1
00:00:00,000 --> 00:00:04,900
If you know what's good for you, now's the time you be quiet, because this man is literally

2
00:00:04,900 --> 00:00:09,120
invited just so that I can watch him, and he's awesome, so we will throw you out.

3
00:00:15,100 --> 00:00:19,940
All right, ladies and gentlemen, I would like to present Quadling, who normally yells.

4
00:00:20,120 --> 00:00:22,840
No, I would like to present Bonlint Sieber.

5
00:00:23,160 --> 00:00:27,660
He is literally invited every year so that all of us running the contest can stare in

6
00:00:27,660 --> 00:00:31,160
his direction with our mouths open, wondering what the hell just happened.

7
00:00:31,760 --> 00:00:36,480
His bio is kind of lame, software engineer by training, Bonlint is a perpetual hacker,

8
00:00:36,920 --> 00:00:41,840
director of vulnerability research at Bastille Networks, and the guy behind Spench.net.

9
00:00:42,120 --> 00:00:46,820
What you should really know is while most of you were discussing burner phones on Twitter,

10
00:00:47,100 --> 00:00:51,880
this man came into our village, I think it was three years ago, three years ago, and

11
00:00:51,880 --> 00:00:55,240
he said, I need to redirect a space probe real quick.

12
00:00:55,340 --> 00:00:57,280
Can anybody loan me a cell card?

13
00:00:57,660 --> 00:01:02,680
And he borrowed a cell card from a random stranger and beamed commands to something

14
00:01:02,680 --> 00:01:05,580
flying in space to try to redirect it around the planet.

15
00:01:06,700 --> 00:01:13,100
So, yeah, if you all are scared of DEF CON and he's like redirecting space, seriously,

16
00:01:13,440 --> 00:01:15,660
sit down, have a good old time.

17
00:01:15,840 --> 00:01:21,660
I would like to, with the proudest feeling in my chest, introduce Bonlint Sieber and

18
00:01:21,660 --> 00:01:26,380
hacking some more of the wireless world, an extension of his rant of hacking the wireless

19
00:01:26,380 --> 00:01:27,280
world last year.

20
00:01:27,660 --> 00:01:31,760
This will be cut off when he damn well feels like it, or sometime late after lunch.

21
00:01:31,980 --> 00:01:36,540
So, sit down and make yourselves at home, meet your neighbors, and thank you, Bonlint.

22
00:01:42,820 --> 00:01:45,160
Thank you for that very kind introduction, Zero.

23
00:01:45,900 --> 00:01:51,760
And thank you all for coming to the wireless village over some of the other excellent talks

24
00:01:51,760 --> 00:01:53,340
that are going on at the moment at DEF CON.

25
00:01:54,960 --> 00:01:57,640
We just heard an interesting story from two DEF CONs.

26
00:01:57,660 --> 00:01:59,160
Is my voice level all right?

27
00:01:59,220 --> 00:01:59,720
Can you hear me okay?

28
00:01:59,820 --> 00:02:00,580
Up the back?

29
00:02:00,640 --> 00:02:00,940
Yeah?

30
00:02:01,020 --> 00:02:01,440
Thumbs up?

31
00:02:01,520 --> 00:02:01,820
Great.

32
00:02:02,700 --> 00:02:03,440
Where's Anders?

33
00:02:03,820 --> 00:02:04,600
Anders, where are you sitting?

34
00:02:04,820 --> 00:02:09,720
Anders is the man that let me tether to his phone so that I could get on the internet

35
00:02:09,720 --> 00:02:15,600
and talk to the computer at Arecibo, at the Arecibo telescope, to beam commands to IC3.

36
00:02:15,800 --> 00:02:20,740
So, thank you very much, Anders, once again, for trusting me with your internet connection.

37
00:02:23,000 --> 00:02:25,400
So, my name's Bonlint, and I'm...

38
00:02:26,900 --> 00:02:27,300
I'm...

39
00:02:27,300 --> 00:02:33,720
I work with the vulnerability research team at Bastille, and I'm going to talk to you today about three things.

40
00:02:33,920 --> 00:02:39,620
Some of them I started introducing last year, and I'd like to show you some of the additional work that I've done on them since then.

41
00:02:40,100 --> 00:02:41,740
One of them briefly will be Inmarsat.

42
00:02:42,940 --> 00:02:47,060
Another one is implementing something that we saw at Cyber Spectrum.

43
00:02:47,260 --> 00:02:53,260
Previously, Cyber Spectrum is a meetup that we run for people interested in sharing projects on software-defined radio,

44
00:02:53,440 --> 00:02:54,940
in the Bay Area in particular.

45
00:02:55,240 --> 00:02:57,160
And then also some new work that I've done.

46
00:02:57,300 --> 00:03:06,220
So, I'm going to show you some of the things that I've done with Radar, and hopefully, my aim today is that everybody understands a little bit more about how Radar processing actually works.

47
00:03:06,500 --> 00:03:09,820
And so, if you have any questions at any point, please feel free to ask.

48
00:03:10,280 --> 00:03:15,280
I will leave you satisfied if I see nodding heads and smiles, as opposed to quizzical looks.

49
00:03:15,360 --> 00:03:17,300
So, let me know how that goes.

50
00:03:18,140 --> 00:03:24,100
And again, most of this time with GNU Radio, and I'll show you some flow graphs and the general concepts behind them.

51
00:03:24,900 --> 00:03:27,120
So, the first item is going to be Inmarsat.

52
00:03:27,300 --> 00:03:31,900
And later on, I'll also be encouraging audience participation.

53
00:03:32,520 --> 00:03:34,120
I've got some live demos here.

54
00:03:34,980 --> 00:03:38,620
So, that hopefully will go well with the gods willing.

55
00:03:38,840 --> 00:03:39,980
So, just a very quick recap.

56
00:03:40,740 --> 00:03:45,520
If you're interested in more of the details, you can visit the last year's talk.

57
00:03:45,840 --> 00:03:47,900
The Inmarsat is a constellation of birds.

58
00:03:48,100 --> 00:03:57,220
And they offer a number of services, one of which is Inmarsat Aero, which is used by airplanes that travel, you know, elsewhere.

59
00:03:57,300 --> 00:04:03,800
Not over terrestrial coverage of normal VHF ground networks, but if they're particularly going over the ocean.

60
00:04:04,840 --> 00:04:07,320
And Inmarsat is a bent pipe design.

61
00:04:07,910 --> 00:04:12,820
So, transmissions will be sent from aircraft to ground stations and back.

62
00:04:13,380 --> 00:04:19,240
It's actually pretty easy to listen to this over these, on the C-band downlink.

63
00:04:19,800 --> 00:04:24,900
You can just get a dish, make your own feed, use an LNA and a bandpass filter.

64
00:04:25,100 --> 00:04:27,280
And then you're able to decode at least one of these.

65
00:04:27,300 --> 00:04:33,820
The P-channel, which is what I looked at to begin with, to decode basically ACAR's star messages going over this link.

66
00:04:34,240 --> 00:04:37,020
And I have to give credit here to my friend Ian Buckley.

67
00:04:37,160 --> 00:04:40,440
This is a setup that he made so that we could receive these signals.

68
00:04:41,640 --> 00:04:43,360
So, again, I won't go through the detail.

69
00:04:43,480 --> 00:04:44,260
I covered it last time.

70
00:04:44,460 --> 00:04:46,200
But you can start demodulating it.

71
00:04:46,260 --> 00:04:48,440
It's a simple GMSK signal.

72
00:04:48,700 --> 00:04:55,740
And then you can work through the process to actually turn what appears to be noise into intelligible data.

73
00:04:55,940 --> 00:04:57,280
So, the first step would be...

74
00:04:57,300 --> 00:05:02,400
And now, I want you to remember the idea of an interleaver or deinterleaver.

75
00:05:02,500 --> 00:05:05,200
So, the idea is that you read data into the rows.

76
00:05:05,520 --> 00:05:10,040
And then once the memory cell is full, then you read it out from the columns.

77
00:05:10,300 --> 00:05:16,140
And the purpose of this for the transmitter is to redistribute the information bits over time.

78
00:05:16,360 --> 00:05:22,940
So, if you have a burst noise source, then it won't knock out all the bits in a single packet.

79
00:05:23,180 --> 00:05:26,540
Those missed bits will be distributed over time through many packets.

80
00:05:26,540 --> 00:05:35,500
And then you can use convolutional decoding or forward error correction to fix those individual erroneous bits over time.

81
00:05:35,640 --> 00:05:38,480
And therefore, you get a much better chance of recovering any corrupted data.

82
00:05:38,760 --> 00:05:44,780
Once you do the forward error correction, some descrambling, and then things look much more structured.

83
00:05:45,100 --> 00:05:48,580
And then on the P channel here, this was a coordination channel.

84
00:05:48,800 --> 00:05:54,580
And some of the messages are actually user messages that contain text information.

85
00:05:54,980 --> 00:05:56,180
And if you're a...

86
00:05:56,540 --> 00:06:02,660
If you're a plane spotter and you like to listen to ACARS, this is essentially the equivalent, but coming down from a geostationary satellite.

87
00:06:03,360 --> 00:06:11,640
And it's interesting because there's a slightly different flavor to them in terms of navigating and sending directions to planes which are, you know, anywhere above the Pacific, for example.

88
00:06:12,020 --> 00:06:18,000
So, just some examples to give you a feel, you know, random human-readable text messages to the crew.

89
00:06:18,200 --> 00:06:22,920
You've got weather reports in standard notation there.

90
00:06:23,840 --> 00:06:26,520
A different form of weather report, commonly used.

91
00:06:26,540 --> 00:06:27,080
They're used in aviation.

92
00:06:27,760 --> 00:06:44,040
You've got AFN, ADSC, CPDLC, which are different sorts of encodings for, you know, messages that pertain to routing and traffic flow and various other sorts of things.

93
00:06:44,900 --> 00:06:47,180
And then this is a good one that I saw.

94
00:06:47,380 --> 00:06:48,740
Remember the Galaxy Note 7?

95
00:06:50,240 --> 00:06:54,700
They would remind everybody over the Inmarsat error P channel as well.

96
00:06:54,700 --> 00:07:01,280
And then a lot of these messages are also printed out on the cockpit printer or sent to the captain.

97
00:07:02,680 --> 00:07:06,920
So, you know, informing them about preferred routes for their flight plan.

98
00:07:07,680 --> 00:07:12,380
This one is a custom message saying that there's really bad weather somewhere.

99
00:07:13,140 --> 00:07:22,160
And this one is interesting because it's actually a step-by-step to, I think, reset one of the displays or subsystems on the aircraft or in the cockpit.

100
00:07:23,080 --> 00:07:24,160
This one actually...

101
00:07:24,700 --> 00:07:26,220
It would be a message that you print out.

102
00:07:26,440 --> 00:07:31,700
And so just before they close the doors, you know, they print out the manifest and then the pilot, the captain signs it.

103
00:07:31,920 --> 00:07:38,240
And so they would, I'm guessing, print this out there and then the captain would sign it on the line there.

104
00:07:39,240 --> 00:07:40,820
Here you'll recognize these numbers.

105
00:07:40,860 --> 00:07:43,460
Who has a guess as to what these codes are?

106
00:07:47,420 --> 00:07:47,820
AKE91536QF.

107
00:07:48,180 --> 00:07:48,800
Any guesses?

108
00:07:50,540 --> 00:07:51,960
I don't think they're TAL numbers.

109
00:07:52,040 --> 00:07:52,800
I think they're something else.

110
00:07:54,880 --> 00:07:55,760
VOR waypoints?

111
00:07:55,900 --> 00:07:57,200
No, not VOR waypoints.

112
00:07:57,340 --> 00:08:03,640
And I might be wrong here, but I think that they're the codes for the luggage cargo containers.

113
00:08:03,940 --> 00:08:05,700
Because if you look on the side, they're all uniquely coded.

114
00:08:05,940 --> 00:08:10,700
So I think these are some of the containers that they have in the belly of the plane.

115
00:08:12,260 --> 00:08:13,240
Here there's an alert.

116
00:08:13,320 --> 00:08:16,620
The gross takeoff weight exceeds some threshold.

117
00:08:18,020 --> 00:08:20,980
You know, again, routing information, reduce the delay.

118
00:08:21,820 --> 00:08:22,800
And this is interesting.

119
00:08:22,960 --> 00:08:24,620
I called Gateway and had them bring the fuel up.

120
00:08:24,700 --> 00:08:28,720
And they brought it up to some value to take into consideration the point through fuel burn increase.

121
00:08:28,880 --> 00:08:31,520
So they're modeling different aspects of the flight.

122
00:08:33,640 --> 00:08:37,960
This is actually data regarding all sorts of different subsystems.

123
00:08:37,960 --> 00:08:45,140
So the auto brake and, you know, various other telemetry that's sent back to airline operations and manufacturers so they can keep track of the equipment.

124
00:08:45,320 --> 00:08:49,720
Because a lot of the stuff there, like the engines too, are all fly by the hour.

125
00:08:49,880 --> 00:08:54,040
So that's all sent back to Rolls Royce or other manufacturers.

126
00:08:54,700 --> 00:09:00,540
And then, you know, various things that are left to the discretion of the captain as to, you know, takeoff or what have you.

127
00:09:01,120 --> 00:09:02,340
So that's all well and good.

128
00:09:02,880 --> 00:09:07,340
And I also spoke previously about decoding all eight channels at the same time.

129
00:09:07,880 --> 00:09:12,680
And once you do that, then you can actually amass quite a lot of information.

130
00:09:12,940 --> 00:09:14,900
And you might not want to read all that stuff.

131
00:09:15,140 --> 00:09:23,320
But maybe you want to actually have a look at some other stuff, which is what was just pointed out here, the waypoints that are part of the flight plans.

132
00:09:23,480 --> 00:09:24,620
Because if you look at this.

133
00:09:24,620 --> 00:09:26,040
You actually see there.

134
00:09:26,100 --> 00:09:28,760
There are north and west coordinates, latitude, longitude.

135
00:09:29,020 --> 00:09:33,580
And then there are these five letter character combinations.

136
00:09:34,100 --> 00:09:37,200
Anybody have any ideas what these five character codes are?

137
00:09:42,000 --> 00:09:42,780
That's right.

138
00:09:42,900 --> 00:09:50,260
They are navigational waypoints on aviation navigation charts that are established around the world.

139
00:09:50,440 --> 00:09:54,460
So they often don't actually send the raw latitude, longitude coordinates.

140
00:09:54,460 --> 00:09:58,280
They just summarize it with these five letter codes because that's well defined.

141
00:09:58,600 --> 00:10:02,780
And so what you can do there is parse that and plot that.

142
00:10:03,040 --> 00:10:08,960
So at any point when the blue arrow appears, I'm going to show you something outside the presentation.

143
00:10:09,420 --> 00:10:16,560
If for some reason I get caught up and forget about the blue arrow and it's there and I haven't shown you whatever I intended to, yell at me and I'll switch.

144
00:10:18,540 --> 00:10:21,580
So you can amass all that data and then parse it.

145
00:10:21,580 --> 00:10:25,000
And then get something like this.

146
00:10:27,080 --> 00:10:32,240
So I parse all that information and then you can convert it into KML for displaying Google Earth.

147
00:10:33,180 --> 00:10:38,940
And the satellite that I was listening to is actually servicing the Pacific Ocean area.

148
00:10:39,200 --> 00:10:45,440
And so you would expect that most of the information flowing through will be regarding planes flying over the Pacific.

149
00:10:45,700 --> 00:10:49,460
And so if you zoom out, you can kind of almost get a feel for the satellite footprint.

150
00:10:49,860 --> 00:10:51,560
And these are all the flight plans.

151
00:10:51,580 --> 00:10:58,620
These are all the flight plans that have been sent to all of the aircraft flying through.

152
00:10:58,740 --> 00:11:00,860
So it's particularly busy over Japan, say.

153
00:11:01,000 --> 00:11:06,940
And if you zoom in a little bit more, then you get all the five character codes coming up.

154
00:11:07,100 --> 00:11:13,720
And you can see by here, I guess this is the registration number for the aircraft.

155
00:11:14,180 --> 00:11:19,900
And so you can see what flight plan that was actually taking.

156
00:11:19,900 --> 00:11:25,900
I think the red dots are the dots that are actually sent with the raw latitude and longitude coordinates.

157
00:11:26,780 --> 00:11:28,040
And you can parse that out.

158
00:11:28,100 --> 00:11:30,280
So just to show you what that looks like in the raw.

159
00:11:37,680 --> 00:11:44,200
The output of the GNU Radio Decoder, it just dumps all this raw information.

160
00:11:44,880 --> 00:11:47,860
And then you have the frequency here because there are eight channels.

161
00:11:47,860 --> 00:11:55,180
And so if you just parse that a little bit, you can get those sorts of excerpts that I showed you before.

162
00:11:55,460 --> 00:11:58,720
And what I did was I just Googled these five letter codes.

163
00:11:59,140 --> 00:12:04,960
And then through various matches online, you find various listings for these for different areas.

164
00:12:05,200 --> 00:12:10,260
So for Russian airspace, there was this kind of text listing.

165
00:12:10,440 --> 00:12:12,320
So I just throw that in a Python file.

166
00:12:13,180 --> 00:12:16,540
And then at the bottom, I just have it go through and parse that.

167
00:12:16,620 --> 00:12:17,840
And it builds up a list of things.

168
00:12:17,840 --> 00:12:18,320
It's a list of waypoints.

169
00:12:18,440 --> 00:12:19,440
Same for Hong Kong, Japan.

170
00:12:20,240 --> 00:12:23,360
Taiwan actually had a JavaScript listing.

171
00:12:23,700 --> 00:12:27,680
So I did a find and replace to make it parsable in Python.

172
00:12:27,880 --> 00:12:32,960
And then build sort of a self-reflexive object to parse all that.

173
00:12:33,280 --> 00:12:37,200
And then the waypoints file just includes all the points.

174
00:12:37,420 --> 00:12:46,600
And then the parser takes the output of this file and then looks for all those coordinates in five-digit, five-character waypoints.

175
00:12:46,600 --> 00:12:47,980
And then just spits out the KML.

176
00:12:48,500 --> 00:12:53,240
So it's kind of interesting to see the system actually working and visualized that way.

177
00:12:54,260 --> 00:12:56,480
So that was the update on InMars.

178
00:12:56,500 --> 00:13:00,240
So any questions about that before I move on to the next thing?

179
00:13:01,140 --> 00:13:01,660
No?

180
00:13:03,360 --> 00:13:03,880
Okay.

181
00:13:07,680 --> 00:13:13,240
So the next thing is actually also aviation related.

182
00:13:14,800 --> 00:13:15,600
It's an implementation.

183
00:13:15,600 --> 00:13:20,280
Implementation of what's been termed as an unselective AM receiver.

184
00:13:20,720 --> 00:13:28,000
And I have to give full credit to Kevin Reed who at Cyberspectrum number 15 demonstrated this with his shiny SDR platform.

185
00:13:28,720 --> 00:13:39,920
Which is a really neat web-based SDR receiver so that you can use it remotely from the actual hardware.

186
00:13:40,460 --> 00:13:44,400
And what he did and he's documented this on his blog there which I encourage you to have a look at.

187
00:13:44,400 --> 00:13:51,800
Is do a really nifty sort of reversal of the common AM demodulator concept.

188
00:13:52,620 --> 00:13:54,680
So switchborg on Twitter.

189
00:13:55,600 --> 00:14:00,340
So with a normal AM demodulator if you want to listen to aircraft.

190
00:14:00,560 --> 00:14:04,300
Listen to interactions between pilots and the tower.

191
00:14:06,320 --> 00:14:08,700
You know for takeoff or approach taxiing.

192
00:14:08,700 --> 00:14:10,000
If you're a bit of a plane spotter.

193
00:14:10,100 --> 00:14:11,640
Then usually you have a receiver.

194
00:14:12,100 --> 00:14:13,620
So a software defined radio.

195
00:14:13,620 --> 00:14:18,240
And with that thick arrow there that's receiving a very large chunk of the spectrum.

196
00:14:18,420 --> 00:14:21,620
So say you might be receiving I don't know 10 megahertz worth.

197
00:14:21,820 --> 00:14:25,840
So you can get a nice big chunk of the arrow spectrum and see all the little transmissions there.

198
00:14:26,440 --> 00:14:29,060
And then you select a channel to listen to.

199
00:14:29,620 --> 00:14:32,600
And so that gets down sampled and filtered.

200
00:14:34,440 --> 00:14:36,940
And then you go complex to mag.

201
00:14:38,160 --> 00:14:40,120
Actually that should be a thin arrow there.

202
00:14:40,180 --> 00:14:40,860
I didn't fix that.

203
00:14:42,140 --> 00:14:43,500
And then what happens is.

204
00:14:43,620 --> 00:14:45,500
Once you do the AMD mod.

205
00:14:45,620 --> 00:14:49,700
Which is really this step here of complex to magnitude.

206
00:14:50,680 --> 00:14:52,640
Then you output that to your speaker.

207
00:14:52,700 --> 00:14:53,660
And you hear the audio.

208
00:14:54,100 --> 00:14:55,340
And what Kevin Reed did.

209
00:14:55,640 --> 00:14:57,740
Was that he swapped these two here.

210
00:14:58,160 --> 00:15:06,320
So what he did was he swapped the Dan sampling and filtering.

211
00:15:06,720 --> 00:15:08,140
With the complex to mag.

212
00:15:08,980 --> 00:15:10,480
And so what that means is.

213
00:15:10,620 --> 00:15:12,600
You essentially take the magnitude.

214
00:15:12,600 --> 00:15:15,320
Of your entire AM spectrum that you're receiving.

215
00:15:15,420 --> 00:15:16,400
Like 10 megahertz worth.

216
00:15:16,900 --> 00:15:21,340
And then you down sample and filter the entire thing.

217
00:15:21,520 --> 00:15:25,340
So you never actually select a channel to down sample to.

218
00:15:25,380 --> 00:15:26,980
You demodulate the entire band.

219
00:15:27,320 --> 00:15:28,580
But because it's AM.

220
00:15:28,920 --> 00:15:30,380
The side effect is.

221
00:15:30,520 --> 00:15:34,520
You end up hearing the strongest transmitter in the band.

