a power threshold defined. 204 00:08:20,680 --> 00:08:22,959 And every note 205 00:08:22,960 --> 00:08:25,059 that is waiting to get the 206 00:08:25,060 --> 00:08:27,039 next opportunity to transmit is basically 207 00:08:27,040 --> 00:08:29,109 waiting until the channel power 208 00:08:29,110 --> 00:08:30,429 goes below the threshold. 209 00:08:31,440 --> 00:08:33,658 This is a very simplistic 210 00:08:33,659 --> 00:08:35,819 shortcut now, but essentially 211 00:08:35,820 --> 00:08:37,619 that's the way it works. 212 00:08:37,620 --> 00:08:39,928 So, for example, this should be 213 00:08:39,929 --> 00:08:42,149 a visualization of instantaneous 214 00:08:42,150 --> 00:08:44,459 power at the antenna input 215 00:08:44,460 --> 00:08:46,579 of a Wi-Fi radio. 216 00:08:46,580 --> 00:08:48,739 And this 217 00:08:48,740 --> 00:08:51,349 is this is not a real 218 00:08:51,350 --> 00:08:52,729 measurement. This is basically something 219 00:08:52,730 --> 00:08:53,730 that I 220 00:08:55,520 --> 00:08:57,649 try, um, I try 221 00:08:57,650 --> 00:08:59,929 to come 222 00:08:59,930 --> 00:09:02,029 up with a visualization as it as 223 00:09:02,030 --> 00:09:04,189 it actually looks like, but cleaned it up 224 00:09:04,190 --> 00:09:05,659 a little bit more. So this is not this is 225 00:09:05,660 --> 00:09:07,099 not the real measurement data here. 226 00:09:07,100 --> 00:09:09,349 But you see that there are these blocks 227 00:09:09,350 --> 00:09:12,139 and these blocks correspond to individual 228 00:09:12,140 --> 00:09:14,509 with the frame transmission's around 229 00:09:14,510 --> 00:09:15,919 the place. 230 00:09:15,920 --> 00:09:18,349 And the green line is is the, 231 00:09:18,350 --> 00:09:20,870 uh, the line from which a. 232 00:09:23,200 --> 00:09:25,119 So this could be sort of the Karasinski 233 00:09:25,120 --> 00:09:27,309 threshold at at our current receiving 234 00:09:27,310 --> 00:09:30,609 device and 235 00:09:30,610 --> 00:09:32,679 what happens is that you see 236 00:09:32,680 --> 00:09:34,839 there, I mean, it's not perfectly 237 00:09:34,840 --> 00:09:36,939 clear, but there are multiple 238 00:09:36,940 --> 00:09:38,379 overlapping all the time. 239 00:09:38,380 --> 00:09:40,479 So there is not like a slot structure 240 00:09:40,480 --> 00:09:42,669 because it's a distributed phenomenon and 241 00:09:42,670 --> 00:09:44,319 everybody has a different perception. 242 00:09:44,320 --> 00:09:46,359 It's really like if we want to have a 243 00:09:46,360 --> 00:09:47,919 clear discussion in that room, it would 244 00:09:47,920 --> 00:09:50,289 be pretty impossible just by our 245 00:09:50,290 --> 00:09:52,569 unamplified unamplified of talking 246 00:09:52,570 --> 00:09:53,499 here. 247 00:09:53,500 --> 00:09:56,259 So what an actual receiver 248 00:09:56,260 --> 00:09:58,029 is trying to understand are those parts 249 00:09:58,030 --> 00:09:59,080 that are non overlapping. 250 00:10:02,610 --> 00:10:04,799 And what he forwards to the upper 251 00:10:04,800 --> 00:10:06,389 layers, basically just the stuff that he 252 00:10:06,390 --> 00:10:08,549 could decode, and 253 00:10:08,550 --> 00:10:09,659 that is sort of the. 254 00:10:12,560 --> 00:10:14,429 The fundamental thing that, uh, 255 00:10:15,470 --> 00:10:17,929 that I want to sort of point out 256 00:10:17,930 --> 00:10:20,059 that there that the whole 257 00:10:20,060 --> 00:10:22,669 channel out the 258 00:10:22,670 --> 00:10:25,259 receiver or the upper layers to don't 259 00:10:25,260 --> 00:10:27,739 don't see the actual 260 00:10:27,740 --> 00:10:29,269 utilization of the channel, 261 00:10:30,290 --> 00:10:31,290 so. 262 00:10:32,260 --> 00:10:35,109 There is a sort of loss of information 263 00:10:35,110 --> 00:10:36,909 that happens there all the time because 264 00:10:36,910 --> 00:10:38,979 our equipment is currently working 265 00:10:38,980 --> 00:10:41,259 like it works, you only get what 266 00:10:41,260 --> 00:10:43,449 you could understand and you get some 267 00:10:43,450 --> 00:10:45,449 signal strength information and annotator 268 00:10:45,450 --> 00:10:46,449 to it. 269 00:10:46,450 --> 00:10:48,669 And this is sort of the philosophical 270 00:10:48,670 --> 00:10:50,799 part here. That is some media 271 00:10:50,800 --> 00:10:53,229 theories, if you like. 272 00:10:53,230 --> 00:10:55,029 Keyword is Marshall McLuhan here. 273 00:10:55,030 --> 00:10:57,369 They say all the technologies we develop 274 00:10:57,370 --> 00:10:59,439 are extensions of our 275 00:10:59,440 --> 00:11:01,119 human capabilities. 