222
00:15:34,780 --> 00:15:35,720
Which is really neat.

223
00:15:37,700 --> 00:15:38,620
Thanks for that.

224
00:15:39,720 --> 00:15:40,320
.

225
00:15:41,800 --> 00:15:42,400
.

226
00:15:42,400 --> 00:15:42,580
.

227
00:15:42,580 --> 00:15:43,300
.

228
00:15:44,120 --> 00:15:44,560
.

229
00:15:44,560 --> 00:15:44,800
Thanks, Ryan.

230
00:15:45,280 --> 00:15:45,880
There we go.

231
00:15:46,440 --> 00:15:48,420
So the idea here is that.

232
00:15:49,400 --> 00:15:51,360
You end up hearing the strongest transmission.

233
00:15:51,560 --> 00:15:52,120
And this works.

234
00:15:52,380 --> 00:15:53,560
Because it's AM, right?

235
00:15:53,680 --> 00:15:54,620
You're demodulating AM.

236
00:15:54,740 --> 00:15:57,360
Because you're simply looking at the power level of your carrier wave.

237
00:15:57,620 --> 00:16:00,320
And if you treat the entire spectrum as your carrier wave.

238
00:16:00,680 --> 00:16:04,120
Then the strongest signal there will cause the largest ripples.

239
00:16:04,240 --> 00:16:05,420
And that's what you'll end up hearing.

240
00:16:07,540 --> 00:16:09,180
So he swapped them like this.

241
00:16:09,360 --> 00:16:11,580
And this is actually really easy to implement.

242
00:16:11,580 --> 00:16:19,660
to implement he did something that was really clever which was take the receive output the

243
00:16:19,660 --> 00:16:27,320
raw complex base band and then apply a slope filter to the spectrum so one path you take out

244
00:16:27,320 --> 00:16:33,700
and you say i want to have the left hand side of the spectrum louder than the right and the other

245
00:16:33,700 --> 00:16:38,900
one you want the right side of the spectrum to be louder than the left and then once you've shaped

246
00:16:38,900 --> 00:16:44,820
your spectrum like that you do the same process so you you do complex to mag or amd modulate the

247
00:16:44,820 --> 00:16:53,920
entire entire spectrum down sample and then you play that into the left speaker and the right

248
00:16:53,920 --> 00:16:59,220
speaker respectively so what ends up happening is stuff that's on the left hand side of the

249
00:16:59,220 --> 00:17:03,140
spectrum you end up hearing out of the left speaker and stuff that's on the right side of

250
00:17:03,140 --> 00:17:07,460
the spectrum you hear out of the rights because so it spatializes the entire am spectrum for you

251
00:17:08,080 --> 00:17:08,560
and

252
00:17:08,900 --> 00:17:13,600
if you're wearing headphones it sounds really cool i've got some speakers here so in a moment i might

253
00:17:13,600 --> 00:17:17,340
might ask for absolute quiet because they're not very loud and they need to be stereo

254
00:17:18,020 --> 00:17:22,400
but i'm going to show you a video here and then i'll show you it interactively

255
00:17:23,080 --> 00:17:24,640
now just before i show you the video

256
00:17:26,980 --> 00:17:31,000
i'll just explain what's going on here so this is the flow graph and this is the whole spectrum

257
00:17:31,680 --> 00:17:36,180
and this is a representation of the frequency response of those

258
00:17:37,080 --> 00:17:38,820
of those shaping filters so

259
00:17:38,900 --> 00:17:43,820
this is not this is the raw baseband that's coming in and this is this is that that shaping and so

260
00:17:43,820 --> 00:17:48,880
there's a blue one and a right and a green one that's the left and the right channel now the

261
00:17:48,880 --> 00:17:54,100
reason why it has these notches in there and you see those notches reflected on the baseband

262
00:17:54,100 --> 00:17:59,960
is because often you you've got birdies and or spurs which are either internal to the receiver

263
00:17:59,960 --> 00:18:04,580
or from from you know man-made noise or from vors which are always transmitting something

264
00:18:05,360 --> 00:18:08,760
or just you know other things that you don't want to listen to and

265
00:18:08,900 --> 00:18:13,100
sometimes they can be powerful and they will drown out legitimate transmissions that you want to listen to

266
00:18:13,100 --> 00:18:19,260
and so i thought it would be cool if you could interactively notch out those frequencies that you didn't want to listen to

267
00:18:19,260 --> 00:18:25,540
and so what this does is and i'll show you you can click on the baseband spectrum on the things that you want to notch out

268
00:18:25,540 --> 00:18:32,020
it'll recompute the filter with those notches included and then apply that to the left and right slopes

269
00:18:32,680 --> 00:18:38,280
and it does it globally so you can change the frequency of your sdr and then it'll actually

270
00:18:38,900 --> 00:18:43,260
compute everything that's in your passband and always keep the stuff out that you don't want to listen to

271
00:18:43,260 --> 00:18:51,100
so if if i could if i have as quiet as you can i'll turn the volume up here so you can hopefully hear it

272
00:18:53,220 --> 00:18:55,040
and so i'm just gonna oh there you go

273
00:18:59,680 --> 00:19:01,980
so notice i clicked on that and the tone went away

274
00:19:07,380 --> 00:19:08,880
i don't know if you can get the sound out of it

275
00:19:08,900 --> 00:19:10,000
i don't know if you can get the stereo effect hopefully you can

276
00:19:12,760 --> 00:19:15,560
so see i clicked there it applied the notch and then

277
00:19:18,080 --> 00:19:25,900
recomputed that and i'm changing the baseband frequency and so it's it's shifting all the notches along to keep them on on the right frequency

278
00:19:30,920 --> 00:19:35,980
and so i moved to to where there's eight cars and it's very loud and powerful you don't want to listen to it so i notch that out

279
00:19:36,520 --> 00:19:37,500
notch that out too

280
00:19:50,320 --> 00:19:54,320
so it's cool right because i'm not actually clicking on the channel that i want to listen to

281
00:19:54,320 --> 00:19:57,380
it just all comes in and you hear the the most powerful thing

282
00:20:02,040 --> 00:20:05,700
and then what's nice is that you can click on remove if you want to remove the notch

283
00:20:05,700 --> 00:20:08,800
and then click on the the the bass bands

284
00:20:08,900 --> 00:20:14,720
and then it'll actually remove that notch recompute the filter and then you get you know in this case the unwanted tone back

285
00:20:16,540 --> 00:20:23,080
so so that's that now the thing is i mentioned that that slope filter and it wasn't immediately

286
00:20:23,080 --> 00:20:28,880
obvious to me how you you generate that and it turns out it's actually really really easy

287
00:20:28,880 --> 00:20:34,700
this is done in a single line of python in a grc block that i'll show you basically you use numpy

288
00:20:35,280 --> 00:20:38,720
and use linspace to create a ramp between zero and one

289
00:20:38,900 --> 00:21:07,280
over the number of uh taps that you want you shift the do an fft shift so it basically rearranges you know one half and the other half back to back you do an inverse for your transform and then you shift it back so what you've done is essentially you say i want this frequency response and then you take the inverse fft and then it gives you the time domain taps that you can apply in a normal fir filter and that that's all it is it's it's really basic

290
00:21:08,900 --> 00:21:12,600
so i've got a blue arrow which means show you the actual thing

291
00:21:14,040 --> 00:21:19,900
now any any questions so far who actually listens to the aero band and listens to

292
00:21:20,440 --> 00:21:25,000
cockpit and tower and all that kind of stuff a few of you all right was there a question over there

293
00:21:29,200 --> 00:21:32,600
just curious with the interactive notch you could do equalization

294
00:21:32,600 --> 00:21:38,140
to null out tones that are continuous but is that too computationally intensive across that

295
00:21:38,900 --> 00:21:45,080
light spectrum so um that's a very good question so i don't have to repeat the question because

296
00:21:45,080 --> 00:21:54,180
he's spoken in the mic usually um yeah so if you remember on that line when it was using a linspace

297
00:21:54,180 --> 00:22:00,660
and have the taps count um you can decrease the number of taps if you're using a higher bandwidth

298
00:22:00,660 --> 00:22:05,200
because naturally yeah the larger bandwidth the like this thing works really hard i think i do 6.25

299
00:22:05,700 --> 00:22:08,280
megahertz off of usurp um

300
00:22:08,900 --> 00:22:13,660
and it it makes it work pretty hard and i i chose a higher tap count so that it would look really

301
00:22:13,660 --> 00:22:18,620
good in the frequency response plot but if you if you make that much smaller it looks far more ugly

302
00:22:18,620 --> 00:22:25,380
the response is obviously rippled but it works okay so you can manage it that way uh so let me

303
00:22:25,380 --> 00:22:31,680
find the flow graph here i'm gonna have a resolution problem so let me just transfer this

304
00:22:34,220 --> 00:22:36,220
make the uh thing fit

305
00:22:41,760 --> 00:22:45,720
all right so i'm looking for am unselected and

306
00:22:46,280 --> 00:22:52,200
turn it off so i'm i'm going to just do a live demo or not i've pre-recorded

307
00:22:52,200 --> 00:23:07,820
so that's it's that and it's that so at the moment i've got a single notch in there

308
00:23:08,900 --> 00:23:10,560
and if i take it out

309
00:23:15,900 --> 00:23:19,520
you can you can hear the tone right you don't want that so let's

310
00:23:20,020 --> 00:23:25,000
and if you're watching the console there um i've got this notches.json that contains the

311
00:23:25,000 --> 00:23:34,440
list so if i click this then it'll load the load the notches calculate the a complex um

312
00:23:36,100 --> 00:23:37,160
uh what did i do

313
00:23:37,160 --> 00:23:41,740
uh a complex filter and then rotate it up to the frequency the baseband frequency that i

314
00:23:41,740 --> 00:23:48,800
selected and then apply that here and what's nice is that you can also change the number of taps

315
00:23:49,520 --> 00:23:55,700
so if i made it 13 instead of 512 then obviously it changes the

316
00:23:57,780 --> 00:24:04,940
the look of your frequency response and one other thing um ian my friend ian buckley observed this

317
00:24:04,940 --> 00:24:07,020
that if you actually add a delay

318
00:24:07,160 --> 00:24:11,180
between the left and the right channel especially if you're on headphones that it makes it

319
00:24:11,180 --> 00:24:16,240
it emphasizes the stereo effect even more i don't know whether we'll necessarily be able to tell here

320
00:24:16,760 --> 00:24:18,380
but if we make i think this was like

321
00:24:22,440 --> 00:24:27,720
yeah you'll need to play with this yourself but you can add a delay to emphasize that

322
00:24:27,720 --> 00:24:32,240
and you can you know mute the left with the right and if you look at the um

323
00:24:33,900 --> 00:24:35,280
the actual response

324
00:24:36,500 --> 00:24:43,480
of um each the output of each filter you can see the baseband uh plus the filter showing that

325
00:24:43,480 --> 00:24:48,720
characteristic you know roll off to one side and the other one and the green lines that are appearing

326
00:24:48,720 --> 00:24:53,700
are just basically selecting the the signal the fft bin that has the strongest energy unit so that

327
00:24:53,700 --> 00:24:58,620
that that's most likely the the audio that you're hearing in that frequency which is being updated at

328
00:24:58,620 --> 00:25:05,260
the bottom there max frequency right then left so again this is just the implementation as well as the

329
00:25:05,260 --> 00:25:10,600
The original idea was Kevin's, but it's, you know, I thought it might be fun to implement

330
00:25:10,600 --> 00:25:11,200
in Gudu Radio.

331
00:25:12,260 --> 00:25:14,440
I can take you through the flow graph very quickly.

332
00:25:15,780 --> 00:25:22,180
I've got a file source here, and then you get your complex baseband output, and as I

333
00:25:22,180 --> 00:25:30,620
mentioned, you, oops, you take one line up to one filter that contains the right channel

334
00:25:30,620 --> 00:25:35,360
taps, and then you take the baseband to the other filter that has the left channel taps,

335
00:25:35,560 --> 00:25:41,500
and what's nice, actually, is that to compute the, you only need to compute one set of taps

336
00:25:41,500 --> 00:25:48,100
because the other channel is just the same taps, but in reverse, which is handy.

337
00:25:49,100 --> 00:25:54,240
And then once you have done the filtering, you immediately do the complex to mag, which

338
00:25:54,240 --> 00:25:59,900
is the AMD mod step, and then you resample it to match the sample.

339
00:26:00,620 --> 00:26:01,660
The sample rate of your sound card.

340
00:26:02,020 --> 00:26:06,500
I've got a DC block in there as well to remove any DC from the AMD modulated signal.

341
00:26:06,720 --> 00:26:11,880
I've got an AGC so that the audio level that comes up the speaker is always more or less

342
00:26:12,660 --> 00:26:13,060
normalized.

343
00:26:13,860 --> 00:26:19,820
A low pass filter so that you, you know, AM is quite narrow band here, so just to focus

344
00:26:19,820 --> 00:26:26,580
on that one channel at the audio pass band, and then it just goes into one channel of

345
00:26:26,580 --> 00:26:28,280
the audio sync, and that's all there is to it.

346
00:26:30,620 --> 00:26:33,340
Additional complexity happens more with that, the interactive thing.

347
00:26:35,680 --> 00:26:40,480
So, the graph there is brought about by, what have we got here?

348
00:26:42,540 --> 00:26:46,920
There's this sort of real-time graph in that plot lib wrapper I made, and so using an

349
00:26:46,920 --> 00:26:51,440
any code block, you just implement the graph like that, and it will cause it to display.

350
00:26:52,260 --> 00:26:59,920
And then every time, every time here, the variable dependencies,

351
00:27:00,620 --> 00:27:01,960
taps left or taps right changes.

352
00:27:02,080 --> 00:27:05,900
This line means create a callback for if those variables change.

353
00:27:06,160 --> 00:27:13,500
Then it does this Python, actually this is multiple lines, but separated by semicolons.

354
00:27:13,700 --> 00:27:21,260
So the first part of it is use SciPy to calculate the frequency response of the left taps, and

355
00:27:21,260 --> 00:27:26,360
then calculate the frequency response of the right taps, and then set the data on the real-time

356
00:27:26,360 --> 00:27:29,520
graph with the FFT shifted.

357
00:27:30,620 --> 00:27:35,760
Log 10 magnitude of the frequency response.

358
00:27:36,220 --> 00:27:39,700
And so, that line, whenever the taps change, will update the graph.

359
00:27:40,300 --> 00:27:48,100
And then in terms of actually clicking on stuff, there's a callback mechanism here.

360
00:27:48,920 --> 00:27:57,660
You know, with these FFT GUI syncs, you can specify a variable to take the frequency that

361
00:27:57,660 --> 00:27:58,180
you clicked on.

362
00:27:59,220 --> 00:28:00,600
So, when that variable changes.

363
00:28:00,600 --> 00:28:08,120
When it changes over here, then I have a little bit of Python that, it's a class, and it has

364
00:28:08,120 --> 00:28:11,140
this add notch function in it, and so it takes the click frequency.

365
00:28:11,480 --> 00:28:15,880
And if remove is disabled, then it adds the notch.

366
00:28:16,080 --> 00:28:19,600
And if remove is enabled, then it removes the notch at that frequency.

367
00:28:20,060 --> 00:28:24,520
And so, there's just a little bit of Python that manages the global database of values.

368
00:28:25,140 --> 00:28:30,580
And once that happens, I think it's, it's done.

369
00:28:30,580 --> 00:28:30,840
It's here.

370
00:28:31,060 --> 00:28:39,280
So, the tap generator class is initialized here, and it basically takes the initial

371
00:28:39,880 --> 00:28:45,820
set of taps, which is just that initial, you know, left and right slope.

372
00:28:46,080 --> 00:28:54,180
And then self.taps is, I think, a reference to the actual taps that the generator will

373
00:28:54,180 --> 00:28:54,740
update.

374
00:28:54,920 --> 00:28:59,240
So, the generator will generate new taps and then call a callback registered within GRC

375
00:28:59,240 --> 00:29:00,060
to update the taps.

376
00:29:00,580 --> 00:29:03,360
And then it, you know, will update in the MFT filter and update on the graph.

377
00:29:04,600 --> 00:29:05,680
And that's it, really.

378
00:29:10,180 --> 00:29:13,460
So, conceptually at least, not, you know, it's pretty elegant.

379
00:29:13,680 --> 00:29:18,280
And again, that's, props to Kevin for figuring that out.

380
00:29:18,480 --> 00:29:21,340
Any questions on any of that so far?

381
00:29:22,340 --> 00:29:23,640
Does it, okay.

382
00:29:29,660 --> 00:29:30,220
All right.

383
00:29:30,580 --> 00:29:37,620
So, next on the agenda, and this is the third and final part, is radar, and particularly

384
00:29:37,620 --> 00:29:38,540
FMCW radar.

385
00:29:38,840 --> 00:29:43,380
I spoke about, a little bit about this last year, but I'd like to show you some more and

386
00:29:43,380 --> 00:29:45,000
explain some more concepts to you as well.

387
00:29:46,240 --> 00:29:50,940
And to do a bit of a recap, who's seen these primary surveillance radars at airports when

388
00:29:50,940 --> 00:29:51,240
you fly?

389
00:29:51,400 --> 00:29:52,480
You know, you see them rotating there?

390
00:29:52,900 --> 00:29:56,480
This is a basic kind of CW radar.

391
00:29:57,460 --> 00:30:00,560
And if you watch here, as it rotates, I'm, I'm going to show you this.

392
00:30:00,580 --> 00:30:02,900
I'm sitting on a hill with a usurp on the frequency.

393
00:30:03,360 --> 00:30:07,720
And on the left-hand side, it's plotting the magnitude of the signal received.

394
00:30:08,020 --> 00:30:16,000
And you can see that it's triggering, play it again, it's triggering on the initial transmission

395
00:30:16,000 --> 00:30:16,700
from the radar.

396
00:30:16,820 --> 00:30:20,580
And the radar then switches to receive mode to listen to any echoes that are coming back

397
00:30:20,580 --> 00:30:22,460
from, in this case, airplanes.

398
00:30:23,960 --> 00:30:29,940
And you can see there that there's also some responses coming back very close in time.

399
00:30:30,580 --> 00:30:35,920
So, I'm picking up both the initial burst of the radar and what turns out to be ground

400
00:30:35,920 --> 00:30:37,100
clutter, mainly nearby.

401
00:30:38,340 --> 00:30:41,760
And so, what is a radar return and how does that work?

402
00:30:41,940 --> 00:30:45,040
So, this is an example return down the bottom.

403
00:30:45,420 --> 00:30:49,720
Just complex IQ that's been, again, turned to AM, essentially, the magnitude.

404
00:30:52,060 --> 00:30:55,940
And the way it works is that the radar sends out what's called a bang, which is that initial

405
00:30:55,940 --> 00:30:56,920
short pulse.

406
00:30:57,280 --> 00:30:58,660
And in this case, it's CW.

407
00:30:58,660 --> 00:31:00,560
So, it's just a continuous pulse.

408
00:31:00,560 --> 00:31:02,080
It's a continuous wave at a single frequency.

409
00:31:02,420 --> 00:31:08,600
And then the radar switches to receive very quickly and then it waits for these echoes

410
00:31:08,600 --> 00:31:13,780
to come back because the radio energy will reflect off, for example, aircraft fuselages.

411
00:31:13,940 --> 00:31:18,740
And some of that energy, just a small portion, will come back and be collected by that massive

412
00:31:18,740 --> 00:31:21,020
dish that's rotating on the radar.

413
00:31:21,160 --> 00:31:24,120
And it's so massive because the signals are so weak and so you want a lot of gain at the

414
00:31:24,120 --> 00:31:24,380
antenna.

415
00:31:25,840 --> 00:31:26,360
And...

416
00:31:27,020 --> 00:31:27,380
Nope.

417
00:31:29,720 --> 00:31:30,540
And then...

418
00:31:30,540 --> 00:31:35,180
And then once that period has elapsed, which is the pulse repetition period, then a new

419
00:31:35,180 --> 00:31:36,040
pulse is sent out.

420
00:31:36,200 --> 00:31:38,220
And then the process just continues.

421
00:31:38,500 --> 00:31:41,820
And this can happen hundreds or thousands of times per second.

422
00:31:46,280 --> 00:31:50,420
And there is some radar concepts which are important to get a hold of.

423
00:31:50,460 --> 00:31:52,720
One is that pulse repetition interval of frequency.

424
00:31:52,920 --> 00:31:57,260
So, that's the amount of time or how many times per second that pulse is sent out.

425
00:31:57,340 --> 00:31:58,740
And then the wait period.

426
00:31:59,460 --> 00:32:00,520
I'm going to call it period.

427
00:32:00,540 --> 00:32:02,800
PRF, which is the number of times it happens per second.

428
00:32:05,140 --> 00:32:08,460
And the width of the actual transmitted pulse.

429
00:32:08,640 --> 00:32:10,960
So for how long the radar is actually transmitting.

430
00:32:11,740 --> 00:32:13,800
That's also important.

431
00:32:15,240 --> 00:32:21,180
So the way you actually figure out how far an object is away from you is by looking at

432
00:32:21,180 --> 00:32:26,160
the delay between when the signal went out and then when you had the signal come back.

433
00:32:26,580 --> 00:32:28,200
And this is called the round trip time.

434
00:32:29,380 --> 00:32:30,520
I'll explain that a little bit later.

435
00:32:30,520 --> 00:32:31,800
I'll explain that in a little bit more detail in a moment.

436
00:32:32,060 --> 00:32:38,020
But if you consider these two diagrams here from RF Cafe, these thick black lines here

437
00:32:38,020 --> 00:32:39,840
are when the transmitted pulse was sent out.

438
00:32:40,640 --> 00:32:42,920
And then you have this return here that comes back.

439
00:32:43,320 --> 00:32:47,620
And there's this notion of unambiguous and ambiguous returns in radar.

440
00:32:47,880 --> 00:32:56,240
So in this diagram, the A diagram, you have a return that occurs before the next transmitted

441
00:32:56,240 --> 00:32:56,500
pulse.

442
00:32:57,120 --> 00:33:00,500
And you can then say that this is an unambiguous return.

443
00:33:00,520 --> 00:33:03,320
Because it happens before the next pulse.

444
00:33:04,300 --> 00:33:10,020
In the B diagram, the real range actually occurs after the second pulse.