276 00:11:01,120 --> 00:11:02,319 And if you think about it, wireless 277 00:11:02,320 --> 00:11:04,479 communication is a little bit could be 278 00:11:04,480 --> 00:11:06,759 said as an extension of our ears 279 00:11:06,760 --> 00:11:08,349 and of our mouth. 280 00:11:08,350 --> 00:11:10,479 And this behavior here is 281 00:11:10,480 --> 00:11:12,879 equivalent to the situation where 282 00:11:12,880 --> 00:11:14,649 we are all having sort of 283 00:11:14,650 --> 00:11:15,789 electromagnetically. 284 00:11:15,790 --> 00:11:17,829 So we have we have ears in the in the two 285 00:11:17,830 --> 00:11:19,899 point four and five gigahertz 286 00:11:19,900 --> 00:11:22,149 spectrum. And when we go into a noisy 287 00:11:22,150 --> 00:11:24,340 room, we basically hear nothing. 288 00:11:25,740 --> 00:11:29,009 We just hear the stuff that someone is 289 00:11:29,010 --> 00:11:31,139 speaking to us, really standing 290 00:11:31,140 --> 00:11:33,199 directly and shouting in our ears. 291 00:11:33,200 --> 00:11:34,559 That's what we can hear. But we don't 292 00:11:34,560 --> 00:11:36,549 hear that there is activity in the room. 293 00:11:36,550 --> 00:11:38,969 We don't look for another room, 294 00:11:38,970 --> 00:11:40,679 for example, because we cannot even 295 00:11:40,680 --> 00:11:42,899 perceive the noise here. 296 00:11:42,900 --> 00:11:45,209 And that's a very long 297 00:11:45,210 --> 00:11:47,819 message saying that we now 298 00:11:47,820 --> 00:11:49,529 made a pitch for open W.T.. 299 00:11:51,210 --> 00:11:53,939 That when you go for 300 00:11:53,940 --> 00:11:56,039 real time graphs so that he has this real 301 00:11:56,040 --> 00:11:58,259 time graph, you before you even connect 302 00:11:58,260 --> 00:12:00,359 to anything, before you even see any 303 00:12:00,360 --> 00:12:02,639 other, before you do any any 304 00:12:02,640 --> 00:12:04,439 kind of association or attachment. 305 00:12:04,440 --> 00:12:05,850 You see now our third graph. 306 00:12:07,580 --> 00:12:09,589 Uh, additional to the two that were there 307 00:12:09,590 --> 00:12:11,719 before and this aircraft just shows you 308 00:12:11,720 --> 00:12:12,889 on the current channel that you're 309 00:12:12,890 --> 00:12:14,299 listening to, this is the level of 310 00:12:14,300 --> 00:12:16,549 activity like nine percent over 311 00:12:16,550 --> 00:12:17,550 time. 312 00:12:20,260 --> 00:12:21,699 I thought about doing a demo here, but I 313 00:12:21,700 --> 00:12:23,259 think it takes too much time, so I'm just 314 00:12:23,260 --> 00:12:24,519 showing you some screenshots here. 315 00:12:26,030 --> 00:12:28,259 We call it cubists lord, 316 00:12:28,260 --> 00:12:30,719 that I will explain later, 317 00:12:30,720 --> 00:12:32,729 but what you actually can do now is we 318 00:12:32,730 --> 00:12:33,730 see. 319 00:12:34,600 --> 00:12:36,789 The actual channel utilization 320 00:12:36,790 --> 00:12:39,009 we see, what are the resources left 321 00:12:39,010 --> 00:12:40,570 that we can use to communicate with? 322 00:12:42,530 --> 00:12:44,959 And this has many 323 00:12:44,960 --> 00:12:47,359 useful I mean, applications 324 00:12:47,360 --> 00:12:50,089 just because it's observable right now 325 00:12:50,090 --> 00:12:52,189 and we thought, well, 326 00:12:52,190 --> 00:12:54,559 isn't it even better 327 00:12:54,560 --> 00:12:57,139 if we can tell about 328 00:12:57,140 --> 00:12:58,140 the situations, 329 00:12:59,600 --> 00:13:01,789 if we can inform 330 00:13:01,790 --> 00:13:03,889 know the actual reasoning 331 00:13:03,890 --> 00:13:06,049 was now we can 332 00:13:06,050 --> 00:13:08,239 observe it at our local node, but 333 00:13:08,240 --> 00:13:10,349 it's still difficult to communicate this 334 00:13:10,350 --> 00:13:12,409 some way, for example, to clients that 335 00:13:12,410 --> 00:13:14,479 want to talk to 336 00:13:14,480 --> 00:13:16,789 us. So it's we we looked for ways 337 00:13:16,790 --> 00:13:19,249 to distribute this this measurement 338 00:13:19,250 --> 00:13:22,399 data somehow in order to 339 00:13:22,400 --> 00:13:24,499 create a global view of what are the 340 00:13:24,500 --> 00:13:25,999 resources in this room currently 341 00:13:26,000 --> 00:13:26,929 available. 342 00:13:26,930 --> 00:13:28,669 And we thought that we should broadcast 343 00:13:28,670 --> 00:13:31,939 these observations using beacons. 