445
00:33:10,160 --> 00:33:13,380
So that's in excess of that interval there.

446
00:33:13,820 --> 00:33:18,240
And so if you didn't know any better, you would assume that this target is actually

447
00:33:18,240 --> 00:33:19,520
really close to your radar.

448
00:33:19,700 --> 00:33:22,120
Where in actual fact it's even further away.

449
00:33:22,340 --> 00:33:24,220
And so you get that sort of false return.

450
00:33:24,520 --> 00:33:26,740
And this is ambiguous range.

451
00:33:27,100 --> 00:33:30,440
And so what's important to note is that the way you design the parameters.

452
00:33:30,440 --> 00:33:36,000
The parameters of your radar will constrain you and will set the unambiguous range that

453
00:33:36,000 --> 00:33:36,640
you can compute.

454
00:33:37,380 --> 00:33:41,180
And so the easy way to think about it is that the round trip time times the speed of

455
00:33:41,180 --> 00:33:43,600
light divided by two is your virtual range.

456
00:33:43,840 --> 00:33:44,660
It's not the real range.

457
00:33:44,800 --> 00:33:45,680
It's the virtual range.

458
00:33:47,780 --> 00:33:49,660
So just keep that in mind.

459
00:33:50,160 --> 00:33:54,880
Now with that radar that you saw spinning, the primary spin radar, I did a recording

460
00:33:54,880 --> 00:33:58,180
and then each of these scan lines is triggered.

461
00:33:58,320 --> 00:34:00,340
There's this red part on the left-hand side.

462
00:34:00,340 --> 00:34:06,020
When it hears a really loud burst, which is attributed to the radar doing its transmission.

463
00:34:06,500 --> 00:34:10,480
And then it keeps listening for a fixed period of time for the return.

464
00:34:10,740 --> 00:34:15,440
So this is all, you know, me sitting on the hillside with my little USOP with just a width

465
00:34:15,440 --> 00:34:15,660
antenna.

466
00:34:15,880 --> 00:34:16,980
Not a big radar dish.

467
00:34:17,840 --> 00:34:22,600
And then it captures a fixed number of samples and then it waits for the next radar transmission

468
00:34:22,600 --> 00:34:23,220
as it rotates.

469
00:34:23,740 --> 00:34:27,780
And what amazed me is that you can actually still pick up with a little width antenna

470
00:34:27,780 --> 00:34:30,320
all of these structures that are a little bit smaller.

471
00:34:30,320 --> 00:34:33,900
They're a little bit further out from the ground clutter, which is causing these immediate

472
00:34:33,900 --> 00:34:35,140
returns on the left-hand side.

473
00:34:36,240 --> 00:34:43,120
And so this is just each scan line and then I just recorded it for a number of rotations.

474
00:34:43,320 --> 00:34:44,760
I think this is actually just a single rotation.

475
00:34:45,240 --> 00:34:48,300
This is 50 mega samples per second capture into RAM disk.

476
00:34:49,200 --> 00:34:56,120
And what's nice is that taking that linear plot and if you unwrap it into polar space

477
00:34:56,540 --> 00:34:59,400
and place it on the radar position itself.

478
00:34:59,480 --> 00:35:00,280
This is the Bay Area.

479
00:35:00,320 --> 00:35:05,940
You can see that the ground clutter actually lines up quite neatly with, you know, the

480
00:35:05,940 --> 00:35:06,940
ground and not in the bay.

481
00:35:07,120 --> 00:35:14,000
What blew my mind is that you actually get returns from the power pylons that cross

482
00:35:14,000 --> 00:35:15,820
the bay there and also from the bridge.

483
00:35:15,980 --> 00:35:20,580
So these large structures return enough energy to be received by just a width antenna on

484
00:35:20,580 --> 00:35:21,020
the USOP.

485
00:35:21,420 --> 00:35:25,440
So, you know, that's the simple concept of CW radar returns.

486
00:35:26,600 --> 00:35:30,080
And a bit of theory here from Wikipedia.

487
00:35:30,100 --> 00:35:30,300
So, you know, I've got a couple of things here.

488
00:35:30,320 --> 00:35:36,140
This is to illustrate why CW radar is not the best kind of radar.

489
00:35:36,340 --> 00:35:37,200
There are better forms.

490
00:35:37,500 --> 00:35:45,180
So, again, this is the plot where you have time and the red CW tone here, brief one is

491
00:35:45,180 --> 00:35:46,320
what's transmitted by the radar.

492
00:35:46,620 --> 00:35:50,420
And these blue patches here are the echoes that come back.

493
00:35:50,780 --> 00:35:58,280
And so if you take a match filter, which is in this case just a filter for the frequency

494
00:35:58,660 --> 00:35:59,860
that was sent out.

495
00:36:01,880 --> 00:36:04,100
Then that means you can ignore everything else.

496
00:36:04,380 --> 00:36:09,200
And if you apply the filter, then you get this plot on the right.

497
00:36:09,380 --> 00:36:10,740
So you ignore the transmitted pulse.

498
00:36:11,200 --> 00:36:16,220
And then the peak of the response here is at the time where that target was.

499
00:36:16,520 --> 00:36:18,160
And then this other target is further out.

500
00:36:18,240 --> 00:36:22,400
And so you get a lower amplitude response, but, again, you get the peak there.

501
00:36:22,580 --> 00:36:24,180
And this is not the best, right?

502
00:36:24,220 --> 00:36:26,440
Because you have this ramp up and ramp down.

503
00:36:26,720 --> 00:36:30,240
And you like to have a much more accurate definition in time.

504
00:36:30,240 --> 00:36:35,820
and the time domain of where that target response actually is. And the bigger issue is that

505
00:36:35,820 --> 00:36:41,260
if you have two targets that are close together like this, so those two targets are now not

506
00:36:41,260 --> 00:36:46,280
separated by as much distance and therefore time, then the response, the echo actually

507
00:36:46,280 --> 00:36:51,140
blows together. And so now it looks like there's only a single target if you were applying

508
00:36:51,140 --> 00:36:55,860
a naive detection strategy. So that's why CW has these drawbacks.

509
00:36:57,020 --> 00:37:03,880
And a simple alternative, and there are many other radar alternatives, is FMCW. And what

510
00:37:03,880 --> 00:37:07,940
that means is it's frequency modulated carry wave. So instead of just sending out a fixed

511
00:37:07,940 --> 00:37:13,780
carry wave, you actually send out a carry wave that increases in frequency or might

512
00:37:13,780 --> 00:37:17,580
decrease in frequency. And this is what it looks like in the time domain. So you start

513
00:37:17,580 --> 00:37:22,420
at a low frequency and then you quickly ramp up and it might be linear, it might not be,

514
00:37:22,480 --> 00:37:25,660
but you can ramp up. And by the end of your transmission, you're

515
00:37:25,860 --> 00:37:31,600
actually transmitting at high frequency. And the reason why FMCW is good is because

516
00:37:32,200 --> 00:37:35,840
these chirps, this is known as a chirp when you have something increasing or decreasing

517
00:37:35,840 --> 00:37:40,780
in frequency, it has the property of strong self-correlation. It's got a really good

518
00:37:40,780 --> 00:37:48,140
auto-correlation property. And what that means is if you take that signal and you mix

519
00:37:48,140 --> 00:37:51,860
it with a copy of itself, then you'll get a really good response. But if you mix it

520
00:37:51,860 --> 00:37:55,760
with anything else, then you'll get a really bad response, which is not the case

521
00:37:55,760 --> 00:38:03,000
for what we just saw with the CW tone, because you get this mixing here and the ambiguous

522
00:38:03,000 --> 00:38:15,120
response. So I have a demo for that. We're going to actually listen to what a chirp is

523
00:38:15,120 --> 00:38:25,080
so that you can get a better idea of what that is like.

524
00:38:28,860 --> 00:38:33,780
And you can see it. So this is, again, in the audio domain. I want to emphasize that

525
00:38:33,780 --> 00:38:39,240
these are general concepts and these are transferable between audio and radio and other things.

526
00:38:39,520 --> 00:38:45,320
So you'll hear the chirp and this is actually the response from my microphone in my laptop.

527
00:38:45,620 --> 00:38:49,240
So as I'm speaking, you can see there. Does anybody want to whistle or make a tone?

528
00:38:52,080 --> 00:38:55,740
There you go. Just to prove to you that it's live. And now I'm going to

529
00:38:55,740 --> 00:38:57,940
let you hear the tone and you'll see it on the spectrum.

530
00:39:02,840 --> 00:39:05,840
So you can't even hear that, but the microphone and the speaker still generate a response.

531
00:39:07,780 --> 00:39:13,360
Okay. So this is a chirp. It's a very slow chirp, but it's a chirp nonetheless. The radar

532
00:39:13,360 --> 00:39:17,120
chirps are much, much quicker, and I'll show you what that sounds like and looks like in

533
00:39:17,120 --> 00:39:25,620
a moment. So this is the frequency response. In the time domain, a common construction

534
00:39:25,740 --> 00:39:32,020
that you use is you use a sawtooth wave. So usually a signal source and you might have

535
00:39:32,020 --> 00:39:36,700
a sine wave. But here I'm using a sawtooth, and I hook that up to a VCO, to a voltage

536
00:39:36,700 --> 00:39:41,500
controlled oscillator. So what I'm effectively doing is I'm putting in a linear ramp, a

537
00:39:41,500 --> 00:39:46,540
linear voltage control, and then the VCO will turn that into a frequency modulated continuous

538
00:39:46,980 --> 00:39:53,920
wave. And we can see what that looks like here. So I'm going to speed up the chirp.

539
00:39:55,740 --> 00:40:04,720
Here, with the, where is it? Do I not have that exposed? Delay, multiplier, low pass

540
00:40:04,720 --> 00:40:10,100
federal. Ah, here. So this will take 10 seconds to do the chirp. I'm going to make it do it

541
00:40:10,100 --> 00:40:19,980
every second. Right? And I'm going to make it do it 10 times a second. Okay. Now, if

542
00:40:19,980 --> 00:40:25,520
you look at it in the time domain, this is the output of the sawtooth generator, right?

543
00:40:25,520 --> 00:40:29,520
So this is not the frequency output yet, the FMCW output. This is just controlling the

544
00:40:29,520 --> 00:40:35,080
VCO. Once you feed that through the VCO, it looks like this. So let me slow that down

545
00:40:35,080 --> 00:40:40,460
so you can see it a little bit better. This is the time domain plot of the complex output

546
00:40:40,460 --> 00:40:44,540
of the VCO block. And you can see that you get a complex sine wave, sine your side, but

547
00:40:44,540 --> 00:40:49,420
it's increasing in frequency and then it resets. So this is the basis of a chirp radar. We're

548
00:40:49,420 --> 00:40:53,560
listening to the chirp here. All right.

549
00:40:58,920 --> 00:41:04,000
So, again, this should be familiar to you now. We saw it in the complex domain, but

550
00:41:04,000 --> 00:41:09,440
here it's in the real domain. Now, in this case, we've been listening to a full duty

551
00:41:09,440 --> 00:41:14,180
cycle chirp generator, so it's constantly generating the chirp. And as soon as one

552
00:41:14,180 --> 00:41:19,960
finishes, the next one starts. In the radar scenario we looked at before, we sent out

553
00:41:19,960 --> 00:41:23,700
a chirp, we were quiet for a while, and then we sent out another chirp. You can do both.

554
00:41:23,860 --> 00:41:24,920
They're both equally valid.

555
00:41:25,520 --> 00:41:29,960
They just require different sorts of hardware configurations. I'm going to start talking

556
00:41:29,960 --> 00:41:38,340
about the full duty cycle continuous version now. What did I say there? Now, the other

557
00:41:38,340 --> 00:41:42,560
thing to remember here, and this is really important, too, is that when we talk about

558
00:41:42,560 --> 00:41:46,140
filters, filters need taps, right? Before with the arrow, you know, the left to right

559
00:41:46,140 --> 00:41:51,820
channel unselective receiver, we generated two series of taps to define two filters and

560
00:41:51,820 --> 00:41:55,500
they imposed their shape on the spectrum. Here, we've got two filters. We've got two

561
00:41:55,520 --> 00:42:00,560
filters. We're going to consider a filter whose taps are the chirp, okay? So usually

562
00:42:00,560 --> 00:42:04,280
there are low pass filters and high pass filters and you can think about how they might shape

563
00:42:04,280 --> 00:42:09,360
the spectrum. You can make a filter of whatever you want. You can even use filters in the

564
00:42:09,360 --> 00:42:14,360
digital domain for like the access control word or a preamble for a digital packet. Here,

565
00:42:14,540 --> 00:42:18,180
our filter is actually going to be the chirp itself. So the idea is that whenever it hears

566
00:42:18,180 --> 00:42:22,400
a chirp and then you put a filter through it, you will get the maximum response from

567
00:42:22,400 --> 00:42:25,500
your filter when the incoming chirp lines up with the filter.

568
00:42:25,520 --> 00:42:35,460
This is the chirp that we've defined in the filter. So, and this is what makes it, I think,

569
00:42:35,480 --> 00:42:41,440
really click. This is like a diagram we had before. So we have the transmitted chirp here

570
00:42:41,440 --> 00:42:45,200
and we have these two echoes, right? Because the echoes are going to be the reflection

571
00:42:45,200 --> 00:42:49,940
of what we transmitted. And because we're using this special filter that has this great

572
00:42:49,940 --> 00:42:55,440
autocorrelation property, we get a really nice well-defined peak here in time that lines

573
00:42:55,520 --> 00:43:01,400
up with the response, the exact response echo that came back here. So you get the

574
00:43:01,400 --> 00:43:06,420
first target and the second target. And this is the really cool thing. Then all this noise

575
00:43:06,420 --> 00:43:12,800
is added, right? This is basically swamping out, completely swamping out the echoes.

576
00:43:13,180 --> 00:43:18,620
And amazingly, when you pass that filter with the chirp through it, the noise disappears

577
00:43:19,120 --> 00:43:24,840
and still you're left with the responses, the peaks at the two targets. And even if

578
00:43:24,840 --> 00:43:25,500
they were overlapped, they're still going to be overlapped. So you can see that the

579
00:43:25,500 --> 00:43:28,540
two targets relate on one another, like we had in the previous plot with just the CW.

580
00:43:28,700 --> 00:43:34,440
You would still be able to disambiguate those two targets. So FMCW is pretty simple. Like,

581
00:43:34,440 --> 00:43:40,660
you can use radar techniques that use coded transmissions. But, you know, this is like

582
00:43:40,660 --> 00:43:46,580
an in-between level where it's still fairly doable with easy construction, even in GRC.

583
00:43:47,560 --> 00:43:55,480
So any questions so far? Question? No questions.

584
00:43:55,500 --> 00:44:01,420
Question about the mapping. How did you get the, I guess, rotational rate of that

585
00:44:01,420 --> 00:44:05,500
radar to project in the... This is going back a couple seconds.

586
00:44:05,620 --> 00:44:10,720
Yeah, no, no. I guessed. I mean, I recorded... I think I can record 70 seconds to 16 gigabytes

587
00:44:10,720 --> 00:44:19,300
of RAM. And I think I guessed. The other one was, does the distance from

588
00:44:19,300 --> 00:44:25,240
the actual radar receiver, you on the hill, give like some kind of distortion to...

589
00:44:25,240 --> 00:44:28,540
Yes. Yeah, exactly. So I can show you later.

590
00:44:28,700 --> 00:44:28,940
Okay.

591
00:44:29,440 --> 00:44:34,800
But if anyone's interested in this notion of the distortion and the actual physical

592
00:44:34,800 --> 00:44:38,940
model of the radar system, and if you consider the path propagation, I'll talk about that

593
00:44:38,940 --> 00:44:45,540
just briefly. But in a Blackhack talk I did a couple of years back, I show a slide

594
00:44:45,540 --> 00:44:52,980
where actually I wrote a program to create a rasterization, a visualization of that

595
00:44:52,980 --> 00:44:53,660
distortion map.

596
00:44:53,760 --> 00:44:54,220
Yeah.

597
00:44:54,220 --> 00:44:57,300
So you basically give it the offset and angle between the receiver and the transmitter,

598
00:44:57,500 --> 00:45:01,580
and then it'll calculate a distortion grid that you... I didn't do it, but you would

599
00:45:01,580 --> 00:45:05,860
then apply that to your returns and then it would undistort everything. So that's possible

600
00:45:05,860 --> 00:45:09,900
to do. But that's a very good point in that we're always talking about virtual range and

601
00:45:09,900 --> 00:45:14,600
virtual time echoes, and so you have to keep in mind how the geometry of your radar is

602
00:45:14,600 --> 00:45:15,540
set up. Next question.

603
00:45:15,840 --> 00:45:17,260
Could you just go back a slide?

604
00:45:17,640 --> 00:45:17,900
Yeah.

605
00:45:18,880 --> 00:45:21,140
So there you've shown the filtering...

606
00:45:21,140 --> 00:45:22,280
This is from Wikipedia, by the way.

607
00:45:22,380 --> 00:45:24,200
Yeah. I realize that. I'm going to ask the question.

608
00:45:24,200 --> 00:45:27,720
The previous Wikipedia slide had the two pulses really close together.

609
00:45:28,040 --> 00:45:28,200
Yep.

610
00:45:28,440 --> 00:45:34,420
Is there a limit for the chirp like decoding, filtering the chirp in terms of frequency

611
00:45:34,420 --> 00:45:37,120
or is there something else? Can you...

612
00:45:37,120 --> 00:45:38,260
To be able to disambiguate?

613
00:45:38,340 --> 00:45:38,580
Yes.

614
00:45:38,600 --> 00:45:43,500
So again, that's all set and constrained by the parameters of your radar system.

615
00:45:43,700 --> 00:45:48,280
So how, why the chirp is, what your signal bandwidth is, what your sample rate of the

616
00:45:48,280 --> 00:45:53,580
ADC is, all that stuff. And I've got a little Jupyter Notebook.

617
00:45:53,660 --> 00:45:54,000
Yeah.

618
00:45:54,200 --> 00:45:57,500
That I'll put online and you can enter in all of those parameters and it'll tell

619
00:45:57,500 --> 00:46:00,020
you what the, what all the constraints of your radar system is.

620
00:46:00,100 --> 00:46:03,560
It's just because the Wikipedia page showed that previous kind of, you know, signals

621
00:46:03,560 --> 00:46:04,020
close together.

622
00:46:04,100 --> 00:46:04,300
Yeah.

623
00:46:04,360 --> 00:46:06,680
I was intrigued to see what the parameters were.

624
00:46:06,820 --> 00:46:10,820
Yeah. I mean, that was some simulation that someone put up, but again, it's all up to

625
00:46:10,820 --> 00:46:11,160
the numbers.

626
00:46:11,300 --> 00:46:11,440
Cheers.

627
00:46:13,040 --> 00:46:17,400
Just a quickie. I mean, would an autocorrelation have given you that rotation rate if it may,

628
00:46:17,540 --> 00:46:19,240
as I can just go look at it and get it right?

629
00:46:19,240 --> 00:46:21,000
Yes. Yeah, it would. Yeah, yeah. I mean, you can...

630
00:46:21,000 --> 00:46:21,140
That's enough structure?

631
00:46:21,360 --> 00:46:22,800
You can... It's...

632
00:46:23,840 --> 00:46:28,480
I don't know. Each capture's like... I mean, it's like 15 gigs or something. So, you could

633
00:46:28,480 --> 00:46:29,420
do it. It would take a little while.

634
00:46:29,580 --> 00:46:29,840
Yeah.

635
00:46:29,840 --> 00:46:35,260
But that's exactly right. The point is raised that if you wanted to figure out the

636
00:46:36,440 --> 00:46:40,860
rotation rate, you know, I guess, because I didn't want to do it. But what you would

637
00:46:40,860 --> 00:46:46,500
do, completely correct, is you would run an autocorrelation with the capture and do a correlation

638
00:46:46,500 --> 00:46:50,940
with itself. And as soon as the similar kind of features match up again, you'll get a response.

639
00:46:51,260 --> 00:46:52,760
At the time, whether...

640
00:46:52,760 --> 00:46:56,840
...where the radar has resumed and completed a single rotation. Thank you very much.

641
00:47:00,540 --> 00:47:04,940
Any other questions? That's great. Thank you for these questions. It's really good. No?

642
00:47:05,040 --> 00:47:09,880
I'll move on then. So, with this continuous or full duty cycle radar system, you have

643
00:47:09,880 --> 00:47:14,420
your CHIRP generator and that's transmitted. So, as you heard here, like, it's being transmitted

644
00:47:14,420 --> 00:47:18,800
out of the speakers. And then the receiver, in this case, in that demonstration, it was

645
00:47:18,800 --> 00:47:21,880
the microphone in my laptop. And there...

646
00:47:22,340 --> 00:47:22,740
Yeah.

647
00:47:22,740 --> 00:47:30,080
...I then mix the input with the CHIRP, right? And because we're receiving and generating

648
00:47:30,080 --> 00:47:35,260
this simultaneously on the same hardware, it means that the clocks are synchronized.

649
00:47:35,620 --> 00:47:40,460
Because I'm using the audio hardware, that doesn't allow you to start transmitting and

650
00:47:40,460 --> 00:47:46,160
receiving the streams at the same time. That's different on, say, a USRP, which I'll show

651
00:47:46,160 --> 00:47:51,040
you next. And that's why I have this delay slot factor that you can change and I'll show

652
00:47:51,040 --> 00:47:51,500
you that in a minute.

653
00:47:51,500 --> 00:47:58,060
But you mix the two together and then you get the deCHIRP signal. And the deCHIRP signal

654
00:47:58,060 --> 00:48:02,860
holds wonders that I'll try to explain to you, but it means you can do some really sophisticated

655
00:48:02,860 --> 00:48:07,760
processing in a pretty simple way and get much more rich information out about what's

656
00:48:07,760 --> 00:48:10,800
going on in the space around you.

657
00:48:11,060 --> 00:48:18,140
So, what's important to note here is that once you deCHIRP the receive signal, then

658
00:48:19,060 --> 00:48:20,780
you get constant tones.