344 00:13:31,940 --> 00:13:33,829 And it turns out if you look into the 345 00:13:33,830 --> 00:13:35,149 standard documents, there is already 346 00:13:35,150 --> 00:13:37,279 defined a way to 347 00:13:37,280 --> 00:13:38,280 do this. 348 00:13:40,900 --> 00:13:44,059 Broadcasting of channel measurements 349 00:13:44,060 --> 00:13:46,419 and this thing, they call it the cupie 350 00:13:46,420 --> 00:13:48,669 is slowed and there's already 351 00:13:48,670 --> 00:13:50,289 a standardized format, how you can 352 00:13:50,290 --> 00:13:52,659 annotate this load in 353 00:13:52,660 --> 00:13:54,819 every beacon that you sent so you'll 354 00:13:54,820 --> 00:13:56,979 get seven bytes of 355 00:13:56,980 --> 00:13:59,289 additional information 356 00:13:59,290 --> 00:14:00,249 to every beacon. 357 00:14:00,250 --> 00:14:02,109 Beacon is normal management information 358 00:14:02,110 --> 00:14:04,480 that is transmitted by 359 00:14:05,590 --> 00:14:08,159 points in infrastructure mode or 360 00:14:08,160 --> 00:14:09,549 I take notes themselves. 361 00:14:11,080 --> 00:14:13,030 So all we did is basically. 362 00:14:16,540 --> 00:14:18,609 Try to implement this 363 00:14:18,610 --> 00:14:21,039 in the software stack on open 364 00:14:21,040 --> 00:14:22,040 at. 365 00:14:23,010 --> 00:14:25,289 We we were aware at that point 366 00:14:25,290 --> 00:14:27,479 in time that some access 367 00:14:27,480 --> 00:14:29,669 points, for example, the 368 00:14:29,670 --> 00:14:31,859 sort of enterprise carrier enterprise 369 00:14:31,860 --> 00:14:34,079 grade Cisco gear is already 370 00:14:34,080 --> 00:14:35,489 doing that. 371 00:14:35,490 --> 00:14:38,189 The reasoning behind is it enables 372 00:14:38,190 --> 00:14:40,559 Voice-Over-IP applications 373 00:14:40,560 --> 00:14:42,989 to have to offer a better 374 00:14:42,990 --> 00:14:44,939 quality of service because the handover 375 00:14:44,940 --> 00:14:45,940 decisions are better. 376 00:14:46,920 --> 00:14:49,979 And so we went forward and 377 00:14:49,980 --> 00:14:52,079 tried to put this into an open 378 00:14:52,080 --> 00:14:53,080 software stack. 379 00:14:55,820 --> 00:14:58,399 The patch itself that does sort of the 380 00:14:58,400 --> 00:14:59,809 main important things is really, really 381 00:14:59,810 --> 00:15:00,739 short. 382 00:15:00,740 --> 00:15:01,740 The problem is that. 383 00:15:03,440 --> 00:15:04,909 I have a background in electrical 384 00:15:04,910 --> 00:15:07,129 engineering, so I'm 385 00:15:07,130 --> 00:15:09,949 really lousy at kind of hacking, 386 00:15:09,950 --> 00:15:12,589 but I have a colleague who 387 00:15:12,590 --> 00:15:14,779 sort of has a computer science degree, 388 00:15:14,780 --> 00:15:16,969 but he's also completely new to 389 00:15:16,970 --> 00:15:18,379 Colonel Space things. 390 00:15:18,380 --> 00:15:20,749 So we pretty much screwed up everything 391 00:15:20,750 --> 00:15:22,879 that we can screw up there 392 00:15:22,880 --> 00:15:24,859 and made all the memory allocation and 393 00:15:24,860 --> 00:15:27,649 all the looking's the wrong way, 394 00:15:27,650 --> 00:15:29,299 but we still were adventurous enough. 395 00:15:29,300 --> 00:15:31,099 So basically, this is this is my 396 00:15:31,100 --> 00:15:32,299 colleague here. 397 00:15:32,300 --> 00:15:33,709 He deserves all the credit for the 398 00:15:33,710 --> 00:15:35,839 coding. I'm just the guy who sort of 399 00:15:35,840 --> 00:15:38,389 suggested it and also 400 00:15:38,390 --> 00:15:40,339 managed to get some some some funding for 401 00:15:40,340 --> 00:15:42,499 it by this new 402 00:15:42,500 --> 00:15:44,539 project called CONFINE. 403 00:15:44,540 --> 00:15:46,669 So we were naive enough to do 404 00:15:46,670 --> 00:15:48,989 our first trial at the at the wireless 405 00:15:48,990 --> 00:15:49,990 Palmach. 406 00:15:50,940 --> 00:15:53,129 In August this year in Maribor 407 00:15:53,130 --> 00:15:54,130 and. 408 00:15:55,260 --> 00:15:56,919 I will never forget the face of Attia 409 00:15:56,920 --> 00:15:59,189 when we had to 410 00:15:59,190 --> 00:16:01,199 sort of quickly find out that our patch 411 00:16:01,200 --> 00:16:03,809 totally Colonel panic's 412 00:16:03,810 --> 00:16:05,909 in a situation as in 413 00:16:05,910 --> 00:16:07,139 the battle. 414 00:16:07,140 --> 00:16:09,569 So I 415 00:16:09,570 --> 00:16:11,939 felt really sorry for them for suggesting 416 00:16:11,940 --> 00:16:14,139 that in any 417 00:16:14,140 --> 00:16:16,289 anyway, we will we 418 00:16:16,290 --> 00:16:17,879 are currently in the face of 419 00:16:19,350 --> 00:16:21,299 sort of polishing it really to make it 420 00:16:22,590 --> 00:16:24,899 sort of to convince subsystem 421 00:16:24,900 --> 00:16:27,120 developers to to let this in. 422 00:16:29,070 --> 00:16:31,289 Because it's a it's a it's a rather nasty 423 00:16:31,290 --> 00:16:33,629 thing that we are updating 424 00:16:33,630 --> 00:16:35,759 fields in a beacon over 425 00:16:35,760 --> 00:16:37,949 time. So this is not not not a usual 426 00:16:37,950 --> 00:16:40,319 thing, but we think it's relevant. 