659
00:48:21,500 --> 00:48:29,500
So, when it deCHIRPS the...when it deCHIRPS itself, you'll just get a DC tone. And then

660
00:48:29,500 --> 00:48:34,900
any other reflections will end up generating other tones at different frequencies. And

661
00:48:34,900 --> 00:48:39,660
that's the key there. The cool thing about CHIRP radar is that once you deCHIRP stuff,

662
00:48:39,960 --> 00:48:45,080
your reflections become tones at different frequencies and that frequency is controlled

663
00:48:45,080 --> 00:48:50,720
by the distance from your radar, which is kind of mind-blowing when you think about

664
00:48:50,720 --> 00:48:54,660
it. It's just a really nice way all the math kind of falls out. And then you can just use

665
00:48:54,660 --> 00:48:56,180
FFTs and then do cool stuff.

666
00:48:56,920 --> 00:48:57,320
.

667
00:49:03,580 --> 00:49:07,760
I don't think so because you're still just doing it on the basis of the response

668
00:49:07,760 --> 00:49:12,340
of the filter. So, you're still looking for a peak response. And you can send out any

669
00:49:12,340 --> 00:49:16,720
kind of signal you wanted to. And I'll demonstrate that actually. The last thing I'll show you

670
00:49:16,720 --> 00:49:19,760
is passive radar using exactly these concepts.

671
00:49:19,760 --> 00:49:26,740
Using digital television signals off airplanes. So, stick around if you want to see that.

672
00:49:26,880 --> 00:49:33,900
That's what this is building up to. So, what's interesting is that in, you know, properly

673
00:49:33,900 --> 00:49:38,140
deployed radar systems, what often happens is that there's more sophisticated RF plumbing.

674
00:49:38,460 --> 00:49:45,020
So, when you have a receiver and you're continually transmitting, you don't want to have to ingest

675
00:49:45,020 --> 00:49:49,720
your local signal because it's always going to be the strongest thing you hear. And considering

676
00:49:49,720 --> 00:49:54,100
that it's going to be eventually fed into an analog to digital converter that has a fixed

677
00:49:54,100 --> 00:49:58,620
dynamic range, you don't want to blow out your receiver. You don't want to blow out the ADC.

678
00:49:58,840 --> 00:50:04,240
And so, what you can do is you can actually mix out or null out your tone that you're

679
00:50:04,240 --> 00:50:14,180
transmitting just using analog RF mixes and blocks or DC blockers, for example. And that

680
00:50:14,180 --> 00:50:19,700
means that you take the power out, like the actual RF power you subtract, not digitally, but you

681
00:50:19,720 --> 00:50:24,720
know, physically electromagnetic domain, and then all you're left with are the responses,

682
00:50:24,900 --> 00:50:31,320
the echoes. And that way you can much better utilize the dynamic range of your ADC and

683
00:50:31,320 --> 00:50:37,440
see reflections further out. So, from Wikipedia as well, this is a nice diagram to illustrate

684
00:50:37,440 --> 00:50:42,060
that kind of RF setup. So, again, you've got your generator, power divider, transmit that

685
00:50:42,060 --> 00:50:49,140
out, and you split that off, bring in the output of your receiver through a preamplifier, mix

686
00:50:49,140 --> 00:50:55,240
out the signal that you transmitted, low-pass filter, amplify again and into your ADC, and

687
00:50:55,240 --> 00:51:01,980
then you do your digital processing. So, that's the simple kind of structure. So, what I'm

688
00:51:01,980 --> 00:51:07,040
going to try and do now is take you through how you do range calculations, and then this

689
00:51:07,040 --> 00:51:13,840
is the magical new part, how to do Doppler calculations using FMCW and GRC. And it's

690
00:51:13,840 --> 00:51:19,120
basically all FFTs. It's, it's, this would appear like this massive mystery machine,

691
00:51:19,140 --> 00:51:22,280
a three-degree black box to me for the longest time. When I finally kind of sat down and

692
00:51:22,280 --> 00:51:27,660
looked at it, it's actually really, really elegant. So, my goal here today is to make

693
00:51:27,660 --> 00:51:32,280
you want to try to understand at least a little bit of this. So, this is a plot in

694
00:51:32,280 --> 00:51:36,580
the frequency domain, right? We talked about chirps. We know that they look like this

695
00:51:36,580 --> 00:51:42,420
on the spectrum. Here you've got time going from left to right and frequency from bottom

696
00:51:42,420 --> 00:51:48,840
to top. And the red signal here is the chirp that you're transmitting, and the green signal

697
00:51:48,840 --> 00:51:52,880
is the echo. So, remember, an echo is just going to send back what you transmitted,

698
00:51:53,020 --> 00:51:57,820
but with a delay in time, right? So, as you can see, the green is delayed a little bit,

699
00:51:58,360 --> 00:52:06,040
delta T, but from the, the original transmission. Now, what the beautiful thing about this,

700
00:52:06,080 --> 00:52:11,340
as I, as I mentioned, is that any time delay will produce a frequency shift at any single

701
00:52:11,960 --> 00:52:17,720
sample in time. So, if we look at time T1, you've got a frequency shift from your transmitted

702
00:52:17,720 --> 00:52:18,400
tone.

703
00:52:18,840 --> 00:52:28,360
At that particular point in time to your received echo. And what happens is, what

704
00:52:28,360 --> 00:52:45,840
was I going to demonstrate to you there? Maybe the whole thing here. Yeah, that's what I

705
00:52:45,840 --> 00:52:48,000
was going to show you. So, going back to the audio demo.

706
00:52:48,840 --> 00:52:54,220
We had this, right? And we know what the VCO input looks like. We know what the VCO

707
00:52:54,220 --> 00:53:02,240
output looks like. And now we're going to look at the FFT of the output of the de-chirping,

708
00:53:02,260 --> 00:53:08,680
of the mixing, right? So, what this is doing is it's taking the chirp, and then it's going

709
00:53:08,680 --> 00:53:14,360
to multiply that by what is being received from the microphone. And what you'll notice

710
00:53:14,360 --> 00:53:18,460
here is if you compare this one, right, to this one.

711
00:53:18,480 --> 00:53:18,820
Yeah.

712
00:53:18,820 --> 00:53:23,280
What it looks like, and hopefully you'll see this, it looks like the spectrum is being

713
00:53:23,280 --> 00:53:28,120
rotated over and over and over again at a fixed rate. See how it looks like the entire

714
00:53:28,120 --> 00:53:33,080
spectrum is shifting? So, the left-hand side here. So, actually, do people want to whistle?

715
00:53:33,180 --> 00:53:35,920
And you'll see your whistles come out, but they'll be diagonal lines here.

716
00:53:45,960 --> 00:53:52,660
So, see, this is me. And this is the cool thing. Our chirp signal, remember I was telling

717
00:53:52,660 --> 00:53:57,840
you it would turn into a fixed tone? Can you all see the fixed tone here? It's that line

718
00:53:58,280 --> 00:54:05,040
coming down the middle at roughly DC, right? That is the tone that we want to receive.

719
00:54:05,040 --> 00:54:11,980
And any other tones that end up, I think in this case, to the left of it, will be echoes

720
00:54:11,980 --> 00:54:19,080
that have come back from around the room into the microphone on the laptop. And so, that

721
00:54:19,080 --> 00:54:23,400
frequency difference, because remember this is in the frequency domain now, that frequency

722
00:54:23,400 --> 00:54:28,480
difference is going to give you your virtual range information, right? Which is pretty,

723
00:54:28,520 --> 00:54:29,080
pretty neat.

724
00:54:29,840 --> 00:54:33,840
Now, what does this look like? I showed this last year, but...

725
00:54:35,040 --> 00:54:39,880
I've hopefully explained things a little better so that it makes more sense. What I'm going

726
00:54:39,880 --> 00:54:41,300
to show you now is...

727
00:54:46,100 --> 00:54:52,060
the same thing, but actually running in the proper mode. See, this is the chirp again.

728
00:54:52,260 --> 00:54:59,720
It's running at a higher rate because we want to be able to update the display more. And

729
00:54:59,720 --> 00:55:03,620
you see when I'm talking, it's still picking me up because it's constantly rotating the

730
00:55:03,620 --> 00:55:03,920
spectrum.

731
00:55:03,920 --> 00:55:09,260
It doesn't look like a proper frequency response now. So, what I'm going to do is get my other

732
00:55:09,260 --> 00:55:10,760
laptop to be the target.

733
00:55:13,420 --> 00:55:13,940
And...

734
00:55:15,280 --> 00:55:17,040
I'm going to set my...

735
00:55:19,080 --> 00:55:21,780
my delay slop so that you see the...

736
00:55:23,520 --> 00:55:25,460
the DC part of it somewhere.

737
00:55:30,360 --> 00:55:31,880
Okay, apparently it's...

738
00:55:32,620 --> 00:55:33,460
not there.

739
00:55:45,680 --> 00:55:50,800
Nate Temple, the genius here, pointed out that I'm trying to move my laptop in front

740
00:55:50,800 --> 00:55:55,060
of speakers that aren't actually outputting anything. I want to use the speakers in my

741
00:55:55,060 --> 00:55:58,340
laptop because that's where the microphone is, not those speakers. Thanks, Nate.

742
00:56:03,700 --> 00:56:05,640
Nate is an absolute legend, by the way.

743
00:56:07,800 --> 00:56:13,560
If you didn't get to see his cyberspectrum talk from the other night, I recommend you

744
00:56:13,560 --> 00:56:16,900
go online and watch it because he has built some amazing stuff.

745
00:56:17,500 --> 00:56:18,680
All right, here we go. This is better now.

746
00:56:20,460 --> 00:56:20,860
So...

747
00:56:22,560 --> 00:56:26,820
what I'm going to do is... this is the DC component, and because I'm talking right,

748
00:56:26,920 --> 00:56:31,420
we're just going to get this noise showing up here. But if I put my laptop here...

749
00:56:33,900 --> 00:56:36,900
then you can see there's that main response there...

750
00:56:37,500 --> 00:56:41,380
that is directly related to the height of the laptop.

751
00:56:43,720 --> 00:56:46,380
Don't clap though because then you'll upset the microphone.

752
00:56:48,940 --> 00:56:49,460
Right?

753
00:56:50,500 --> 00:56:51,020
So...

754
00:56:52,660 --> 00:56:55,060
That's alright. I planned that.

755
00:56:56,320 --> 00:56:57,580
Um...so what's happening here?

756
00:56:57,700 --> 00:56:59,520
A chirp is being sent out from the laptop.

757
00:56:59,780 --> 00:57:01,080
I mean, this is...

758
00:57:01,080 --> 00:57:03,100
a chirp, those

759
00:57:03,800 --> 00:57:05,480
air molecules are being vibrated

760
00:57:05,480 --> 00:57:07,360
and then that audio wave

761
00:57:07,360 --> 00:57:08,860
is moving up, hitting

762
00:57:08,860 --> 00:57:10,500
the bottom side of my laptop

763
00:57:10,500 --> 00:57:12,980
and then being reflected back down in the microphone.

764
00:57:13,280 --> 00:57:15,340
And amazingly, the microphone can discern

765
00:57:15,340 --> 00:57:17,200
even though it's hearing itself,

766
00:57:17,380 --> 00:57:19,440
it can still discern the response

767
00:57:19,440 --> 00:57:21,480
from an object way out here.

768
00:57:21,660 --> 00:57:23,180
So that's the power of

769
00:57:24,040 --> 00:57:24,600
FMCW.

770
00:57:24,980 --> 00:57:27,460
Now, there was another little thing in there

771
00:57:27,460 --> 00:57:29,440
that was on my screen. I will show you that.

772
00:57:29,440 --> 00:57:31,020
That's actually the Doppler plot

773
00:57:31,020 --> 00:57:33,520
and I'll come to that. What that gives you

774
00:57:33,520 --> 00:57:35,580
is not only range information, but it gives

775
00:57:35,580 --> 00:57:37,460
you velocity information, so you can tell how

776
00:57:37,460 --> 00:57:38,760
quickly a target is moving.

777
00:57:41,100 --> 00:57:41,900
And that's

778
00:57:41,900 --> 00:57:43,200
really cool when it comes to aircraft.

779
00:57:44,760 --> 00:57:45,740
So, what I want

780
00:57:45,740 --> 00:57:47,600
to show you here now is this plot, where

781
00:57:47,600 --> 00:57:49,560
instead of looking at the original frequency domain

782
00:57:49,560 --> 00:57:51,200
what we received, this is the domain

783
00:57:51,860 --> 00:57:53,260
once we've done the de-chirping.

784
00:57:53,440 --> 00:57:55,360
So we have that DC component here

785
00:57:55,360 --> 00:57:57,740
and then in the green line, this is now our

786
00:57:58,460 --> 00:57:59,220
target return

787
00:57:59,220 --> 00:58:01,000
tone, right? Because we've de-chirped it.

788
00:58:01,220 --> 00:58:03,020
And so you have this difference in frequency.

789
00:58:03,280 --> 00:58:04,800
So the frequency change implies

790
00:58:05,500 --> 00:58:07,380
that time delay, and therefore

791
00:58:07,380 --> 00:58:09,300
virtual range. Now,

792
00:58:09,380 --> 00:58:11,240
what's neat is that you can take

793
00:58:11,240 --> 00:58:13,180
the de-chirped output

794
00:58:13,180 --> 00:58:15,460
and then pass it through an FFT,

795
00:58:15,620 --> 00:58:17,640
right? Because an FFT will give you

796
00:58:17,640 --> 00:58:18,760
the energy

797
00:58:19,320 --> 00:58:21,380
in each bin. And so the trick

798
00:58:21,380 --> 00:58:23,420
here is that you use an FFT

799
00:58:24,120 --> 00:58:25,260
you know how you

800
00:58:25,260 --> 00:58:27,240
have a certain size in your FFT,

801
00:58:27,340 --> 00:58:28,820
a certain number of points that you

802
00:58:29,220 --> 00:58:31,360
form. The number of points in your FFT

803
00:58:31,360 --> 00:58:33,460
you can make the same

804
00:58:33,460 --> 00:58:34,680
number of samples

805
00:58:35,180 --> 00:58:37,640
as what is made up in a single chirp.

806
00:58:37,780 --> 00:58:39,440
Right? So you decide how long your chirp

807
00:58:39,440 --> 00:58:41,080
is, and then you set your FFT

808
00:58:41,680 --> 00:58:43,380
in terms of samples to be exactly

809
00:58:43,380 --> 00:58:45,500
the same amount of time. And what

810
00:58:45,500 --> 00:58:47,480
that means is that every single time you do

811
00:58:47,480 --> 00:58:49,300
one full chirp, you take one

812
00:58:49,300 --> 00:58:51,400
FFT. And

813
00:58:52,500 --> 00:58:53,440
I'm trying to illustrate

814
00:58:53,440 --> 00:58:54,520
that here by these

815
00:58:55,160 --> 00:58:57,360
lines. So if you think about

816
00:58:57,360 --> 00:58:59,140
these over time, this is like a waterfall.

817
00:58:59,220 --> 00:59:01,380
These are your frequency

818
00:59:01,380 --> 00:59:03,340
bins that you get out of your

819
00:59:03,340 --> 00:59:05,580
FFT transform. And in this case

820
00:59:05,580 --> 00:59:07,160
the echo return, the green line

821
00:59:07,160 --> 00:59:09,500
ends up falling into, in this

822
00:59:09,500 --> 00:59:11,260
case, arbitrary bin number 5.

823
00:59:11,700 --> 00:59:13,360
And remember how we said that

824
00:59:13,360 --> 00:59:15,560
differences in frequency will imply

825
00:59:15,560 --> 00:59:17,400
your virtual range.

826
00:59:17,880 --> 00:59:19,240
Here, the

827
00:59:19,240 --> 00:59:21,560
bin that it falls into will give you

828
00:59:21,560 --> 00:59:23,260
the range

829
00:59:23,260 --> 00:59:25,100
information for that particular bin.

830
00:59:25,280 --> 00:59:26,920
So here we can say bin 5

831
00:59:26,920 --> 00:59:29,020
and we can calculate

832
00:59:29,220 --> 00:59:31,260
how far away that is. And

833
00:59:31,260 --> 00:59:33,080
to calculate how far away that is,

834
00:59:33,180 --> 00:59:35,160
you know how long one sample lasts

835
00:59:35,160 --> 00:59:37,360
because you set a sample

836
00:59:37,360 --> 00:59:39,140
rate on your SDR, right?

837
00:59:39,200 --> 00:59:41,080
So 10 megahertz or whatever, 1 megahertz.

838
00:59:41,280 --> 00:59:43,260
So you can calculate 1 over that to

839
00:59:43,260 --> 00:59:45,540
calculate the period of a single sample.

840
00:59:46,120 --> 00:59:47,140
And each bin

841
00:59:47,140 --> 00:59:48,440
in terms of duration

842
00:59:49,980 --> 00:59:51,300
maps up with the

843
00:59:51,300 --> 00:59:52,800
duration of a single sample.

844
00:59:53,000 --> 00:59:55,300
And if you know how long it is, if we're

845
00:59:55,300 --> 00:59:56,840
talking about RF now and not audio,

846
00:59:57,060 --> 00:59:59,140
you can multiply it by the speed of light.

847
00:59:59,220 --> 01:00:00,860
And that gives you your round trip

848
01:00:01,560 --> 01:00:03,220
time and you can calculate your virtual range.

849
01:00:03,500 --> 01:00:05,340
So it's cool because you can map

850
01:00:05,340 --> 01:00:07,260
to frequency really easily just by doing

851
01:00:07,260 --> 01:00:09,320
an FFT and then you can map that

852
01:00:09,320 --> 01:00:10,760
directly back to virtual range.

853
01:00:11,080 --> 01:00:13,240
So does that kind of

854
01:00:13,240 --> 01:00:15,120
make sense so far? I see people nodding, which is good.

855
01:00:15,980 --> 01:00:17,260
Now, the wild thing is

856
01:00:17,260 --> 01:00:18,780
in this instance,

857
01:00:19,100 --> 01:00:21,000
you know, this is a target at a fixed

858
01:00:21,000 --> 01:00:22,540
range and it's not changing.

859
01:00:22,840 --> 01:00:24,980
Here, you might have two

860
01:00:24,980 --> 01:00:26,660
different targets moving

861
01:00:26,660 --> 01:00:28,440
toward and away from

862
01:00:28,440 --> 01:00:30,100
your radar receiver.

863
01:00:30,420 --> 01:00:32,020
And so you would expect a plot like this.

864
01:00:32,080 --> 01:00:33,140
And we kind of saw that before

865
01:00:33,960 --> 01:00:36,260
with the FM chirp. Actually, one thing I was

866
01:00:36,260 --> 01:00:38,080
going to try was, I don't have another

867
01:00:38,080 --> 01:00:40,120
laptop, but maybe if somebody has a laptop, we can try it

868
01:00:40,120 --> 01:00:42,160
at the end. We can have two here and then if you

869
01:00:42,160 --> 01:00:44,040
move them separately, you'd get obviously

870
01:00:44,040 --> 01:00:46,240
two different crossing returns like that.

871
01:00:49,480 --> 01:00:50,180
Maybe that's what

872
01:00:50,180 --> 01:00:52,180
I was going to demo, but I'll do that later

873
01:00:52,180 --> 01:00:54,040
because time is short.

874
01:00:54,860 --> 01:00:56,200
So, again, as I was

875
01:00:56,200 --> 01:00:57,600
saying, the sample rate sets

876
01:00:57,600 --> 01:00:59,660
the duration of a single sample

877
01:00:59,660 --> 01:01:01,620
and therefore limits your range

878
01:01:01,620 --> 01:01:03,700
resolution. Because if you consider, you

879
01:01:03,700 --> 01:01:05,600
get discretized bins coming out

880
01:01:05,600 --> 01:01:07,460
of an FFT, and

881
01:01:07,460 --> 01:01:09,620
the length, the total length

882
01:01:09,620 --> 01:01:11,680
of that is going to be your unambiguous

883
01:01:11,680 --> 01:01:13,180
possible round trip time.

884
01:01:13,500 --> 01:01:15,200
And so your pulse repetition

885
01:01:15,200 --> 01:01:17,440
frequency sets your unambiguous range,

886
01:01:18,160 --> 01:01:19,460
but the sample rate

887
01:01:19,460 --> 01:01:21,340
also sets the

888
01:01:22,460 --> 01:01:23,020
distinct

889
01:01:23,020 --> 01:01:25,620
range that you can resolve within a single bin.

890
01:01:25,760 --> 01:01:27,580
So as you step from one bin to the other, you can

891
01:01:27,600 --> 01:01:29,580
to the next. That's going to be some amount of range

892
01:01:29,580 --> 01:01:31,280
that you go, you know, from one

893
01:01:31,280 --> 01:01:33,460
to the next. And so what you have to consider

894
01:01:33,460 --> 01:01:35,460
is that targets, depending upon your radar system,

895
01:01:35,900 --> 01:01:37,520
might be differently

896
01:01:37,520 --> 01:01:39,520
spaced, but they might fall

897
01:01:39,520 --> 01:01:41,560
within the range of a single

898
01:01:41,560 --> 01:01:43,300
bin. And so you'll get

899
01:01:43,300 --> 01:01:45,360
one return for multiple targets

900
01:01:45,360 --> 01:01:47,580
in a single bin, and that's ambiguous.

901
01:01:47,980 --> 01:01:49,400
So you want to do something which is

902
01:01:49,400 --> 01:01:51,500
Doppler processing to make

903
01:01:51,500 --> 01:01:52,240
that unambiguous.

904
01:01:54,240 --> 01:01:55,540
And as I was saying, you've got

905
01:01:55,540 --> 01:01:57,520
different radar geometries. So you've got

906
01:01:57,600 --> 01:01:59,500
a monostatic system where the transmitter

907
01:01:59,500 --> 01:02:01,540
and receiver are co-located, and so it's just

908
01:02:01,540 --> 01:02:03,280
a direct line of sight round trip time.

909
01:02:03,580 --> 01:02:05,220
But if you have a bistatic radar system

910
01:02:05,220 --> 01:02:07,340
where the receiver is separate from the transmitter,

911
01:02:07,520 --> 01:02:09,480
then you have to take into account the

912
01:02:09,480 --> 01:02:11,480
geometry, because that delay is going

913
01:02:11,480 --> 01:02:13,280
to be, you know, different.

914
01:02:13,400 --> 01:02:15,440
You might have a target way out here, and you'll have the signal bounce

915
01:02:15,440 --> 01:02:17,420
out and then come back, or it might be in between the two

916
01:02:17,420 --> 01:02:19,520
and so on. So important

917
01:02:19,520 --> 01:02:21,300
to keep that in mind. So

918
01:02:21,300 --> 01:02:23,000
what I was talking about there is that

919
01:02:23,000 --> 01:02:25,300
with RF, the speed of light, we know what that is.

920
01:02:26,440 --> 01:02:27,420
And for a

921
01:02:27,600 --> 01:02:29,440
single sample, that's going to be a

922
01:02:29,440 --> 01:02:31,560
very large distance, because we're talking about the speed of light.

923
01:02:32,780 --> 01:02:33,580
300,000

924
01:02:33,580 --> 01:02:35,480
kilometers per second. So even at

925
01:02:35,480 --> 01:02:37,260
a high sample rate, megahertz,

926
01:02:37,420 --> 01:02:39,440
a single sample is still going to allow

927
01:02:39,440 --> 01:02:40,960
light to travel a long way.

928
01:02:41,700 --> 01:02:43,140
Or in this case, the RF energy.

929
01:02:43,900 --> 01:02:45,420
And so what we want to do is

930
01:02:45,420 --> 01:02:47,520
you can increase the sample rate to give you

931
01:02:47,520 --> 01:02:49,360
better range resolution, but there comes

932
01:02:49,360 --> 01:02:51,340
a point where if you're just using an SDR

933
01:02:51,340 --> 01:02:53,600
and a laptop, you can't go to, you know, gigahertz

934
01:02:53,600 --> 01:02:54,860
worth of bandwidth.

935
01:02:55,040 --> 01:02:56,660
You will end up being constrained.

936
01:02:57,680 --> 01:02:59,420
By your bus and by your processing

937
01:02:59,420 --> 01:03:01,660
speed. So what you should consider

938
01:03:01,660 --> 01:03:03,280
then with an FFT here is that

939
01:03:03,280 --> 01:03:04,340
this is an FFT

940
01:03:05,140 --> 01:03:06,840
and you have these range bins

941
01:03:06,840 --> 01:03:09,600
and then every time the chirp finishes it, it does

942
01:03:09,600 --> 01:03:11,520
the transform and then it just starts and

943
01:03:11,520 --> 01:03:12,920
keeps cycling like that over and over.

944
01:03:13,120 --> 01:03:14,880
So you can see that cycle there.

945
01:03:16,140 --> 01:03:17,540
And we're talking about

946
01:03:18,280 --> 01:03:19,160
multiple targets

947
01:03:19,160 --> 01:03:20,540
fitting into a single

948
01:03:21,780 --> 01:03:22,460
range bin.

949
01:03:24,200 --> 01:03:25,540
And so once you do the FFT

950
01:03:25,540 --> 01:03:27,460
and you were to take the magnitude, you would

951
01:03:27,460 --> 01:03:29,400
basically get an energy response at

952
01:03:29,400 --> 01:03:30,260
that range.

953
01:03:33,680 --> 01:03:35,420
So as I was saying, you might have two echoes

954
01:03:35,420 --> 01:03:37,140
that fall into the same range because they're not

955
01:03:37,140 --> 01:03:39,280
the same target. They're slightly different, but

956
01:03:39,280 --> 01:03:41,380
from the point of view of your receiver, they

957
01:03:41,380 --> 01:03:43,260
end up giving you the same round

958
01:03:43,260 --> 01:03:43,760
trip time.

959
01:03:46,660 --> 01:03:47,300
And the other

960
01:03:47,300 --> 01:03:49,640
problem too is that you might have

961
01:03:49,640 --> 01:03:51,400
a radar system, and I'll

962
01:03:51,400 --> 01:03:53,260
show you this with a passive one, where the

963
01:03:54,220 --> 01:03:55,480
transmitter is so strong

964
01:03:55,900 --> 01:03:57,220
even out to many range

965
01:03:57,220 --> 01:03:59,260
bins that it completely swamps

966
01:03:59,260 --> 01:03:59,940
your receiver

967
01:04:01,080 --> 01:04:01,860
in the

968
01:04:03,400 --> 01:04:05,260
for, you know, if you were just

969
01:04:05,260 --> 01:04:07,240
doing this transform. And what's nice is that

970
01:04:07,240 --> 01:04:08,720
you can do Doppler processing,

971
01:04:08,940 --> 01:04:11,000
which reveals this hidden information

972
01:04:11,000 --> 01:04:12,940
working in the phase information

973
01:04:12,940 --> 01:04:14,760
that's output from the Fourier transform

974
01:04:14,760 --> 01:04:16,760
that will actually show you these

975
01:04:16,760 --> 01:04:19,140
targets, even though ordinarily you wouldn't

976
01:04:19,140 --> 01:04:20,780
see them because you're swamped by

977
01:04:20,780 --> 01:04:23,180
your local transmitter or ground clutter.

978
01:04:25,520 --> 01:04:27,120
So yeah, the clutter could have

979
01:04:27,220 --> 01:04:29,380
affected that bin, so you don't see them

980
01:04:29,380 --> 01:04:31,520
if you're just looking at range information, or

981
01:04:31,520 --> 01:04:33,740
it'll take out your entire transform.

982
01:04:34,680 --> 01:04:36,860
So the Doppler effect. We've heard about this.

983
01:04:38,640 --> 01:04:41,060
And we know that it will cause a, usually

984
01:04:41,060 --> 01:04:43,500
a shift in frequency. You know, the classic example

985
01:04:43,500 --> 01:04:47,860
is an ambulance that's driving by. And in this

986
01:04:47,860 --> 01:04:49,900
case, what I'm gonna try to illustrate to you

987
01:04:49,900 --> 01:04:52,660
is how it changes the phase of the

988
01:04:52,660 --> 01:04:55,700
echo that comes back to the radar. That's the

989
01:04:55,700 --> 01:04:56,200
key here.

990
01:04:56,200 --> 01:04:58,860
So, we get a phase change due to motion.

991
01:05:00,460 --> 01:05:02,680
And again, if anybody, any, any questions so

992
01:05:02,680 --> 01:05:06,700
far? Any, any comments? If you want me to

993
01:05:06,700 --> 01:05:10,180
repeat anything, please, please ask. Don't be,

994
01:05:10,220 --> 01:05:12,300
don't be afraid, because it took me a while to

995
01:05:12,300 --> 01:05:18,000
figure this out. Yeah. Yeah, so it, it depends

996
01:05:18,000 --> 01:05:20,860
on sort of how you conceptualize it and, and

997
01:05:20,860 --> 01:05:23,120
how you receive it. In this particular case,

998
01:05:23,260 --> 01:05:25,700
there's one you think about. I mean, I'll illustrate it

999
01:05:25,700 --> 01:05:29,280
here and, and hopefully it'll make sense. So the

1000
01:05:29,280 --> 01:05:32,740
idea is, with Doppler processing, that you receive not

1001
01:05:32,740 --> 01:05:35,620
just echoes from a single chirp, but you do

1002
01:05:35,620 --> 01:05:38,820
it for multiple chirps over some arbitrary period that

1003
01:05:38,820 --> 01:05:42,200
you select. That'll also have implications on what you

1004
01:05:42,200 --> 01:05:44,800
can resolve. But this is referred to as the integration

1005
01:05:44,800 --> 01:05:46,780
time. So let's say you can have integration time

1006
01:05:46,780 --> 01:05:49,020
of like a second. And let's say you're sending out

1007
01:05:49,020 --> 01:05:52,000
a hundred pulses per second. So you collect a hundred

1008
01:05:52,000 --> 01:05:54,960
of those returns, and then you process them in one

1009
01:05:54,960 --> 01:05:55,240
batch.

1010
01:05:55,700 --> 01:05:59,760
And then from that you extract Doppler information. And

1011
01:05:59,760 --> 01:06:01,320
the key here, and I'll illustrate this for you,

1012
01:06:01,360 --> 01:06:05,020
is that you, instead of just ending where we

1013
01:06:05,020 --> 01:06:07,260
had before, where you do an FFT of your

1014
01:06:07,940 --> 01:06:10,780
range response bins, you build up then a hundred

1015
01:06:10,780 --> 01:06:15,800
of those range response transforms, right. So you imagine

1016
01:06:15,800 --> 01:06:18,120
they're in rows. And then you run FFTs on

1017
01:06:18,120 --> 01:06:20,200
the columns. So what you end up doing is

1018
01:06:20,200 --> 01:06:24,500
you do a second series of FFTs on each individual

1019
01:06:24,500 --> 01:06:28,620
range over the integration period. And I'll show you

1020
01:06:28,620 --> 01:06:31,620
what that looks like. So the other thing to do

1021
01:06:31,620 --> 01:06:34,540
is remember, again, the geometry. So this is Doppler velocity

1022
01:06:35,520 --> 01:06:38,820
with respect to your radar system, not with respect

1023
01:06:38,820 --> 01:06:41,440
to the moving object. So if you consider this

1024
01:06:41,440 --> 01:06:43,580
scenario on the left here, you've got a radar transmitted

1025
01:06:43,580 --> 01:06:47,660
receiver, sends out a pulse. The airplane is moving tangentially

1026
01:06:47,660 --> 01:06:50,420
at that moment to the radar receiver. So it's

1027
01:06:50,420 --> 01:06:54,460
not going to impart any Doppler shift on the return, because the

1028
01:06:54,460 --> 01:06:57,780
there's no velocity component in the same plane as

1029
01:06:57,780 --> 01:07:01,860
your radar signal. Whereas in the second scenario, in the

1030
01:07:01,860 --> 01:07:04,000
middle, if the plane's moving toward the radar receiver,

1031
01:07:04,160 --> 01:07:06,720
it's going to impart the biggest Doppler shift on the

1032
01:07:06,720 --> 01:07:09,980
return. And if it's offset by some amount, then

1033
01:07:09,980 --> 01:07:14,000
even though it's, you know, not entirely tangential, like

1034
01:07:14,000 --> 01:07:16,100
in the first scenario, it's still going to impart

1035
01:07:16,100 --> 01:07:18,280
a little bit, because there is still a component there

1036
01:07:18,280 --> 01:07:24,340
that matches with the, the plane. So with Doppler processing, if you

1037
01:07:24,460 --> 01:07:27,700
consider the scenario where at any one time the

1038
01:07:27,700 --> 01:07:30,620
target is moving toward the radar there, in, at

1039
01:07:30,620 --> 01:07:33,760
T zero, that pulse is sent out. At round trip

1040
01:07:33,760 --> 01:07:36,180
time divided by two, the reflection is sent back.

1041
01:07:36,360 --> 01:07:38,760
And at the round trip time, you get the return.

1042
01:07:38,920 --> 01:07:41,820
And you can see that throughout that period, the

1043
01:07:41,820 --> 01:07:46,280
signal amplitude decreases, because of course, there's that loss through,

1044
01:07:46,320 --> 01:07:49,140
through free space and, and on the, on the

1045
01:07:49,140 --> 01:07:51,480
reflector. And what I want to illustrate to you

1046
01:07:51,480 --> 01:07:52,680
here is when that.

1047
01:07:55,140 --> 01:07:58,620
return comes back at a particular point in time, right?

1048
01:07:58,700 --> 01:08:00,140
It's always going to come back at a particular point,

1049
01:08:00,220 --> 01:08:02,660
which implies a particular range, right?

1050
01:08:02,700 --> 01:08:03,480
So there's a range there.

1051
01:08:03,740 --> 01:08:06,440
There's going to be a phase difference,

1052
01:08:07,080 --> 01:08:08,240
which might be zero.

1053
01:08:08,420 --> 01:08:10,280
There might be effectively no phase difference

1054
01:08:10,280 --> 01:08:11,760
or some phase difference

1055
01:08:11,760 --> 01:08:13,580
between the signal that you sent out

1056
01:08:13,580 --> 01:08:15,300
and the signal that comes back.

1057
01:08:15,600 --> 01:08:19,220
And if you consider the actual propagation of a sine wave,

1058
01:08:19,280 --> 01:08:21,160
even though, you know, they're photons and whatnot,

1059
01:08:23,020 --> 01:08:25,640
then depending upon where that plane is,

1060
01:08:25,720 --> 01:08:29,560
if it's here or it's like a meter over here, offset,

1061
01:08:30,920 --> 01:08:32,700
the wavefront will hit it,

1062
01:08:33,600 --> 01:08:37,440
and then, you know, if you imagine that wavefront hitting it,

1063
01:08:37,480 --> 01:08:38,780
it'll hit it at a particular phase

1064
01:08:38,780 --> 01:08:40,560
and then it'll reflect at that phase.

1065
01:08:40,900 --> 01:08:42,840
And if you move that plane a little bit,

1066
01:08:42,980 --> 01:08:46,120
then the tone will hit the plane

1067
01:08:46,120 --> 01:08:47,300
at a slightly different phase

1068
01:08:47,300 --> 01:08:48,920
and reflect at a different phase.

1069
01:08:49,100 --> 01:08:51,140
And so in this particular...

1070
01:08:51,160 --> 01:08:52,180
In this particular case, we're lucky

1071
01:08:52,180 --> 01:08:55,040
because everything happens to line up on the phase boundary,

1072
01:08:55,340 --> 01:08:56,920
and so when you get the response,

1073
01:08:57,220 --> 01:08:58,840
it actually ends up being the same phase.

1074
01:08:59,020 --> 01:09:00,400
So in here, it's zero degrees.

1075
01:09:00,700 --> 01:09:03,700
And remember, this is all in a single range bin.

1076
01:09:04,400 --> 01:09:06,800
Now, you could always say if a plane is traveling,

1077
01:09:06,900 --> 01:09:08,420
you'll see it move through the range bins

1078
01:09:08,420 --> 01:09:10,900
and that implies that it's moving with some velocity,

1079
01:09:11,120 --> 01:09:13,280
but this is all within a single range bin

1080
01:09:13,280 --> 01:09:15,120
because we want to disambiguate targets

1081
01:09:15,120 --> 01:09:16,180
and get velocity information.

1082
01:09:17,420 --> 01:09:19,440
So you've got transmitted pulse, received pulse,

1083
01:09:19,560 --> 01:09:20,180
no phase difference.

1084
01:09:20,620 --> 01:09:21,140
Now, if you're in a plane,

1085
01:09:21,140 --> 01:09:23,580
now let's consider the plane has moved a meter

1086
01:09:23,580 --> 01:09:25,060
and, you know,

1087
01:09:26,440 --> 01:09:28,240
it also depends upon your wavelength.

1088
01:09:29,500 --> 01:09:30,860
That has implications, too.

1089
01:09:31,880 --> 01:09:34,660
But here, we're talking about the same scenario.

1090
01:09:34,840 --> 01:09:36,760
And if you notice carefully here,

1091
01:09:37,660 --> 01:09:41,320
the wave that is returned is slightly offset.

1092
01:09:41,560 --> 01:09:41,820
See that?

1093
01:09:42,020 --> 01:09:43,800
So see there and then there.

1094
01:09:43,940 --> 01:09:45,560
It's a slightly different phase on that wave.

1095
01:09:45,880 --> 01:09:47,000
And so when it comes back,

1096
01:09:47,100 --> 01:09:49,400
the phase difference is going to be slightly different as well.

1097
01:09:49,980 --> 01:09:51,060
And when we...

1098
01:09:51,140 --> 01:09:52,200
compare them,

1099
01:09:52,300 --> 01:09:54,360
we get some arbitrary phase offset.

1100
01:09:55,060 --> 01:09:56,340
Now, let's consider...

1101
01:09:56,340 --> 01:09:59,000
These are two distinct pulses.

1102
01:09:59,600 --> 01:10:00,960
Now, let's consider four

1103
01:10:01,500 --> 01:10:03,600
in more or less successive time order.

1104
01:10:03,820 --> 01:10:04,960
So the first one comes back.

1105
01:10:05,280 --> 01:10:07,460
The second one comes back with a difference.

1106
01:10:07,700 --> 01:10:08,640
Third one comes back.

1107
01:10:08,900 --> 01:10:10,440
In this case, it's the same

1108
01:10:10,440 --> 01:10:11,500
because the plane moved again

1109
01:10:11,500 --> 01:10:13,800
and imparted another phase change.

1110
01:10:14,100 --> 01:10:14,880
And then again,

1111
01:10:15,020 --> 01:10:17,200
we've got a different phase offset

1112
01:10:17,200 --> 01:10:19,640
in the fourth period there.

1113
01:10:20,220 --> 01:10:20,900
And remember,

1114
01:10:21,140 --> 01:10:22,980
every chirp period that gets sent out,

1115
01:10:23,080 --> 01:10:26,800
we get this different energy coming back to the receiver.

1116
01:10:27,160 --> 01:10:29,880
It's the same chirp or the same signal.

1117
01:10:30,380 --> 01:10:32,280
The magnitude will look the same,

1118
01:10:32,320 --> 01:10:34,340
but the phase will be different.

1119
01:10:34,640 --> 01:10:35,980
And what you have to think about is

1120
01:10:35,980 --> 01:10:37,460
when you do FFTs,

1121
01:10:37,460 --> 01:10:38,720
when we look at them on the waterfall

1122
01:10:38,720 --> 01:10:41,240
or when we usually think about them,

1123
01:10:41,880 --> 01:10:42,800
what are you doing?

1124
01:10:43,260 --> 01:10:44,880
You're taking the FFT,

1125
01:10:45,000 --> 01:10:47,000
which gives you always a complex output,

1126
01:10:47,680 --> 01:10:49,580
and you always end up taking the magnitude

1127
01:10:49,580 --> 01:10:50,580
of that complex output.

1128
01:10:51,140 --> 01:10:52,220
And the effect of that is?

1129
01:10:53,360 --> 01:10:53,840
Anyone?

1130
01:10:54,940 --> 01:10:55,880
You lose data.

1131
01:10:56,060 --> 01:10:56,580
What do you lose?

1132
01:10:57,020 --> 01:10:58,440
You lose the phase information.

1133
01:10:58,900 --> 01:11:00,400
So here we want to use that phase information.

1134
01:11:02,220 --> 01:11:04,060
And the integration period is important

1135
01:11:04,060 --> 01:11:05,260
because that's the period

1136
01:11:05,260 --> 01:11:07,400
over which you're monitoring these phase changes

1137
01:11:07,400 --> 01:11:09,320
in each of the range bins.

1138
01:11:09,660 --> 01:11:10,420
Does that make sense?

1139
01:11:12,380 --> 01:11:15,080
So the idea is to get that phase information

1140
01:11:15,080 --> 01:11:16,620
from each FFT bin

1141
01:11:16,620 --> 01:11:18,400
over that integration period.

1142
01:11:18,700 --> 01:11:19,580
And then,

1143
01:11:21,380 --> 01:11:22,460
think about this.

1144
01:11:23,840 --> 01:11:26,940
What is a changing phase over time?

1145
01:11:29,060 --> 01:11:30,340
It's a rotating...

1146
01:11:30,920 --> 01:11:32,000
What's the magic P word?

1147
01:11:33,940 --> 01:11:35,040
It's a phasor, right?

1148
01:11:35,140 --> 01:11:36,740
So you've got a phasor going around.

1149
01:11:36,860 --> 01:11:38,700
And I think I have a blue arrow here, yes.

1150
01:11:39,260 --> 01:11:41,640
So, I mean, just to illustrate that a little bit more,

1151
01:11:41,680 --> 01:11:42,820
very, very simply,

1152
01:11:43,380 --> 01:11:44,320
in GNU Radio,

1153
01:11:44,460 --> 01:11:46,200
because we usually use fast sample rates

1154
01:11:46,200 --> 01:11:47,240
and fast...

1155
01:11:48,300 --> 01:11:49,420
in terms of spinning,

1156
01:11:50,240 --> 01:11:50,920
you know,

1157
01:11:50,920 --> 01:11:53,880
everything is high frequency.

1158
01:11:54,060 --> 01:11:55,300
Here I've got a signal source,

1159
01:11:55,500 --> 01:11:56,240
which is a sine wave.

1160
01:11:56,740 --> 01:11:59,240
The frequency is 0.2 hertz, okay?

1161
01:11:59,380 --> 01:12:01,360
So that phasor is rotating very slowly.

1162
01:12:01,580 --> 01:12:03,680
And we can very easily illustrate that.

1163
01:12:04,280 --> 01:12:05,580
This is an IQ plot.

1164
01:12:05,720 --> 01:12:07,480
It's a scope plot in XY mode.

1165
01:12:07,700 --> 01:12:09,780
And you can see the samples.

1166
01:12:10,820 --> 01:12:13,500
You know, this is a normal sine wave, right?

1167
01:12:15,900 --> 01:12:16,700
Where are you?

1168
01:12:17,740 --> 01:12:18,400
I get...

1169
01:12:18,400 --> 01:12:19,520
No, it takes a long time

1170
01:12:19,520 --> 01:12:20,260
because it's...

1171
01:12:20,260 --> 01:12:20,900
because it's...

1172
01:12:20,900 --> 01:12:21,260
slow.

1173
01:12:21,440 --> 01:12:22,060
There it is.

1174
01:12:22,160 --> 01:12:23,620
Let's go, I think, strip chart.

1175
01:12:26,580 --> 01:12:27,420
Yeah, there it is.

1176
01:12:27,560 --> 01:12:30,040
So we have the normal phasor,

1177
01:12:30,080 --> 01:12:32,180
the normal sine wave that's being generated.

1178
01:12:32,460 --> 01:12:33,440
But if you look at it on XY,

1179
01:12:34,660 --> 01:12:36,840
then it's just going to be a phase

1180
01:12:36,840 --> 01:12:38,580
that's rotating at a fixed rate

1181
01:12:38,580 --> 01:12:41,280
around your unit circle on the complex plane.

1182
01:12:41,600 --> 01:12:43,320
So what I want you to think about

1183
01:12:43,320 --> 01:12:44,360
is exactly this

1184
01:12:44,360 --> 01:12:47,760
in a single range bin, okay?

1185
01:12:47,880 --> 01:12:49,520
So you've got your echo coming back,

1186
01:12:49,540 --> 01:12:50,460
it's changing in phase,

1187
01:12:50,460 --> 01:12:51,880
and so you'll get a bit of this

1188
01:12:51,880 --> 01:12:53,200
in that range bin.