427 00:16:40,320 --> 00:16:41,969 So we want to go for it. 428 00:16:41,970 --> 00:16:44,099 And what it, for example, allows this 429 00:16:44,100 --> 00:16:46,559 rebroadcasting is something like that. 430 00:16:46,560 --> 00:16:48,719 So this is a few that you can 431 00:16:48,720 --> 00:16:50,459 get when you use, uh, 432 00:16:51,540 --> 00:16:54,119 a very nice tool called Horst. 433 00:16:54,120 --> 00:16:55,979 Highly optimized video scanning tool is 434 00:16:55,980 --> 00:16:56,980 one way. 435 00:16:57,980 --> 00:17:00,449 So Bruno wrote it or is maintaining 436 00:17:00,450 --> 00:17:02,999 it. And it's it's pretty useful 437 00:17:03,000 --> 00:17:05,309 when you're doing, uh, debugging of, 438 00:17:05,310 --> 00:17:07,469 uh, Metrolink's 439 00:17:07,470 --> 00:17:09,088 and now we made a little patch to it. 440 00:17:09,089 --> 00:17:11,669 So this the part on the right side is new 441 00:17:11,670 --> 00:17:13,078 in the pattern, right. That just says you 442 00:17:13,079 --> 00:17:15,269 OK, this and these Mac addresses now 443 00:17:15,270 --> 00:17:17,039 see from their point of view, these and 444 00:17:17,040 --> 00:17:19,439 these loads, if it 445 00:17:19,440 --> 00:17:21,509 if if you look, they are 446 00:17:21,510 --> 00:17:23,568 all operating on the same channel, 447 00:17:23,569 --> 00:17:25,328 they are all within communication range. 448 00:17:25,329 --> 00:17:27,568 So this is this is an output 449 00:17:27,569 --> 00:17:29,039 you see just with your normal Wipha 450 00:17:29,040 --> 00:17:30,450 receiver. So I could 451 00:17:31,590 --> 00:17:33,119 I could demonstrate it right here with a 452 00:17:33,120 --> 00:17:35,519 normal with with a normal wife radio. 453 00:17:35,520 --> 00:17:37,229 I see access points around here 454 00:17:37,230 --> 00:17:39,539 broadcasting the channel Load Fuze, 455 00:17:39,540 --> 00:17:41,189 and I see that they're different. 456 00:17:41,190 --> 00:17:42,190 So this. 457 00:17:43,200 --> 00:17:45,389 This is a perfect demonstration of it's 458 00:17:45,390 --> 00:17:46,979 a distributed phenomenon, everybody has a 459 00:17:46,980 --> 00:17:48,119 different perception of noise. 460 00:17:49,670 --> 00:17:50,810 Why is this all 461 00:17:52,520 --> 00:17:54,169 so relevant? 462 00:17:54,170 --> 00:17:55,490 I mean, it's relevant for. 463 00:17:57,430 --> 00:17:58,869 The cases that were mentioned in the 464 00:17:58,870 --> 00:18:00,909 intro, for example, but there is another 465 00:18:00,910 --> 00:18:02,609 thing that I'm 466 00:18:04,570 --> 00:18:06,279 there's another reason why I wanted to 467 00:18:06,280 --> 00:18:07,539 sort of. 468 00:18:07,540 --> 00:18:09,130 Make this point here, and there is, 469 00:18:10,960 --> 00:18:13,479 um, currently 470 00:18:13,480 --> 00:18:15,639 there is a lot of activity 471 00:18:15,640 --> 00:18:17,719 around the five, so there 472 00:18:17,720 --> 00:18:19,359 are lots of things happening because of 473 00:18:19,360 --> 00:18:21,549 it, because of technology operating 474 00:18:21,550 --> 00:18:22,749 in this. 475 00:18:22,750 --> 00:18:25,119 They call it, uh, you and 476 00:18:25,120 --> 00:18:27,429 I and the uni 477 00:18:27,430 --> 00:18:30,169 bands and this activity, 478 00:18:30,170 --> 00:18:31,170 uh. 479 00:18:33,700 --> 00:18:35,769 One part of it is due to 480 00:18:35,770 --> 00:18:38,229 the cellular industry 481 00:18:38,230 --> 00:18:40,539 looking at it from, 482 00:18:40,540 --> 00:18:41,540 uh. 483 00:18:43,880 --> 00:18:46,159 So they are they are looking at ways now 484 00:18:46,160 --> 00:18:48,799 very concretely to use these. 485 00:18:50,060 --> 00:18:51,230 Unlicenced bands. 486 00:18:52,240 --> 00:18:54,309 The terms are LTU or 487 00:18:54,310 --> 00:18:55,810 license assisted exis. 488 00:18:59,130 --> 00:19:00,130 And. 489 00:19:00,930 --> 00:19:02,909 I mean, it's a matter of regulation 490 00:19:02,910 --> 00:19:05,429 whether this will actually be, 491 00:19:05,430 --> 00:19:06,430 uh. 492 00:19:08,200 --> 00:19:09,699 Put in place. 493 00:19:11,410 --> 00:19:13,719 But it's expected 494 00:19:13,720 --> 00:19:15,639 that there I mean, the whole point of 495 00:19:15,640 --> 00:19:17,769 doing this is in order to 496 00:19:18,790 --> 00:19:19,790 have more control. 