1189
01:12:53,700 --> 01:12:54,580
And of course,

1190
01:12:54,600 --> 01:12:56,220
what happens when you take the FFT

1191
01:12:56,220 --> 01:12:58,060
of a sine wave in this case?

1192
01:12:58,560 --> 01:13:01,060
You'll get a response in some bin

1193
01:13:01,060 --> 01:13:02,720
that maps to the frequency of the thing.

1194
01:13:02,960 --> 01:13:04,100
So what does that imply?

1195
01:13:04,220 --> 01:13:05,100
Can anybody tell me?

1196
01:13:05,520 --> 01:13:07,340
Once you end up taking the FFTs of the columns,

1197
01:13:07,600 --> 01:13:09,840
maybe I'll show you the next slide

1198
01:13:09,840 --> 01:13:11,000
and somebody can tell me.

1199
01:13:15,080 --> 01:13:16,340
So we did all that.

1200
01:13:19,640 --> 01:13:20,040
Okay.

1201
01:13:20,040 --> 01:13:20,760
Okay, so we...

1202
01:13:22,360 --> 01:13:24,180
Oh, microphone stopped working?

1203
01:13:26,820 --> 01:13:27,260
Hello?

1204
01:13:27,440 --> 01:13:28,380
Oh, there we go, testing.

1205
01:13:30,140 --> 01:13:31,120
It's not even wireless,

1206
01:13:31,140 --> 01:13:32,080
so we couldn't have blamed somebody

1207
01:13:32,080 --> 01:13:33,260
trying to dust my mic.

1208
01:13:35,260 --> 01:13:35,700
Right.

1209
01:13:38,260 --> 01:13:39,720
So consider this, then.

1210
01:13:41,060 --> 01:13:43,280
Each row here is a single FFT

1211
01:13:43,280 --> 01:13:45,820
that we took on an echo that came back, right?

1212
01:13:46,660 --> 01:13:48,180
Over our integration period,

1213
01:13:48,300 --> 01:13:49,600
we build up multiple rows.

1214
01:13:50,180 --> 01:13:51,380
So these are all the transforms.

1215
01:13:51,500 --> 01:13:52,980
Here, we're just taking an eight-point transform.

1216
01:13:53,980 --> 01:13:56,260
And so we build up each row

1217
01:13:56,260 --> 01:13:57,460
and have a number of columns

1218
01:13:57,460 --> 01:14:00,060
that maps to our integration period.

1219
01:14:01,380 --> 01:14:03,020
Then, to get the Doppler information,

1220
01:14:03,520 --> 01:14:04,240
what do you do?

1221
01:14:07,500 --> 01:14:08,320
You look at the face,

1222
01:14:08,340 --> 01:14:09,860
but what's the next step here?

1223
01:14:10,360 --> 01:14:11,280
I mean, I've already said it,

1224
01:14:11,300 --> 01:14:12,400
so I'm hoping somebody was listening.

1225
01:14:13,140 --> 01:14:14,880
You take FFTs of what?

1226
01:14:15,540 --> 01:14:16,240
The columns.

1227
01:14:16,660 --> 01:14:19,760
So you build up this memory cell, right?

1228
01:14:20,040 --> 01:14:21,660
And then once you've built up

1229
01:14:21,660 --> 01:14:23,380
and you've populated all of these

1230
01:14:23,380 --> 01:14:27,360
with the complex outputs of the first FFT part

1231
01:14:27,360 --> 01:14:28,400
of the range information,

1232
01:14:28,780 --> 01:14:31,120
you take the FFTs column-wise.

1233
01:14:32,040 --> 01:14:33,620
And so what you're doing is

1234
01:14:33,620 --> 01:14:34,560
you're doing an FFT

1235
01:14:34,560 --> 01:14:37,160
to look at the phase changes of the returns

1236
01:14:37,160 --> 01:14:38,980
at each range bin.

1237
01:14:39,600 --> 01:14:41,160
And what does that mean?

1238
01:14:41,760 --> 01:14:43,320
The Doppler shift

1239
01:14:43,320 --> 01:14:46,080
and the movement of the target

1240
01:14:46,080 --> 01:14:47,280
imparts the phase change,

1241
01:14:47,520 --> 01:14:50,020
and that causes that phase alteration.

1242
01:14:50,660 --> 01:14:52,760
and the faster or slower,

1243
01:14:53,580 --> 01:14:55,820
and whether it's forward or reversed,

1244
01:14:56,040 --> 01:14:57,360
rotating in that phase or,

1245
01:14:58,000 --> 01:15:00,620
tells you how fast the object is moving.

1246
01:15:01,340 --> 01:15:03,660
And so if you have two targets

1247
01:15:03,660 --> 01:15:06,100
that end up ambiguously,

1248
01:15:07,200 --> 01:15:08,900
ending up in the same range bin,

1249
01:15:09,060 --> 01:15:11,260
but they're maybe flying in different directions

1250
01:15:11,260 --> 01:15:12,100
or different speeds,

1251
01:15:12,700 --> 01:15:14,720
you can use Doppler processing

1252
01:15:14,720 --> 01:15:16,600
to then disambiguate them.

1253
01:15:16,860 --> 01:15:19,460
So instead of then just having a 1D plot,

1254
01:15:19,460 --> 01:15:20,420
you end up with a 2D plot,

1255
01:15:20,500 --> 01:15:21,360
and I'll show you what that looks like.

1256
01:15:22,280 --> 01:15:24,840
So you run the FFTs for each of these columns.

1257
01:15:25,620 --> 01:15:27,380
And then you get,

1258
01:15:27,920 --> 01:15:28,320
oops.

1259
01:15:30,140 --> 01:15:32,300
Now, remember how I said remember this slide

1260
01:15:32,300 --> 01:15:33,420
earlier in the presentation?

1261
01:15:33,720 --> 01:15:35,580
What was I showing you when I showed that slide?

1262
01:15:36,800 --> 01:15:37,600
An interleaver.

1263
01:15:39,000 --> 01:15:39,700
Look familiar?

1264
01:15:41,120 --> 01:15:41,520
Interleaver?

1265
01:15:41,900 --> 01:15:44,040
So what you can do in GNU Radio is,

1266
01:15:44,440 --> 01:15:46,740
I just wrote a simple row column interleaver block,

1267
01:15:46,940 --> 01:15:49,360
and it takes the output of the first FFT state,

1268
01:15:49,360 --> 01:15:51,100
and then once it fills the interleaver,

1269
01:15:51,220 --> 01:15:52,900
it outputs everything into another FFT,

1270
01:15:53,300 --> 01:15:56,400
and then you've got all your data,

1271
01:15:56,600 --> 01:15:58,440
the range and velocity information.

1272
01:16:00,960 --> 01:16:02,080
Okay, so interleaver.

1273
01:16:02,260 --> 01:16:04,500
And so this is a neat little tweak

1274
01:16:04,500 --> 01:16:05,780
that I made to my interleaver.

1275
01:16:06,200 --> 01:16:07,520
Usually, with an interleaver,

1276
01:16:07,540 --> 01:16:09,000
you need to fill up the entire interleaver,

1277
01:16:09,060 --> 01:16:10,280
and then you can read it out, right?

1278
01:16:10,360 --> 01:16:11,360
So you fill up all the rows,

1279
01:16:11,500 --> 01:16:12,900
and then you read it out all the columns.

1280
01:16:13,220 --> 01:16:15,400
What I did with my interleaver is that you can set,

1281
01:16:16,940 --> 01:16:18,480
I can't remember what the variable is,

1282
01:16:18,560 --> 01:16:19,340
but you can actually go to my interleaver,

1283
01:16:19,340 --> 01:16:20,460
and you can cause it to read out more quickly.

1284
01:16:20,780 --> 01:16:23,760
So it'll actually fill up,

1285
01:16:23,880 --> 01:16:24,860
let's say there are eight rows,

1286
01:16:25,140 --> 01:16:27,040
instead of outputting all the data

1287
01:16:27,040 --> 01:16:28,020
once it's filled up eight rows,

1288
01:16:29,620 --> 01:16:31,100
it fills four rows,

1289
01:16:32,120 --> 01:16:33,120
moves everything up,

1290
01:16:33,200 --> 01:16:34,980
fills up the last four rows, moves everything up,

1291
01:16:35,000 --> 01:16:36,180
and then outputs all eight.

1292
01:16:36,480 --> 01:16:39,100
And then it fills the next four, outputs all eight,

1293
01:16:39,220 --> 01:16:40,600
and then fills up four, and output.

1294
01:16:40,740 --> 01:16:42,920
And so effectively you get twice the output

1295
01:16:42,920 --> 01:16:44,880
coming out of your interleaver, which means that

1296
01:16:44,880 --> 01:16:46,820
even though you might have a long integration period,

1297
01:16:46,940 --> 01:16:49,040
like 10 seconds, we don't want to be

1298
01:16:49,340 --> 01:16:51,400
waiting 10 seconds to update a pretty plot,

1299
01:16:51,980 --> 01:16:53,140
we want to update it much more frequently.

1300
01:16:53,380 --> 01:16:55,280
So you have your interleaver spit the data out

1301
01:16:55,280 --> 01:16:55,800
more frequently.

1302
01:16:56,840 --> 01:16:58,760
So the DCHIP signal goes into an FFT,

1303
01:16:58,920 --> 01:16:59,760
you get your range information,

1304
01:17:00,220 --> 01:17:02,560
goes into the interleaver, you do another FFT,

1305
01:17:02,800 --> 01:17:06,860
and then magically you get your velocity information.

1306
01:17:07,420 --> 01:17:08,940
So now what I'm going to show you is

1307
01:17:08,940 --> 01:17:10,700
the original audio demo

1308
01:17:11,100 --> 01:17:13,280
that I had with the laptop here,

1309
01:17:13,480 --> 01:17:15,440
and then that plot that I didn't show you

1310
01:17:15,440 --> 01:17:17,760
is going to be the output of all of this

1311
01:17:17,760 --> 01:17:18,760
in a 2D image.

1312
01:17:19,580 --> 01:17:22,360
That gives you the Doppler information.

1313
01:17:28,300 --> 01:17:34,640
Okay, where are we?

1314
01:17:34,820 --> 01:17:41,580
And I will take Nate's wonderful advice again.

1315
01:17:42,200 --> 01:17:43,240
Use my laptop,

1316
01:17:43,420 --> 01:17:44,040
speakers and microphone.

1317
01:17:45,300 --> 01:17:46,660
Let me just fix up the,

1318
01:17:46,660 --> 01:17:48,160
so see before,

1319
01:17:48,840 --> 01:17:50,740
this is the DC component,

1320
01:17:50,800 --> 01:17:51,620
it's actually wrapped around,

1321
01:17:51,860 --> 01:17:54,120
because the microphone on the audio card

1322
01:17:54,120 --> 01:17:55,400
starts sampling at a different time

1323
01:17:55,400 --> 01:17:57,580
to when the output,

1324
01:17:59,200 --> 01:18:00,460
the digital Allen converter started,

1325
01:18:00,680 --> 01:18:02,740
they're offset in frequency space

1326
01:18:02,740 --> 01:18:04,120
because they're offset in time,

1327
01:18:04,280 --> 01:18:06,520
and so I need to adjust for that

1328
01:18:06,520 --> 01:18:07,280
manually myself.

1329
01:18:07,780 --> 01:18:08,880
Let me just do that.

1330
01:18:11,380 --> 01:18:11,860
Here.

1331
01:18:15,320 --> 01:18:15,800
Okay.

1332
01:18:17,420 --> 01:18:18,220
All right.

1333
01:18:23,540 --> 01:18:24,840
All right, there's that.

1334
01:18:24,960 --> 01:18:26,080
It should look familiar to you.

1335
01:18:26,280 --> 01:18:29,060
And now let me show you the magic here.

1336
01:18:29,160 --> 01:18:29,840
Hopefully it'll work.

1337
01:18:31,280 --> 01:18:32,120
All right, here we go.

1338
01:18:32,760 --> 01:18:37,560
So this is the Doppler plot.

1339
01:18:38,360 --> 01:18:39,820
You're going to see a little bit more of this

1340
01:18:39,820 --> 01:18:41,300
because I'm just going to show you

1341
01:18:41,300 --> 01:18:43,980
some brief previous experiments after this.

1342
01:18:44,120 --> 01:18:45,340
What I want you to think about here

1343
01:18:45,340 --> 01:18:45,780
is that the Doppler plot

1344
01:18:45,780 --> 01:18:47,980
is that the waterfall behind it

1345
01:18:47,980 --> 01:18:49,340
is showing you, effectively,

1346
01:18:49,520 --> 01:18:51,120
virtual range over time, right?

1347
01:18:51,580 --> 01:18:52,300
The Doppler plot,

1348
01:18:52,460 --> 01:18:54,740
the vertical axis here

1349
01:18:54,740 --> 01:18:56,320
is the range information.

1350
01:18:57,060 --> 01:18:58,580
The horizontal axis

1351
01:18:58,580 --> 01:19:00,080
is the velocity information.

1352
01:19:00,820 --> 01:19:03,400
So stuff there,

1353
01:19:03,640 --> 01:19:04,840
actually you see as I'm moving around

1354
01:19:04,840 --> 01:19:07,660
I'm actually causing the echo to change.

1355
01:19:07,980 --> 01:19:09,960
So it's surprisingly sensitive.

1356
01:19:12,160 --> 01:19:13,240
So yeah.

1357
01:19:14,140 --> 01:19:15,700
See how we've got the strong DC

1358
01:19:15,780 --> 01:19:17,420
part there on the left?

1359
01:19:17,520 --> 01:19:19,440
That maps to vertically that strong

1360
01:19:19,440 --> 01:19:20,560
bright spot at the top there.

1361
01:19:21,060 --> 01:19:23,000
So what I'm going to do is I might

1362
01:19:23,000 --> 01:19:24,540
need to change

1363
01:19:26,240 --> 01:19:27,740
the color mapping

1364
01:19:27,740 --> 01:19:28,160
here.

1365
01:19:32,160 --> 01:19:32,900
Yeah, that's better.

1366
01:19:33,060 --> 01:19:34,900
All right, let's see how this works.

1367
01:19:35,100 --> 01:19:36,840
So remember, what you'll see is

1368
01:19:37,480 --> 01:19:38,580
I'm going to be moving

1369
01:19:40,680 --> 01:19:41,640
the target

1370
01:19:41,640 --> 01:19:42,620
at a particular velocity.

1371
01:19:42,840 --> 01:19:44,860
Maybe up and down, a little bit slower,

1372
01:19:45,780 --> 01:19:47,900
and you'll see a bright spot appear there

1373
01:19:47,900 --> 01:19:49,740
either on the left or on the right,

1374
01:19:49,760 --> 01:19:51,140
depending if I'm going forward or backward.

1375
01:19:51,380 --> 01:19:53,240
And the distance from the center point

1376
01:19:53,240 --> 01:19:55,660
is going to imply how quickly I'm moving it.

1377
01:19:56,360 --> 01:19:57,280
Let's see if this works.

1378
01:19:59,380 --> 01:20:01,160
So there's my target. It's a bright spot, right?

1379
01:20:01,340 --> 01:20:03,020
And notice that I moved it here.

1380
01:20:04,040 --> 01:20:05,720
And now it's settled, so I'm going to move it more.

1381
01:20:05,980 --> 01:20:07,860
See, now it's on the left because I'm moving it up.

1382
01:20:09,180 --> 01:20:10,040
And now I'm going to stop

1383
01:20:10,040 --> 01:20:11,540
and it's going to come back because I've stopped moving it.

1384
01:20:11,700 --> 01:20:13,040
Now you still get the range return.

1385
01:20:13,560 --> 01:20:15,700
See the bright spot in the middle on the center line?

1386
01:20:15,780 --> 01:20:17,660
But I'm not moving it, so there's no doppel...

1387
01:20:17,660 --> 01:20:19,960
Well, I am, actually, because I can't keep it totally steady.

1388
01:20:20,240 --> 01:20:22,500
But it's that sensitive that even as it's waving here,

1389
01:20:22,540 --> 01:20:23,880
you can see that it's moving left and right.

1390
01:20:24,160 --> 01:20:25,440
So if I go like this,

1391
01:20:26,460 --> 01:20:28,180
then you get the doppel response.

1392
01:20:28,520 --> 01:20:30,080
See? And that's just using the audio.

1393
01:20:30,580 --> 01:20:31,700
Now, who can...

1394
01:20:32,160 --> 01:20:33,580
Apart from the multiple echoes,

1395
01:20:33,980 --> 01:20:36,260
who can tell me why we're getting multiple echoes here?

1396
01:20:38,840 --> 01:20:40,240
Wait. Skylar, hang on a second.

1397
01:20:40,360 --> 01:20:41,880
Anyone else with a hand?

1398
01:20:42,780 --> 01:20:43,540
Guy at the back?

1399
01:20:43,980 --> 01:20:45,440
Audience member 2 in the back.

1400
01:20:45,780 --> 01:20:48,120
audience member 2 in the back.

1401
01:20:48,760 --> 01:20:49,280
audience member 2 in the back.

1402
01:20:49,280 --> 01:20:50,600
Say it anyway.

1403
01:20:51,200 --> 01:20:52,700
audience member 2 in the back.

1404
01:20:53,680 --> 01:20:54,920
Yeah, because at least...

1405
01:20:55,420 --> 01:20:56,720
because we didn't have a microphone.

1406
01:20:56,840 --> 01:20:59,660
But what the two gentlemen were saying

1407
01:20:59,660 --> 01:21:00,720
was that

1408
01:21:01,340 --> 01:21:02,660
they're so close that

1409
01:21:02,660 --> 01:21:05,100
the audio's going to bounce off the bottom of my laptop,

1410
01:21:05,160 --> 01:21:06,360
it's going to enter the microphone,

1411
01:21:06,480 --> 01:21:08,280
but it's also going to bounce off my bottom laptop

1412
01:21:08,280 --> 01:21:09,600
and it'll keep bouncing back and forth.

1413
01:21:09,780 --> 01:21:11,300
So we get the multiple returns there.

1414
01:21:11,300 --> 01:21:12,780
The strongest one's going to be the direct...

1415
01:21:12,780 --> 01:21:14,120
you know, the shortest part.

1416
01:21:14,540 --> 01:21:16,360
The other thing is look carefully

1417
01:21:16,360 --> 01:21:17,620
what happens to the bright spot

1418
01:21:17,620 --> 01:21:19,900
in relation to the edges of this image

1419
01:21:19,900 --> 01:21:20,920
as I move it, right?

1420
01:21:21,100 --> 01:21:23,340
I'm going to move it very slowly to keep it in there

1421
01:21:23,340 --> 01:21:25,420
and maybe a bit further out so we don't get all the returns.

1422
01:21:26,120 --> 01:21:27,540
Right? So now it's going to be on the left.

1423
01:21:27,700 --> 01:21:29,380
I'm going to move it fast and watch what happens.

1424
01:21:33,060 --> 01:21:33,720
Who said that?

1425
01:21:34,320 --> 01:21:35,560
That's right. It's aliasing.

1426
01:21:35,980 --> 01:21:37,060
Why are we getting aliasing?

1427
01:21:46,240 --> 01:21:47,360
It's aliasing.

1428
01:21:47,360 --> 01:21:48,600
Yeah, that's exactly right.

1429
01:21:48,800 --> 01:21:51,280
So what is happening here

1430
01:21:51,280 --> 01:21:53,000
is that the pulse

1431
01:21:53,000 --> 01:21:54,320
repetition frequency, right?

1432
01:21:54,360 --> 01:21:55,540
How often we send out pulses

1433
01:21:56,800 --> 01:21:57,360
and

1434
01:22:01,460 --> 01:22:03,120
that'll basically constrain us

1435
01:22:03,120 --> 01:22:04,280
in terms of

1436
01:22:05,020 --> 01:22:06,820
the maximum unambiguous

1437
01:22:07,580 --> 01:22:09,280
velocity that you can make out.

1438
01:22:09,560 --> 01:22:11,240
So once you exceed that velocity

1439
01:22:11,240 --> 01:22:12,900
just like with a normal signal

1440
01:22:12,900 --> 01:22:15,020
if it's above your Nyquist

1441
01:22:15,020 --> 01:22:17,020
divided by 2, or Nyquist

1442
01:22:17,020 --> 01:22:19,180
if it's complex, it will then

1443
01:22:19,180 --> 01:22:20,840
wrap around because

1444
01:22:20,840 --> 01:22:23,240
you can't disambiguate above that.

1445
01:22:23,320 --> 01:22:24,360
It's exactly the same principle.

1446
01:22:24,600 --> 01:22:26,840
So this is another constraint that's applied to

1447
01:22:26,840 --> 01:22:28,980
the radar system depending upon the parameters

1448
01:22:28,980 --> 01:22:30,080
that you pick initially.

1449
01:22:30,400 --> 01:22:32,960
So you have to, as a radar designer

1450
01:22:32,960 --> 01:22:35,020
define what your

1451
01:22:36,000 --> 01:22:37,040
maximum unambiguous

1452
01:22:37,920 --> 01:22:39,060
range should be

1453
01:22:39,060 --> 01:22:41,300
and then you have to

1454
01:22:41,300 --> 01:22:42,480
balance that with your

1455
01:22:42,910 --> 01:22:44,480
maximum unambiguous velocity.

1456
01:22:44,840 --> 01:22:45,840
Because you can't have both.

1457
01:22:46,060 --> 01:22:48,260
It's got to be either one way or the other.

1458
01:22:51,100 --> 01:22:51,600
Yeah.

1459
01:22:51,680 --> 01:22:52,460
So that's

1460
01:22:53,060 --> 01:22:54,040
basically the...

1461
01:22:55,760 --> 01:22:56,880
I need the sound on.

1462
01:22:57,060 --> 01:22:58,820
Again, very good advice from the audience.

1463
01:23:01,440 --> 01:23:03,120
So any questions about this?

1464
01:23:04,420 --> 01:23:04,920
No?

1465
01:23:05,080 --> 01:23:07,000
So just a reminder then, remember how

1466
01:23:07,060 --> 01:23:09,060
we were basically doing the

1467
01:23:09,060 --> 01:23:10,700
interleaving and doing the FFT?