497 00:19:24,610 --> 00:19:27,669 So cellular industry is looking for ways 498 00:19:27,670 --> 00:19:29,769 to expand their domain of 499 00:19:29,770 --> 00:19:32,109 control into the into the unlicensed 500 00:19:32,110 --> 00:19:33,969 bands. And the problem is, if this 501 00:19:33,970 --> 00:19:36,439 happens, since our Wi-Fi 502 00:19:36,440 --> 00:19:39,399 radios cannot understand those 503 00:19:39,400 --> 00:19:41,799 differently formatted signals, 504 00:19:41,800 --> 00:19:44,019 we will hear basically nothing if this 505 00:19:44,020 --> 00:19:45,849 activity in the neighborhood starts to 506 00:19:45,850 --> 00:19:46,850 appear. 507 00:19:49,420 --> 00:19:50,420 So. 508 00:19:51,190 --> 00:19:53,619 These are put 509 00:19:53,620 --> 00:19:56,139 these are some some some early, 510 00:19:56,140 --> 00:19:59,079 uh, measurements that people did 511 00:19:59,080 --> 00:20:01,329 that show clearly, that the presence of 512 00:20:01,330 --> 00:20:03,609 such in this case is an LTU 513 00:20:03,610 --> 00:20:05,979 base station will greatly diminish 514 00:20:05,980 --> 00:20:07,599 all the performance that we can out of 515 00:20:07,600 --> 00:20:09,730 the five gigahertz bands using Wi-Fi. 516 00:20:11,540 --> 00:20:13,639 Again, it's a matter of regulation, 517 00:20:13,640 --> 00:20:15,889 the regulation, how how how 518 00:20:15,890 --> 00:20:18,199 strict you will be in this 519 00:20:18,200 --> 00:20:20,269 equal access and coexistence if 520 00:20:20,270 --> 00:20:21,739 they really take it, if you really take 521 00:20:21,740 --> 00:20:24,259 coexistence and equal access seriously, 522 00:20:24,260 --> 00:20:26,929 there is no point in doing LTU 523 00:20:26,930 --> 00:20:29,419 or license assisted access 524 00:20:29,420 --> 00:20:31,579 simply because if you want to have an 525 00:20:31,580 --> 00:20:33,649 equal equal access, 526 00:20:33,650 --> 00:20:35,629 you basically have to use the same medium 527 00:20:35,630 --> 00:20:37,889 access protocol, listen before talk 528 00:20:37,890 --> 00:20:40,399 the same rights to every note and then 529 00:20:40,400 --> 00:20:42,709 the whole point practically vanishes 530 00:20:42,710 --> 00:20:44,899 because the channel 531 00:20:44,900 --> 00:20:46,999 is not such 532 00:20:47,000 --> 00:20:49,519 different between those two standards. 533 00:20:49,520 --> 00:20:50,520 And I think that it's. 534 00:20:51,520 --> 00:20:53,619 Partly because of this, that 535 00:20:53,620 --> 00:20:55,839 now this this this discussion has 536 00:20:55,840 --> 00:20:57,490 started because. 537 00:21:01,400 --> 00:21:02,749 No, wait, let me skip this. 538 00:21:03,870 --> 00:21:04,870 Uh. 539 00:21:05,650 --> 00:21:08,049 The FCC and the European Commission 540 00:21:08,050 --> 00:21:10,269 have just or the FCC 541 00:21:10,270 --> 00:21:11,709 is about to join the European Commission 542 00:21:11,710 --> 00:21:13,569 already has put new law 543 00:21:15,130 --> 00:21:18,189 into place that now seems to 544 00:21:18,190 --> 00:21:19,190 try to. 545 00:21:22,660 --> 00:21:24,480 Uh, make it 546 00:21:25,950 --> 00:21:28,049 I would say if you look at 547 00:21:28,050 --> 00:21:30,179 the new law, it says that they want to 548 00:21:30,180 --> 00:21:33,509 make it impossible that people change 549 00:21:33,510 --> 00:21:35,939 the settings on their routers 550 00:21:35,940 --> 00:21:37,469 and they're especially worried about the 551 00:21:37,470 --> 00:21:38,549 uni bands again. 552 00:21:40,950 --> 00:21:43,259 And the official 553 00:21:43,260 --> 00:21:44,999 reason it always is about the weather out 554 00:21:45,000 --> 00:21:46,469 there, but I don't I don't think it's a 555 00:21:46,470 --> 00:21:48,869 coincidence that they are now that 556 00:21:48,870 --> 00:21:51,539 there is now a law being put into place 557 00:21:51,540 --> 00:21:53,729 that restricts the behavior 558 00:21:53,730 --> 00:21:55,589 of or restricts the freedoms that we 559 00:21:55,590 --> 00:21:56,590 have. 560 00:21:57,600 --> 00:21:59,489 And at the same time, there's a there's 561 00:21:59,490 --> 00:22:01,859 another secondary user 562 00:22:01,860 --> 00:22:03,869 sort of to say the primary uses is to 563 00:22:05,390 --> 00:22:07,199 the radars here in the secondary use of 564 00:22:07,200 --> 00:22:08,519 what all this wi fi. 565 00:22:08,520 --> 00:22:10,859 And now it appears that another secondary 566 00:22:10,860 --> 00:22:13,079 user wants all 567 00:22:13,080 --> 00:22:15,029 Wi-Fi users to be as restrictive as 568 00:22:15,030 --> 00:22:16,769 possible. That that that's how I 569 00:22:16,770 --> 00:22:17,940 interpret the whole thing. 570 00:22:21,940 --> 00:22:24,309 Yo, this 10 little patch 571 00:22:24,310 --> 00:22:25,310 will allow us. 572 00:22:27,050 --> 00:22:29,299 If that's the big if 573 00:22:29,300 --> 00:22:31,039 if we are still allowed to run open 574 00:22:31,040 --> 00:22:32,869 software on these things, this channel 575 00:22:32,870 --> 00:22:35,389 will at least allow us to detect 576 00:22:35,390 --> 00:22:36,390 that there are 577 00:22:37,460 --> 00:22:39,589 nearby licenses that exist or else to 578 00:22:39,590 --> 00:22:41,329 base stations that may be the cause of 579 00:22:41,330 --> 00:22:42,330 problems there. 580 00:22:44,720 --> 00:22:47,059 I think I should come I have 581 00:22:47,060 --> 00:22:49,159 a little comment to make, but I 582 00:22:49,160 --> 00:22:51,979 think I'll skip it. And basically 583 00:22:51,980 --> 00:22:53,539 the message I'm trying to make here is. 584 00:22:55,420 --> 00:22:58,089 We should treat these spectrum resources 585 00:22:58,090 --> 00:23:00,279 as our comments, so. 586 00:23:02,080 --> 00:23:04,089 Sort of it's. 587 00:23:05,960 --> 00:23:08,149 It's about the right for free 588 00:23:08,150 --> 00:23:10,459 speech in 589 00:23:10,460 --> 00:23:11,749 in an electromagnetic way. 590 00:23:12,800 --> 00:23:14,899 These are the same sort of 591 00:23:14,900 --> 00:23:16,639 the force of the freedoms that Stallmann 592 00:23:16,640 --> 00:23:17,809 talks about. 593 00:23:17,810 --> 00:23:19,999 We need them also, and we have 594 00:23:20,000 --> 00:23:23,149 to ensure that we still retain 595 00:23:23,150 --> 00:23:25,459 the freedom to express 596 00:23:25,460 --> 00:23:26,710 electromagnetically sort of. 597 00:23:27,920 --> 00:23:28,920 That's my point of. 598 00:23:39,080 --> 00:23:40,080 Christians. 599 00:23:42,410 --> 00:23:43,410 OK, 600 00:23:45,170 --> 00:23:47,589 so we have some more time for questions. 601 00:23:49,710 --> 00:23:51,629 If you want to ask a question and you are 602 00:23:51,630 --> 00:23:53,759 listening on the stream, please go 603 00:23:53,760 --> 00:23:55,679 in the IOC. 604 00:23:55,680 --> 00:23:57,569 We have an angel listening there and 605 00:23:57,570 --> 00:24:00,539 reading the questions out loud. 606 00:24:00,540 --> 00:24:01,829 If you have a question in the room, 607 00:24:01,830 --> 00:24:04,769 please come to one of the six 608 00:24:04,770 --> 00:24:06,959 microphones and I see we already have 609 00:24:06,960 --> 00:24:08,729 the first one on microphone number two. 610 00:24:09,740 --> 00:24:12,239 I'm just a small question regarding 611 00:24:12,240 --> 00:24:14,309 to yours truly where you 612 00:24:14,310 --> 00:24:16,589 made a pitch, where you see on the right 613 00:24:16,590 --> 00:24:17,849 side the cube says, 614 00:24:18,990 --> 00:24:21,749 I didn't fully understand because 615 00:24:21,750 --> 00:24:24,329 you just said that this curious 616 00:24:24,330 --> 00:24:26,429 is broadcasted on the beacon's. 617 00:24:26,430 --> 00:24:28,469 But you could clearly see that normal 618 00:24:28,470 --> 00:24:31,109 stations were also shown their 619 00:24:31,110 --> 00:24:32,290 point of view of 620 00:24:33,900 --> 00:24:36,209 how much interference they see how 621 00:24:36,210 --> 00:24:37,210 this system. 622 00:24:38,850 --> 00:24:40,439 Because if it's only the access point, 623 00:24:40,440 --> 00:24:42,419 sending it in the pictures, how can you 624 00:24:42,420 --> 00:24:45,209 see the station if you 625 00:24:45,210 --> 00:24:46,559 know it's not a station for you? 626 00:24:46,560 --> 00:24:48,119 It's a basically what I'm what I wanted 627 00:24:48,120 --> 00:24:50,039 to say is that you see several access 628 00:24:50,040 --> 00:24:52,109 points there and 629 00:24:52,110 --> 00:24:53,999 sort of all of the all of those give 630 00:24:54,000 --> 00:24:55,979 their individual views. 631 00:24:55,980 --> 00:24:58,199 But that gives you all access 632 00:24:58,200 --> 00:24:59,779 points all on the same page. 633 00:24:59,780 --> 00:25:00,929 OK, yeah. 634 00:25:00,930 --> 00:25:02,519 Second question, because you said that 635 00:25:02,520 --> 00:25:04,949 some vendors implement this 636 00:25:04,950 --> 00:25:06,269 already in the standards. 637 00:25:06,270 --> 00:25:07,229 Yeah. 638 00:25:07,230 --> 00:25:09,269 If somebody is using the patch. 639 00:25:10,410 --> 00:25:12,539 Are these interoperability comparable? 640 00:25:12,540 --> 00:25:14,939 So if I have a group of workstation Cisco 641 00:25:14,940 --> 00:25:16,169 workstation. Yeah. 642 00:25:16,170 --> 00:25:17,669 It's OK. Yeah. 643 00:25:17,670 --> 00:25:18,670 Thanks. 