1468
01:23:10,840 --> 01:23:12,340
That plot that we saw

1469
01:23:12,340 --> 01:23:14,060
is basically the output

1470
01:23:14,840 --> 01:23:16,720
of that entire process.

1471
01:23:17,000 --> 01:23:17,900
So each of those

1472
01:23:17,900 --> 01:23:20,080
each of the rows

1473
01:23:20,660 --> 01:23:22,380
was an output of

1474
01:23:22,850 --> 01:23:23,520
an FFT.

1475
01:23:27,540 --> 01:23:28,540
So that's the audio.

1476
01:23:29,340 --> 01:23:30,840
It's almost time so

1477
01:23:31,400 --> 01:23:32,920
if you're interested I'll show you

1478
01:23:32,920 --> 01:23:33,980
an SDR version.

1479
01:23:35,100 --> 01:23:36,020
I've got some

1480
01:23:37,060 --> 01:23:38,040
directional antennas here.

1481
01:23:40,240 --> 01:23:41,200
I'll try and show you

1482
01:23:41,200 --> 01:23:42,300
because what I've done is

1483
01:23:42,300 --> 01:23:44,760
I take the bins

1484
01:23:44,760 --> 01:23:46,540
that map to the zero range

1485
01:23:46,540 --> 01:23:49,020
because remember I was saying light travels really fast, right?

1486
01:23:49,200 --> 01:23:50,720
So even though I think I'm using

1487
01:23:50,720 --> 01:23:52,480
1 megahertz sample rate

1488
01:23:52,480 --> 01:23:54,980
and so 1 over 1 million

1489
01:23:54,980 --> 01:23:57,220
is still going to be 100 meters

1490
01:23:57,220 --> 01:23:58,680
or 60 meters or whatever it is.

1491
01:23:59,520 --> 01:24:00,240
And so

1492
01:24:00,860 --> 01:24:02,520
we wouldn't be able to demonstrate

1493
01:24:03,060 --> 01:24:04,100
a change in range

1494
01:24:04,100 --> 01:24:06,960
but we can demonstrate Doppler because Doppler will still

1495
01:24:07,060 --> 01:24:08,400
work in the zero bin

1496
01:24:08,400 --> 01:24:09,780
if we're nice and close.

1497
01:24:10,080 --> 01:24:12,880
And what I do is I then do an inverse transform

1498
01:24:12,880 --> 01:24:14,320
of that FFT at the zero bin

1499
01:24:14,320 --> 01:24:16,120
and then I put that to the speakers

1500
01:24:16,120 --> 01:24:17,420
and it makes like a theremin

1501
01:24:17,420 --> 01:24:19,900
so we can wave something back and forth

1502
01:24:19,900 --> 01:24:21,460
and turn it into a musical instrument.

1503
01:24:21,800 --> 01:24:24,180
But I'll show you that last in case we run out of time.

1504
01:24:24,360 --> 01:24:25,680
You can stick around if you want to see it.

1505
01:24:26,700 --> 01:24:27,880
Right, back to the slides.

1506
01:24:32,040 --> 01:24:32,680
Okay, so

1507
01:24:33,300 --> 01:24:35,020
that's stuff that you can do at home

1508
01:24:35,020 --> 01:24:36,740
and I want to show you some stuff

1509
01:24:37,060 --> 01:24:38,780
that's used in practice and how you can actually

1510
01:24:39,660 --> 01:24:40,720
decode that stuff as well.

1511
01:24:41,060 --> 01:24:42,660
So one is CODAR

1512
01:24:42,660 --> 01:24:45,240
which is HF radar used to map

1513
01:24:45,240 --> 01:24:46,860
the surface of the ocean

1514
01:24:46,860 --> 01:24:48,300
and map currents

1515
01:24:48,300 --> 01:24:50,580
and you can go on various universities websites

1516
01:24:50,580 --> 01:24:52,980
that have these set up and look at interactive

1517
01:24:52,980 --> 01:24:53,320
live

1518
01:24:54,600 --> 01:24:56,500
plots of currents over the ocean.

1519
01:24:56,660 --> 01:24:58,320
The idea is that you send out an HF wave

1520
01:24:58,320 --> 01:24:59,820
it'll hit crests of

1521
01:24:59,820 --> 01:25:02,840
an RF wave, it'll hit the crests of ocean waves

1522
01:25:02,840 --> 01:25:04,840
which will reflect them back to the receiver

1523
01:25:04,840 --> 01:25:06,900
and those ocean waves will also have a

1524
01:25:07,060 --> 01:25:08,220
velocity component to them

1525
01:25:08,220 --> 01:25:10,780
which will then change the phase and once you do

1526
01:25:10,780 --> 01:25:12,300
doppler integration over time

1527
01:25:12,300 --> 01:25:14,480
then you can both get range and

1528
01:25:14,960 --> 01:25:16,540
velocity information of ocean waves.

1529
01:25:17,260 --> 01:25:18,960
And then they have them

1530
01:25:18,960 --> 01:25:19,960
distributed in certain

1531
01:25:21,440 --> 01:25:22,480
ways and

1532
01:25:23,020 --> 01:25:24,320
they did something interesting which is

1533
01:25:24,840 --> 01:25:26,580
gating the transmitter

1534
01:25:26,580 --> 01:25:28,840
so if you look on an ordinary plot

1535
01:25:28,840 --> 01:25:30,300
it looks like it's just a chirp

1536
01:25:30,300 --> 01:25:32,620
but if you look very closely with the right parameters

1537
01:25:32,620 --> 01:25:34,860
you can actually see they gate the transmitter

1538
01:25:34,860 --> 01:25:36,840
on and off to give it that dead time

1539
01:25:37,060 --> 01:25:37,800
to receive a return

1540
01:25:38,840 --> 01:25:40,840
and the side effect of this is that

1541
01:25:40,840 --> 01:25:43,040
when you take an FFT you actually end up getting

1542
01:25:43,040 --> 01:25:44,780
sidebands on your main signal

1543
01:25:44,780 --> 01:25:46,760
and you might think they're returns but they're

1544
01:25:46,760 --> 01:25:48,820
just a result of doing the

1545
01:25:48,820 --> 01:25:51,100
frequency transform of this gated

1546
01:25:51,100 --> 01:25:52,780
waveform and

1547
01:25:52,780 --> 01:25:54,440
if you do the math

1548
01:25:54,900 --> 01:25:56,940
they're like millions of kilometers away in terms

1549
01:25:56,940 --> 01:25:58,840
of range and so the stuff that you want to see

1550
01:25:58,840 --> 01:26:00,020
is really really close in.

1551
01:26:01,680 --> 01:26:02,900
So that's what it looks like

1552
01:26:02,900 --> 01:26:04,580
and you've got your main chirp and

1553
01:26:04,580 --> 01:26:07,000
return hidden in the middle and those

1554
01:26:07,060 --> 01:26:08,960
symmetric lines going out

1555
01:26:08,960 --> 01:26:10,840
are the

1556
01:26:10,840 --> 01:26:12,660
AM sidebands that you can forget about.

1557
01:26:12,940 --> 01:26:14,740
So I showed this before

1558
01:26:15,280 --> 01:26:17,040
there are a bunch of HF

1559
01:26:17,040 --> 01:26:18,480
CODAR stations running on the

1560
01:26:18,840 --> 01:26:20,760
spectrum all the time so you can pick any one of them

1561
01:26:20,760 --> 01:26:22,980
I basically talked about all that

1562
01:26:24,180 --> 01:26:25,140
and a shout out

1563
01:26:25,140 --> 01:26:26,680
again to Peter Bellings and

1564
01:26:26,680 --> 01:26:28,940
Mo Wheatley, Iris Space SpectraView

1565
01:26:28,940 --> 01:26:30,640
actually Peter made these

1566
01:26:30,640 --> 01:26:32,960
Vivaldi antennas, really cool if you want to check them out

1567
01:26:33,480 --> 01:26:34,260
after the talk

1568
01:26:34,260 --> 01:26:35,760
and they've used

1569
01:26:37,060 --> 01:26:38,800
SpectraView and his SDRs

1570
01:26:38,800 --> 01:26:40,120
to produce these amazing plots

1571
01:26:40,120 --> 01:26:42,860
where you're not actually looking at the ocean returns

1572
01:26:42,860 --> 01:26:44,760
anymore, you're looking at the returns

1573
01:26:44,760 --> 01:26:46,820
of the ionosphere because it's HF

1574
01:26:46,820 --> 01:26:47,560
and will still

1575
01:26:48,760 --> 01:26:50,380
travel all the way up to the ionosphere

1576
01:26:50,380 --> 01:26:52,020
and come back on a return

1577
01:26:52,480 --> 01:26:54,500
and then this is one of

1578
01:26:55,140 --> 01:26:56,220
Iris Space's tweets

1579
01:26:57,880 --> 01:26:58,940
really really cool stuff

1580
01:26:59,580 --> 01:27:01,140
and then this is in SpectraView

1581
01:27:01,140 --> 01:27:02,960
producing on the plot so in this way you can

1582
01:27:02,960 --> 01:27:04,700
use the signal in an unintended

1583
01:27:04,700 --> 01:27:07,040
way to look at natural phenomena

1584
01:27:07,060 --> 01:27:08,780
that are going around the earth

1585
01:27:08,780 --> 01:27:10,660
and another

1586
01:27:11,360 --> 01:27:13,120
very very amazing

1587
01:27:13,120 --> 01:27:14,300
SDR guy, Juha

1588
01:27:15,560 --> 01:27:16,120
Vierian

1589
01:27:17,920 --> 01:27:19,000
he's done incredible

1590
01:27:19,000 --> 01:27:21,240
stuff, he's used this kind of

1591
01:27:21,240 --> 01:27:22,980
UHF radar to map

1592
01:27:22,980 --> 01:27:24,600
the surface of the moon

1593
01:27:24,600 --> 01:27:27,100
amazing paper, he's done something

1594
01:27:27,100 --> 01:27:28,420
really cool stuff with ionosons

1595
01:27:29,060 --> 01:27:30,800
I've tried my hand at this

1596
01:27:30,800 --> 01:27:33,260
and so you also get those multiple returns

1597
01:27:33,860 --> 01:27:35,340
this is sort of frequency over time

1598
01:27:35,340 --> 01:27:37,040
as an ionoson sweeps the ionosphere

1599
01:27:37,060 --> 01:27:37,880
across the entire HF spectrum

1600
01:27:37,880 --> 01:27:39,820
you can similarly de-chirp it

1601
01:27:39,820 --> 01:27:40,840
because it's just a chirp

1602
01:27:40,840 --> 01:27:43,160
and then get these multiple reflections

1603
01:27:43,160 --> 01:27:44,020
from the ionosphere

1604
01:27:46,280 --> 01:27:48,560
we went out hiking

1605
01:27:49,320 --> 01:27:49,880
and

1606
01:27:50,820 --> 01:27:52,440
I thought after the hike

1607
01:27:53,760 --> 01:27:54,700
my wife

1608
01:27:54,700 --> 01:27:56,480
was very kind and she just

1609
01:27:56,480 --> 01:27:57,640
read her book in the car

1610
01:27:57,640 --> 01:27:59,420
I set my laptop up in the boot

1611
01:27:59,420 --> 01:28:02,060
and received Kodar

1612
01:28:02,060 --> 01:28:03,260
from around that area

1613
01:28:03,260 --> 01:28:06,120
you may recall me showing this previously

1614
01:28:07,060 --> 01:28:07,140
so this is the

1615
01:28:07,780 --> 01:28:09,000
x-axis is time

1616
01:28:10,100 --> 01:28:11,260
y-axis is frequency

1617
01:28:11,260 --> 01:28:12,960
and so this is the Kodar over time

1618
01:28:12,960 --> 01:28:14,560
and you can see the AM side bands again

1619
01:28:14,560 --> 01:28:16,800
and there's some information hidden in there

1620
01:28:16,800 --> 01:28:17,920
we're only interested in the center one

1621
01:28:17,920 --> 01:28:19,140
if you zoom in

1622
01:28:19,820 --> 01:28:22,700
in this case I don't think I had a GPS DO

1623
01:28:22,700 --> 01:28:24,240
they're GPS disciplines

1624
01:28:24,240 --> 01:28:25,400
so you want to have a GPS 2

1625
01:28:25,400 --> 01:28:27,040
so that the phase information is correct

1626
01:28:27,040 --> 01:28:29,120
because remember if you're running your interleaver

1627
01:28:29,120 --> 01:28:29,920
and the doppler processing

1628
01:28:29,920 --> 01:28:32,320
you want to make sure everything is perfectly aligned and stable

1629
01:28:37,300 --> 01:28:38,740
so there's something going on there

1630
01:28:38,740 --> 01:28:39,380
I didn't know where it was

1631
01:28:40,580 --> 01:28:41,420
again you know

1632
01:28:41,420 --> 01:28:44,420
that's range information effectively in the middle

1633
01:28:44,420 --> 01:28:46,740
so you're getting returns somehow

1634
01:28:47,440 --> 01:28:48,820
and then with Ian Buckley

1635
01:28:48,820 --> 01:28:50,380
we went out one night and set up

1636
01:28:50,380 --> 01:28:52,460
and did some captures with a GPS DO

1637
01:28:53,360 --> 01:28:53,760
and

1638
01:28:55,180 --> 01:28:56,760
you know we had a long wire set up

1639
01:28:56,760 --> 01:28:57,960
and a custom made dipole

1640
01:28:57,960 --> 01:28:59,280
I'm going to show you a video here

1641
01:28:59,280 --> 01:29:02,240
this is again same kind of output from GNU Radio

1642
01:29:02,240 --> 01:29:03,480
using that same plotting thing

1643
01:29:03,480 --> 01:29:06,720
but it's a modified SEL sync but also outputs

1644
01:29:07,060 --> 01:29:07,520
a series of bitmaps

1645
01:29:08,620 --> 01:29:11,100
and so this is the same

1646
01:29:11,100 --> 01:29:12,920
everything that I've spoken about so far

1647
01:29:12,920 --> 01:29:14,180
exactly the same process

1648
01:29:14,180 --> 01:29:16,500
but applied to the HF CODA

1649
01:29:17,140 --> 01:29:19,540
returns that I just received on my SDR

1650
01:29:19,540 --> 01:29:21,080
now the disclaimer is

1651
01:29:21,080 --> 01:29:23,360
I have no idea exactly what we're looking at

1652
01:29:23,360 --> 01:29:25,100
I mean obviously there's range and velocity

1653
01:29:25,100 --> 01:29:25,880
information in there

1654
01:29:25,880 --> 01:29:28,280
and you can see that there's some cool stuff happening

1655
01:29:28,280 --> 01:29:29,320
like there are these

1656
01:29:30,240 --> 01:29:31,280
returns here

1657
01:29:31,280 --> 01:29:34,840
and in terms of velocity which is on the y-axis

1658
01:29:34,840 --> 01:29:36,100
things are moving

1659
01:29:37,060 --> 01:29:38,060
direction away and toward

1660
01:29:39,000 --> 01:29:40,040
the receiver

1661
01:29:40,040 --> 01:29:42,840
so there's stuff happening and there are weaker

1662
01:29:42,840 --> 01:29:43,800
responses out there

1663
01:29:44,380 --> 01:29:47,020
I don't know whether this is purely from the ocean waves

1664
01:29:47,020 --> 01:29:48,280
or it's purely from the ionosphere

1665
01:29:48,280 --> 01:29:50,960
I meant to do the calculations but I didn't get a chance

1666
01:29:50,960 --> 01:29:52,300
if you know

1667
01:29:52,900 --> 01:29:54,380
what's contributing to this

1668
01:29:55,060 --> 01:29:56,220
I'd be really curious

1669
01:29:57,020 --> 01:29:58,780
but you know you can basically

1670
01:29:58,780 --> 01:30:00,600
take the parameters of the radar system and calculate

1671
01:30:00,600 --> 01:30:02,440
how far those returns are

1672
01:30:03,040 --> 01:30:04,160
and what the implications are

1673
01:30:04,160 --> 01:30:06,820
and you know it's this information that they

1674
01:30:07,060 --> 01:30:07,760
take in the CODA system

1675
01:30:07,760 --> 01:30:11,080
and then map to figuring out

1676
01:30:11,080 --> 01:30:11,960
the ocean currents

1677
01:30:11,960 --> 01:30:14,340
from information gathered from multiple points

1678
01:30:14,340 --> 01:30:15,420
like you saw on that map before

1679
01:30:15,420 --> 01:30:17,740
so you can generate cool plots

1680
01:30:17,740 --> 01:30:19,560
the last thing I want to show you is

1681
01:30:19,560 --> 01:30:21,580
the television version with airplanes

1682
01:30:22,140 --> 01:30:24,240
so this is using passive radar again

1683
01:30:24,240 --> 01:30:26,520
so with CODA I was using someone else's signal

1684
01:30:26,840 --> 01:30:28,060
in this case I'm using

1685
01:30:28,060 --> 01:30:31,560
digital ATSC terrestrial TV signal

1686
01:30:31,560 --> 01:30:33,040
and remember

1687
01:30:33,740 --> 01:30:35,280
like I said with the CHIRP

1688
01:30:35,280 --> 01:30:37,040
if you have a fixed thing that you're looking at

1689
01:30:37,060 --> 01:30:38,380
and you want to run through your filter

1690
01:30:39,060 --> 01:30:40,960
in the case of ATSC

1691
01:30:40,960 --> 01:30:41,840
happily

1692
01:30:41,840 --> 01:30:44,600
in the middle between picture information

1693
01:30:44,600 --> 01:30:47,120
there is this synchronization sequence

1694
01:30:47,120 --> 01:30:48,960
of a fixed length and it's documented

1695
01:30:48,960 --> 01:30:50,980
so this now becomes

1696
01:30:50,980 --> 01:30:52,440
the filter

1697
01:30:52,900 --> 01:30:54,960
not FMCW CHIRP anymore

1698
01:30:54,960 --> 01:30:56,800
but the standards based

1699
01:30:58,080 --> 01:30:59,200
synchronization sequence

1700
01:30:59,200 --> 01:31:00,220
for ATSC

1701
01:31:00,220 --> 01:31:03,060
that we end up filtering and correlating on

1702
01:31:03,980 --> 01:31:04,960
and this is

1703
01:31:04,960 --> 01:31:07,040
in all TV signals and what's available

1704
01:31:07,060 --> 01:31:08,780
what's nice is they use incredibly powerful

1705
01:31:08,780 --> 01:31:10,120
transmitters that'll

1706
01:31:11,060 --> 01:31:12,120
broadcast the signal

1707
01:31:12,120 --> 01:31:13,320
in a broad range

1708
01:31:14,560 --> 01:31:16,560
I had this set up first in my house

1709
01:31:16,560 --> 01:31:18,840
where I had this directional antenna pointing out to a highway

1710
01:31:19,720 --> 01:31:20,820
this is the view from

1711
01:31:20,820 --> 01:31:22,720
our place, there's the 280 highway

1712
01:31:22,720 --> 01:31:24,340
in San Francisco that runs there

1713
01:31:25,160 --> 01:31:26,660
cars are moving back and forth

1714
01:31:26,660 --> 01:31:28,740
they might end up being decent radar returns

1715
01:31:28,740 --> 01:31:30,880
and so you can see how they're moving there

1716
01:31:31,480 --> 01:31:32,840
one direction of traffic on top

1717
01:31:32,840 --> 01:31:33,640
the other on the bottom

1718
01:31:33,640 --> 01:31:35,840
they're metallic so we might get some returns

1719
01:31:38,100 --> 01:31:38,900
and that's

1720
01:31:39,820 --> 01:31:40,960
that's the receiver

1721
01:31:40,960 --> 01:31:42,880
site, the transmitter site

1722
01:31:42,880 --> 01:31:45,020
is over the hill which is nice because we don't get

1723
01:31:45,020 --> 01:31:46,200
the direct path from the transmitter

1724
01:31:46,200 --> 01:31:48,940
you get the reflections from stuff

1725
01:31:48,940 --> 01:31:51,180
that's and the direct path is attenuated

1726
01:31:51,180 --> 01:31:53,020
so this is where

1727
01:31:53,020 --> 01:31:54,760
I want to illustrate the power of Doppler processing

1728
01:31:54,760 --> 01:31:57,140
because in this case

1729
01:31:57,140 --> 01:31:59,100
Doppler and velocity is on the Y

1730
01:31:59,100 --> 01:32:00,160
and range is on the X

1731
01:32:00,660 --> 01:32:02,860
what this is doing is it's correlating that

1732
01:32:02,860 --> 01:32:05,100
known synchronization sequence in the signal

1733
01:32:05,100 --> 01:32:06,040
and then

1734
01:32:07,180 --> 01:32:08,220
that initial

1735
01:32:08,220 --> 01:32:10,600
the strongest return from

1736
01:32:10,600 --> 01:32:11,920
the filter for that

1737
01:32:11,920 --> 01:32:13,400
ends up being in the center there

1738
01:32:13,400 --> 01:32:16,320
so you can ignore everything to the left because that's all picture information

1739
01:32:16,860 --> 01:32:18,400
and then stuff that's here

1740
01:32:18,860 --> 01:32:20,920
this is basically time zero in terms of

1741
01:32:20,920 --> 01:32:21,840
the synchronization signal

1742
01:32:21,840 --> 01:32:24,700
now if we weren't doing any

1743
01:32:24,700 --> 01:32:26,680
Doppler processing, if we didn't care about velocity

1744
01:32:26,680 --> 01:32:29,040
and we only got the range information

1745
01:32:29,040 --> 01:32:30,860
look at this range line

1746
01:32:30,860 --> 01:32:31,800
the green line

1747
01:32:32,340 --> 01:32:34,340
there's nothing discernible going on there

1748
01:32:34,340 --> 01:32:36,000
you will get reflections

1749
01:32:37,060 --> 01:32:38,780
off static structures that are large

1750
01:32:38,780 --> 01:32:39,960
and will reflect a lot of the signal

1751
01:32:39,960 --> 01:32:42,480
but if you're talking about cars and aeroplanes

1752
01:32:42,480 --> 01:32:44,500
the radar cross section is so small

1753
01:32:44,930 --> 01:32:47,300
they won't be able to send any appreciable signal