644 00:25:19,390 --> 00:25:21,489 OK, next question from microphone 645 00:25:21,490 --> 00:25:23,619 number four, hello, thank 646 00:25:23,620 --> 00:25:24,219 you for the talk. 647 00:25:24,220 --> 00:25:25,629 Just a quick question. You had a child, 648 00:25:25,630 --> 00:25:27,639 were you short holiday before performance 649 00:25:27,640 --> 00:25:30,169 grades? If base station is Nehama, 650 00:25:30,170 --> 00:25:32,259 yeah. Do you think anybody 651 00:25:32,260 --> 00:25:34,109 will go for it? Because everybody uses 652 00:25:34,110 --> 00:25:35,269 for the public outcry? 653 00:25:35,270 --> 00:25:36,969 I can't imagine that can ever be reality. 654 00:25:36,970 --> 00:25:38,009 Right. 655 00:25:38,010 --> 00:25:39,180 I mean, my would stop working 656 00:25:40,450 --> 00:25:40,959 will see. 657 00:25:40,960 --> 00:25:42,549 I mean, I also I'm yeah. 658 00:25:42,550 --> 00:25:43,929 I'm having the same view. I'm just 659 00:25:43,930 --> 00:25:45,039 puzzled. Thank you. 660 00:25:46,980 --> 00:25:49,319 OK, I see we have a question 661 00:25:49,320 --> 00:25:51,109 up there at microphone number five. 662 00:25:52,350 --> 00:25:54,659 All right, I 663 00:25:54,660 --> 00:25:56,969 have a short question 664 00:25:56,970 --> 00:25:58,229 regarding 665 00:25:59,640 --> 00:26:02,309 measurements. Have you looked at 666 00:26:02,310 --> 00:26:03,310 the 667 00:26:05,070 --> 00:26:07,139 how this behaves 668 00:26:07,140 --> 00:26:09,899 with throughput from access points? 669 00:26:09,900 --> 00:26:11,489 So if you can actually 670 00:26:12,600 --> 00:26:15,959 detect before even you can 671 00:26:15,960 --> 00:26:17,849 before you transmit data how the 672 00:26:17,850 --> 00:26:19,619 throughput will be? 673 00:26:19,620 --> 00:26:21,479 No, but that's an excellent idea. 674 00:26:21,480 --> 00:26:23,699 I mean, building an estimated 675 00:26:23,700 --> 00:26:26,249 four expectable throughput 676 00:26:26,250 --> 00:26:28,169 would be now possible with that. 677 00:26:28,170 --> 00:26:29,309 That was. Yeah. 678 00:26:30,460 --> 00:26:32,019 Excellent, thanks. 679 00:26:32,020 --> 00:26:33,619 Good microphone number four. 680 00:26:34,800 --> 00:26:37,119 Hello, what data do you get 681 00:26:37,120 --> 00:26:39,249 from the hardware and is it similar 682 00:26:39,250 --> 00:26:41,149 from all the devices? 683 00:26:41,150 --> 00:26:43,089 Yeah, the chips, the data. 684 00:26:43,090 --> 00:26:45,219 We get our cycle counters, 685 00:26:45,220 --> 00:26:46,389 they are called cycle counters. 686 00:26:46,390 --> 00:26:48,459 And we for this, we're using the business 687 00:26:48,460 --> 00:26:50,349 cycle counters or to receive cycle, 688 00:26:50,350 --> 00:26:51,249 whatever we like. 689 00:26:51,250 --> 00:26:53,589 And these are registers that are being 690 00:26:53,590 --> 00:26:55,989 sort of in most drivers 691 00:26:55,990 --> 00:26:57,879 they are being copied in. 692 00:26:57,880 --> 00:26:59,619 So they are available via the net link 693 00:26:59,620 --> 00:27:00,620 interface. 694 00:27:01,240 --> 00:27:03,759 And you dispatch 695 00:27:03,760 --> 00:27:05,409 just reads them out in this place. 696 00:27:05,410 --> 00:27:07,329 And it's a very simple thing, actually. 697 00:27:07,330 --> 00:27:10,389 And most most current hardware 698 00:27:10,390 --> 00:27:12,639 or most current drivers are 699 00:27:12,640 --> 00:27:13,779 exporting this feature, so. 700 00:27:16,330 --> 00:27:18,339 So microphone number one. 701 00:27:20,230 --> 00:27:22,389 Do you have an idea how we can keep this 702 00:27:22,390 --> 00:27:23,799 spectrum in the comments? 703 00:27:25,180 --> 00:27:27,309 So I said, I think question, 704 00:27:27,310 --> 00:27:29,079 I think the only way we have is to make 705 00:27:29,080 --> 00:27:30,999 make it visible, what would we do with 706 00:27:31,000 --> 00:27:33,069 it? So this is if you 707 00:27:33,070 --> 00:27:35,139 take this further, we can see, OK, my 708 00:27:35,140 --> 00:27:36,369 neighborhood is full of stuff 709 00:27:37,660 --> 00:27:39,879 and 90 percent of it is El 710 00:27:39,880 --> 00:27:42,009 Tube station. So I will certainly 711 00:27:42,010 --> 00:27:43,119 try to do something about it. 712 00:27:43,120 --> 00:27:45,039 Right. But if you cannot see it, you 713 00:27:45,040 --> 00:27:46,040 cannot act. 714 00:27:47,320 --> 00:27:49,539 So now we 715 00:27:50,590 --> 00:27:52,389 push in very shortly a question from the 716 00:27:52,390 --> 00:27:53,390 Internet. 717 00:27:54,130 --> 00:27:56,509 Do these chipsets matter 718 00:27:56,510 --> 00:27:58,689 all interfaces or just use 719 00:27:58,690 --> 00:28:00,310 it as a Wi-Fi? 720 00:28:01,470 --> 00:28:03,689 Sorry, can you say it again to these 721 00:28:03,690 --> 00:28:06,089 chips that measure all interfaces 722 00:28:06,090 --> 00:28:08,159 or just user usage of other 723 00:28:08,160 --> 00:28:09,160 wildfire's? 724 00:28:10,640 --> 00:28:12,199 Was no. 725 00:28:12,200 --> 00:28:14,529 I don't know if the 726 00:28:14,530 --> 00:28:16,639 chips use only 727 00:28:16,640 --> 00:28:18,409 the user interface. 728 00:28:18,410 --> 00:28:20,659 All in all interfaces or just 729 00:28:20,660 --> 00:28:23,269 the user interface, especially interfaces 730 00:28:23,270 --> 00:28:25,729 not just by wi fi 731 00:28:25,730 --> 00:28:26,469 trust Wi-Fi. 732 00:28:26,470 --> 00:28:28,729 And so that's what I would 733 00:28:28,730 --> 00:28:29,959 say. 734 00:28:29,960 --> 00:28:30,960 I don't get it. 735 00:28:32,240 --> 00:28:33,979 So just Wi-Fi. 736 00:28:33,980 --> 00:28:34,980 OK, 737 00:28:36,050 --> 00:28:37,579 microphone number four. 738 00:28:37,580 --> 00:28:39,649 Yeah. Um, so have you taken a 739 00:28:39,650 --> 00:28:42,319 look at a pair of Qualcomm's 740 00:28:42,320 --> 00:28:44,989 spectrum analysis mode for this? 741 00:28:44,990 --> 00:28:47,749 And does it actually qualify 742 00:28:47,750 --> 00:28:50,139 to supplement your data? 743 00:28:50,140 --> 00:28:51,889 Sure. It's a nice feature. 744 00:28:51,890 --> 00:28:54,079 The the when we 745 00:28:54,080 --> 00:28:55,099 played with it, we just had the 746 00:28:55,100 --> 00:28:57,769 impression that only sparsity samples. 747 00:28:57,770 --> 00:29:00,559 So it's not only about every 748 00:29:00,560 --> 00:29:01,679 half a second or something. 749 00:29:01,680 --> 00:29:03,909 Yeah. It's for for estimating channel 750 00:29:03,910 --> 00:29:05,540 load. It's not really that use of new. 751 00:29:06,690 --> 00:29:09,449 But also to get a look at 752 00:29:09,450 --> 00:29:11,219 how you can classify what kind of 753 00:29:11,220 --> 00:29:13,289 interferer that is, that should 754 00:29:13,290 --> 00:29:15,749 also be put into visualization, 755 00:29:15,750 --> 00:29:17,879 because there's no I know that 756 00:29:17,880 --> 00:29:20,159 IBM has options for that on their 757 00:29:20,160 --> 00:29:23,009 checks, but there's no real option on 758 00:29:23,010 --> 00:29:25,139 open WRP now 759 00:29:25,140 --> 00:29:25,749 to do it. 760 00:29:25,750 --> 00:29:27,809 And it's really not that easy 761 00:29:27,810 --> 00:29:28,949 to do it by hand. 762 00:29:28,950 --> 00:29:30,959 So, yeah, would be great to have an 763 00:29:30,960 --> 00:29:31,629 option for that. 764 00:29:31,630 --> 00:29:34,289 If someone is interested, please come. 765 00:29:34,290 --> 00:29:35,309 Let's talk. 766 00:29:35,310 --> 00:29:37,499 OK, we have time for one last 767 00:29:37,500 --> 00:29:38,159 question. 768 00:29:38,160 --> 00:29:39,450 Microphone number four, please. 769 00:29:40,600 --> 00:29:41,600 Yeah. 770 00:29:42,150 --> 00:29:43,679 Yeah, I just. 771 00:29:43,680 --> 00:29:46,499 Did you have experience with five years 772 00:29:46,500 --> 00:29:48,569 and is there is 773 00:29:48,570 --> 00:29:50,489 a possibility to get the information from 774 00:29:50,490 --> 00:29:52,769 the client also about 775 00:29:52,770 --> 00:29:54,989 inter activities 776 00:29:54,990 --> 00:29:56,189 in there from the. 777 00:29:58,470 --> 00:30:00,569 Which client would you refer to, 778 00:30:00,570 --> 00:30:02,489 if you like your mobile phone or 779 00:30:05,460 --> 00:30:07,349 you mean that if a client is typically 780 00:30:07,350 --> 00:30:08,309 not sending Bekins? 781 00:30:08,310 --> 00:30:09,539 Is that what you mean? Yeah. 782 00:30:09,540 --> 00:30:11,519 Yeah, that's right. 783 00:30:11,520 --> 00:30:13,679 Yeah, yeah. I mean that we 784 00:30:13,680 --> 00:30:15,779 can take it to other management 785 00:30:15,780 --> 00:30:17,399 firms like the pro-Brexit. 786 00:30:17,400 --> 00:30:19,199 Still send out all the time. 787 00:30:19,200 --> 00:30:21,179 Mm. It would be a seven bytes addition 788 00:30:21,180 --> 00:30:22,479 which is pretty small. 789 00:30:22,480 --> 00:30:24,689 OK, let's just OK 790 00:30:24,690 --> 00:30:27,569 for everybody who still has a question, 791 00:30:27,570 --> 00:30:29,819 Paul will be in the foyer afterwards 792 00:30:29,820 --> 00:30:31,709 so you can meet him there and ask your 793 00:30:31,710 --> 00:30:32,910 leftover questions. 794 00:30:34,990 --> 00:30:36,549 Thank you very much. 795 00:30:36,550 --> 00:30:38,289 It was really nice talk. 796 00:30:38,290 --> 00:30:40,180 So the last round of applause for.