1754
01:32:47,300 --> 01:32:49,040
back to that small directional antenna

1755
01:32:49,040 --> 01:32:50,620
so there's nothing there

1756
01:32:50,620 --> 01:32:52,540
once you add velocity information

1757
01:32:53,050 --> 01:32:54,520
you can see that green is

1758
01:32:54,520 --> 01:32:56,300
we go from white to red to green

1759
01:32:56,300 --> 01:32:59,340
it's the color spectrum

1760
01:32:59,340 --> 01:33:00,840
the black and the blue there

1761
01:33:01,700 --> 01:33:03,360
are basically down the other end

1762
01:33:03,360 --> 01:33:05,220
of your power range

1763
01:33:05,220 --> 01:33:06,360
still within the

1764
01:33:07,060 --> 01:33:08,300
dynamic range of your ADC

1765
01:33:08,300 --> 01:33:10,840
but it's much more easy to

1766
01:33:10,840 --> 01:33:11,500
discern stuff

1767
01:33:11,500 --> 01:33:13,240
so actually that's the target right there

1768
01:33:13,240 --> 01:33:14,240
I'll go back to the beginning

1769
01:33:14,240 --> 01:33:16,220
oh look did you see

1770
01:33:17,140 --> 01:33:18,880
that's not on my display

1771
01:33:18,880 --> 01:33:20,540
let me try that one more time

1772
01:33:21,620 --> 01:33:23,920
so at the beginning here some funky stuff happens

1773
01:33:23,920 --> 01:33:25,060
as the interleaver fills up

1774
01:33:25,060 --> 01:33:26,820
because there's still empty data in the interleaver

1775
01:33:26,820 --> 01:33:28,020
and it kind of settles

1776
01:33:29,380 --> 01:33:31,260
and just look here and you'll see

1777
01:33:31,260 --> 01:33:33,080
little spots appear

1778
01:33:33,700 --> 01:33:36,180
and I'm pretty sure they're cars on the highway

1779
01:33:37,060 --> 01:33:37,940
so that's the TV signal

1780
01:33:37,940 --> 01:33:39,040
there's another one there

1781
01:33:39,040 --> 01:33:40,440
and then they appear very briefly

1782
01:33:40,440 --> 01:33:41,920
because if you think of them like a mirror

1783
01:33:41,920 --> 01:33:44,280
you want to end up seeing the specular reflection

1784
01:33:44,280 --> 01:33:47,400
off the car and the car has to be at the right angle

1785
01:33:47,400 --> 01:33:48,560
you know the door or the windshield

1786
01:33:48,560 --> 01:33:50,560
or whatever it is has to be at the right angle

1787
01:33:51,020 --> 01:33:52,520
both with respect to my receiver

1788
01:33:52,520 --> 01:33:54,720
and the TV transmitter to produce the reflection

1789
01:33:55,840 --> 01:33:57,580
another important point here is that

1790
01:33:57,580 --> 01:33:58,780
I have a GPSDO

1791
01:33:58,780 --> 01:34:01,720
displaying the oscillator in my receiver

1792
01:34:01,720 --> 01:34:03,440
but that's still not good enough

1793
01:34:03,440 --> 01:34:04,700
there is still some phase offset

1794
01:34:07,060 --> 01:34:08,000
a series of blocks I have

1795
01:34:08,000 --> 01:34:10,180
that measure the number of samples

1796
01:34:10,180 --> 01:34:13,160
between each successive primary return

1797
01:34:13,160 --> 01:34:15,440
of that ATSC synchronization sequence

1798
01:34:15,440 --> 01:34:19,060
and there's a polyphase resampler

1799
01:34:19,060 --> 01:34:23,480
but it uses a long double ratio

1800
01:34:23,480 --> 01:34:24,220
which is

1801
01:34:24,220 --> 01:34:26,580
it's not a float, it's not a double

1802
01:34:26,580 --> 01:34:27,720
it's a long double

1803
01:34:27,720 --> 01:34:29,640
because there are so many decimal points back

1804
01:34:29,640 --> 01:34:32,180
where it has to adjust the rate

1805
01:34:32,180 --> 01:34:34,060
to be locked on exactly to the signal

1806
01:34:34,060 --> 01:34:36,460
and not have the range information wander

1807
01:34:37,320 --> 01:34:38,760
and then you have to track the phase

1808
01:34:38,760 --> 01:34:41,180
of the primary and it gets a bit messy

1809
01:34:41,180 --> 01:34:41,740
if you want to do it well

1810
01:34:41,740 --> 01:34:44,480
but you can see there's stuff going on in there

1811
01:34:45,380 --> 01:34:47,020
now to prove

1812
01:34:47,020 --> 01:34:48,660
this a little bit more with the airplanes

1813
01:34:50,160 --> 01:34:51,120
Ian and I went out

1814
01:34:51,120 --> 01:34:51,780
to this car park

1815
01:34:52,560 --> 01:34:53,720
and we parked there

1816
01:34:53,720 --> 01:34:55,780
and this is in this park here

1817
01:34:56,400 --> 01:34:57,680
which happens to be here

1818
01:34:58,120 --> 01:35:00,540
right under almost the approach path

1819
01:35:00,540 --> 01:35:02,780
for planes coming to land at San Francisco

1820
01:35:02,780 --> 01:35:04,640
so the geometry there is really nice

1821
01:35:04,640 --> 01:35:05,780
we've got a TV transmitter

1822
01:35:07,060 --> 01:35:09,080
elsewhere in the north of the bay

1823
01:35:09,080 --> 01:35:10,620
and then we're down here

1824
01:35:11,300 --> 01:35:13,300
receiving the reflections quite close

1825
01:35:13,300 --> 01:35:14,160
to the aircraft

1826
01:35:14,900 --> 01:35:16,560
and that's the kind of setup there

1827
01:35:16,560 --> 01:35:18,200
we're just using a normal TV antenna

1828
01:35:18,200 --> 01:35:19,200
plugged into the USIRP

1829
01:35:20,120 --> 01:35:23,260
and then that's the stuff going on there

1830
01:35:24,020 --> 01:35:25,180
so I've got a couple of videos

1831
01:35:25,180 --> 01:35:27,180
and we'll see some interesting things

1832
01:35:27,180 --> 01:35:28,000
so that's a plane

1833
01:35:28,000 --> 01:35:28,680
see that

1834
01:35:31,480 --> 01:35:32,720
and it's aliasing

1835
01:35:32,720 --> 01:35:34,840
because we can't control

1836
01:35:34,840 --> 01:35:36,100
the pulse repetition frequency

1837
01:35:37,060 --> 01:35:38,240
we can't control the constraints

1838
01:35:38,240 --> 01:35:39,480
of unambiguous velocity

1839
01:35:39,480 --> 01:35:41,660
because that's defined by the ATC spec

1840
01:35:42,820 --> 01:35:44,120
but you can see

1841
01:35:45,320 --> 01:35:46,260
these returns

1842
01:35:46,900 --> 01:35:47,780
that are coming off planes

1843
01:35:47,780 --> 01:35:48,660
and what's interesting is

1844
01:35:50,280 --> 01:35:51,900
in terms of range

1845
01:35:51,900 --> 01:35:53,940
these ones are actually further out

1846
01:35:53,940 --> 01:35:55,460
we wouldn't see them again

1847
01:35:55,460 --> 01:35:57,840
if we didn't do the doppler processing

1848
01:35:57,840 --> 01:35:58,960
but because we do

1849
01:35:59,760 --> 01:36:02,620
we basically ignore the zero velocity line

1850
01:36:02,620 --> 01:36:03,620
and these targets are moving

1851
01:36:03,620 --> 01:36:06,620
so they make themselves apparent on the rest of the plot

1852
01:36:07,720 --> 01:36:09,240
let me go to the next one

1853
01:36:15,560 --> 01:36:17,320
some of them are more obvious than others

1854
01:36:17,320 --> 01:36:17,980
I think they get

1855
01:36:18,660 --> 01:36:20,780
so you can see there that line

1856
01:36:20,780 --> 01:36:22,940
and the reason why it's strict like that

1857
01:36:22,940 --> 01:36:24,460
remember I'm reading out of my interleaver

1858
01:36:24,460 --> 01:36:26,120
faster than the integration period

1859
01:36:26,120 --> 01:36:29,160
so you basically end up sort of smushing together

1860
01:36:29,680 --> 01:36:30,400
smoothing together

1861
01:36:30,910 --> 01:36:32,280
one integration period with the next

1862
01:36:32,280 --> 01:36:33,960
and so you'll basically get the blending

1863
01:36:33,960 --> 01:36:35,740
of velocity information as you go

1864
01:36:40,460 --> 01:36:42,020
I haven't quite got my head around that

1865
01:36:43,860 --> 01:36:46,640
it's some artifact of the system

1866
01:36:46,640 --> 01:36:47,820
oh look there's something there

1867
01:36:49,820 --> 01:36:50,940
yeah I have to think about that

1868
01:36:50,940 --> 01:36:51,560
a bit more

1869
01:36:51,560 --> 01:36:52,700
it wasn't nearly obvious to me

1870
01:36:53,640 --> 01:36:54,440
why that's there

1871
01:36:55,720 --> 01:36:56,840
any ideas from anyone?

1872
01:37:00,300 --> 01:37:01,880
yeah so there's a return down there

1873
01:37:01,880 --> 01:37:03,200
I don't know

1874
01:37:03,200 --> 01:37:04,500
I don't know

1875
01:37:04,500 --> 01:37:05,000
I don't know

1876
01:37:05,000 --> 01:37:05,940
I don't know

1877
01:37:05,940 --> 01:37:06,380
I don't know

1878
01:37:06,380 --> 01:37:06,680
I don't know

1879
01:37:07,340 --> 01:37:08,200
I don't know

1880
01:37:08,200 --> 01:37:10,500
some of them are really obvious

1881
01:37:10,500 --> 01:37:12,120
what was interesting is that

1882
01:37:12,120 --> 01:37:13,980
the planes that are overhead are obviously

1883
01:37:13,980 --> 01:37:14,820
going to be very close

1884
01:37:14,820 --> 01:37:17,800
I think one pixel here is 30 meters

1885
01:37:18,820 --> 01:37:19,580
somewhere around there

1886
01:37:19,580 --> 01:37:20,300
or maybe 40 meters

1887
01:37:21,600 --> 01:37:24,760
and so they all end up being quite close

1888
01:37:24,760 --> 01:37:26,480
especially as they're flying in overhead

1889
01:37:27,060 --> 01:37:28,880
but what really blew my mind later

1890
01:37:28,880 --> 01:37:29,300
is that

1891
01:37:29,300 --> 01:37:31,760
I can't remember which of these videos it is unfortunately

1892
01:37:31,760 --> 01:37:32,780
it might just appear

1893
01:37:32,780 --> 01:37:35,020
but there are dots that appeared further out

1894
01:37:37,160 --> 01:37:38,500
with slowly changing velocity.

1895
01:37:38,880 --> 01:37:40,640
And as it turned out, we looked over to the bay,

1896
01:37:40,740 --> 01:37:41,740
and of course there's San Francisco,

1897
01:37:41,940 --> 01:37:43,380
but what's on the other side of the bay?

1898
01:37:44,220 --> 01:37:45,580
Oakland is on the other side of the bay.

1899
01:37:45,660 --> 01:37:48,360
And there's a plane that was just smoothly cruising

1900
01:37:48,360 --> 01:37:50,700
into land at Oakland, but because of the geometry,

1901
01:37:51,820 --> 01:37:54,240
it was headed in a slightly different direction

1902
01:37:54,240 --> 01:37:57,560
than our planes coming into SFO,

1903
01:37:57,740 --> 01:38:00,080
and so it appeared differently on the plot,

1904
01:38:00,140 --> 01:38:01,660
both in terms of velocity and range.

1905
01:38:04,520 --> 01:38:06,400
Oh yeah, so there's something out there.

1906
01:38:12,740 --> 01:38:16,480
Yeah, so any questions so far?

1907
01:38:19,600 --> 01:38:21,240
So you saw that streak there.

1908
01:38:23,000 --> 01:38:25,100
I think this one might be a decent one.

1909
01:38:30,700 --> 01:38:31,140
Question?

1910
01:38:31,360 --> 01:38:34,420
If your plane's spinning on the same SDR,

1911
01:38:34,680 --> 01:38:36,080
do you not need a DCS?

1912
01:38:36,160 --> 01:38:36,860
Is that all there?

1913
01:38:37,840 --> 01:38:38,460
Correct, yeah.

1914
01:38:38,580 --> 01:38:39,960
So if you're transmitting and receiving

1915
01:38:39,960 --> 01:38:40,980
from the same piece of hardware,

1916
01:38:42,060 --> 01:38:43,540
everything's inherently going to be synchronized,

1917
01:38:43,560 --> 01:38:45,480
so you don't need to worry about using a GPS DO.

1918
01:38:46,300 --> 01:38:48,360
And I'll show you that if you want to stick around.

1919
01:38:54,180 --> 01:39:01,860
You mean that vertical line there?

1920
01:39:02,940 --> 01:39:03,980
So I think at the moment,

1921
01:39:05,320 --> 01:39:06,740
because there's a lot of artifacts happen,

1922
01:39:07,060 --> 01:39:08,940
during that disciplining process,

1923
01:39:09,200 --> 01:39:10,260
not the GPS DO one,

1924
01:39:10,340 --> 01:39:11,480
it's the other one that I talked about

1925
01:39:11,480 --> 01:39:13,220
where it measures the number of samples

1926
01:39:13,220 --> 01:39:15,580
between two of the primary returns

1927
01:39:15,580 --> 01:39:16,620
and it measures the phase.

1928
01:39:16,740 --> 01:39:18,080
There's some additional processing going on there,

1929
01:39:18,160 --> 01:39:20,280
and as it's doing that very, very fine adjustment

1930
01:39:20,280 --> 01:39:21,160
of the resampling,

1931
01:39:21,700 --> 01:39:24,500
it produces all these artifacts in the FFTs,

1932
01:39:24,500 --> 01:39:27,600
and that's why you get crap in there.

1933
01:39:29,660 --> 01:39:30,460
Maybe, question?

1934
01:39:30,640 --> 01:39:32,800
What's the frequency of the FFTs?

1935
01:39:34,360 --> 01:39:36,920
So I don't remember the exact frequency of this television,

1936
01:39:36,920 --> 01:39:39,640
but it's, you know, there are a plethora of stations

1937
01:39:39,640 --> 01:39:41,500
throughout the UHF band.

1938
01:39:42,720 --> 01:39:44,340
And then, based on the frequency,

1939
01:39:44,540 --> 01:39:47,660
that has an implication on how the radar cross-section will look,

1940
01:39:47,740 --> 01:39:49,940
so everything is interrelated there,

1941
01:39:50,040 --> 01:39:51,460
and here we're constrained, obviously,

1942
01:39:51,520 --> 01:39:53,520
by the station frequency and the spec.

1943
01:39:55,800 --> 01:39:58,320
So anyway, that brings me to the end of the talk.

1944
01:39:58,520 --> 01:40:00,880
If you want to hang around, I'll show you the SDR version.

1945
01:40:01,420 --> 01:40:03,180
If you want to hang around now, I'll show you.

1946
01:40:03,220 --> 01:40:04,080
It'll just take a couple of minutes.

1947
01:40:04,220 --> 01:40:05,600
Or you can clap now and head to lunch.

1948
01:40:05,660 --> 01:40:06,340
Totally up to you.

1949
01:40:07,420 --> 01:40:08,440
All right, I'll do it quickly.

1950
01:40:09,560 --> 01:40:12,100
So here, we've got,

1951
01:40:13,480 --> 01:40:14,780
let me find it here.

1952
01:40:17,140 --> 01:40:18,560
Oh, one thing I want to show you quickly.

1953
01:40:18,780 --> 01:40:23,680
This is the live processing of the coder, right?

1954
01:40:23,740 --> 01:40:25,680
So the HF, HF1.

1955
01:40:27,560 --> 01:40:29,780
So this should look familiar from the video.

1956
01:40:31,780 --> 01:40:33,140
What I wanted to show you here is,

1957
01:40:33,180 --> 01:40:35,360
in this plot, you know how I was talking about

1958
01:40:35,360 --> 01:40:36,540
usually when we take FFTs,

1959
01:40:36,920 --> 01:40:37,520
we always take the magnitude

1960
01:40:37,520 --> 01:40:39,280
and we always forget the phase information?

1961
01:40:39,740 --> 01:40:42,600
And this plot is showing you phase information only.

1962
01:40:43,000 --> 01:40:44,700
So consider it like a waterfall,

1963
01:40:44,900 --> 01:40:47,280
but instead of showing the magnitude, it's phase.

1964
01:40:47,620 --> 01:40:50,400
And remember how we had those interesting shapes coming up

1965
01:40:50,400 --> 01:40:51,980
and, you know, they're a strong return,

1966
01:40:52,220 --> 01:40:53,840
so there's going to be some information there?

1967
01:40:54,140 --> 01:40:55,580
This looks like noise,

1968
01:40:55,800 --> 01:40:58,600
because there's nothing strong there.

1969
01:40:58,780 --> 01:41:01,700
But amazingly, you can see this phase structure

1970
01:41:02,400 --> 01:41:04,920
in these range bins in the return.

1971
01:41:05,360 --> 01:41:06,900
So, you know, it's a good thing.

1972
01:41:07,620 --> 01:41:10,240
I haven't, I'd never looked at a phase plot like that before,

1973
01:41:10,380 --> 01:41:12,580
a phase waterfall, and it kind of looks cool.

1974
01:41:16,000 --> 01:41:16,440
Yeah.

1975
01:41:16,900 --> 01:41:19,680
So let me find the, I think it's this one.

1976
01:41:22,040 --> 01:41:23,320
I think it's this one.

1977
01:41:25,620 --> 01:41:26,280
Try this.

1978
01:41:26,880 --> 01:41:28,680
So this is two mega samples per second.

1979
01:41:29,480 --> 01:41:30,740
Oh, let me show you that,

1980
01:41:35,440 --> 01:41:36,320
that workbook.

1981
01:41:42,320 --> 01:41:46,740
No, no, no, there's another thing.

1982
01:41:46,740 --> 01:41:47,280
I watch it all the time.

1983
01:41:47,280 --> 01:41:47,920
Are there any movies in it?

1984
01:41:48,000 --> 01:41:56,160
Ah, we've got a little girl band.

1985
01:42:00,860 --> 01:42:01,320
Right.

1986
01:42:01,320 --> 01:42:01,620
But it's a two mega sample,

1987
01:42:02,680 --> 01:42:03,700
but it's fairly expensive.

1988
01:42:03,700 --> 01:42:05,160
It's, I think there's a 90 percent variety

1989
01:42:05,960 --> 01:42:06,900
What you can see is three美 oops,

1990
01:42:06,900 --> 01:42:11,460
You put in your pulse repetition frequency, you know, you basically can plug in anything

1991
01:42:11,850 --> 01:42:15,240
you want into the equation and then it will output all this information.

1992
01:42:15,620 --> 01:42:18,560
So for a given frequency, it gives you the wavelength, the range resolution.

1993
01:42:19,040 --> 01:42:21,740
For a given pulse repetition frequency, it will give you pulse duration.

1994
01:42:22,680 --> 01:42:27,140
Given bandwidth, it will give you an ambiguous range, an ambiguous Doppler, an ambiguous

1995
01:42:27,140 --> 01:42:33,380
velocity and then, you know, you can figure out the scales on all of these pictures from

1996
01:42:33,380 --> 01:42:33,860
that.

1997
01:42:36,060 --> 01:42:37,380
Let's see what's happening here.

1998
01:42:39,000 --> 01:42:40,160
Oh, it's running, cool.

1999
01:42:42,580 --> 01:42:50,620
So what I've got is I've got this flow graph here and then I've got the plot here.

2000
01:42:52,720 --> 01:42:57,100
Nothing is showing which is a little bit, oh, no, of course, nothing is showing.

2001
01:43:00,020 --> 01:43:03,620
Skyler correctly points out, thank you, another audience contribution.

2002
01:43:03,620 --> 01:43:08,700
The default configuration here is everything is off because I don't want to blow anything

2003
01:43:08,700 --> 01:43:08,920
up.

2004
01:43:09,940 --> 01:43:12,140
So, and I've got these two directional antennas here.

2005
01:43:13,960 --> 01:43:15,780
One is the transmitter and one is the receiver.

2006
01:43:16,060 --> 01:43:24,060
So I can put them next to each other and I'm just going to do a very quick setup and put

2007
01:43:24,060 --> 01:43:24,860
my bag in between.

2008
01:43:25,240 --> 01:43:27,120
I didn't, hopefully that won't affect things.

2009
01:43:27,820 --> 01:43:32,640
And I'm using a B200 Mini here in a custom 3D-printed case.

2010
01:43:33,160 --> 01:43:33,600
So I'm going to put it here.

2011
01:43:33,600 --> 01:43:38,700
And let me try doing this.

2012
01:43:39,660 --> 01:43:42,440
This is really arbitrary.

2013
01:43:43,880 --> 01:43:44,960
Something like that.

2014
01:43:45,760 --> 01:43:46,340
Okay.

2015
01:43:47,360 --> 01:43:49,300
And now, no guarantees.

2016
01:43:50,260 --> 01:43:55,200
It works also better if you're using separate transmit-receive USURP.

2017
01:43:55,300 --> 01:43:59,480
So, for example, I did it with a USURP with a MIMO cable so you can synchronize the clocks,

2018
01:43:59,660 --> 01:44:03,380
but the actual boards are different and they won't interfere with them.

2019
01:44:03,440 --> 01:44:03,580
Okay.

2020
01:44:05,540 --> 01:44:08,100
So, I'm going to do a simple setup.

2021
01:44:08,100 --> 01:44:08,600
I'm going to set the gain.

2022
01:44:08,980 --> 01:44:10,160
I think I use about 55.

2023
01:44:11,140 --> 01:44:12,940
Same gain I'll go, I don't know, 33.

2024
01:44:13,640 --> 01:44:18,720
And then I'm going to set my, sorry, I can bring it over here.

2025
01:44:20,460 --> 01:44:22,380
My amplitude of the signal of .7.

2026
01:44:22,800 --> 01:44:27,500
And what we've got now is, hopefully this is familiar to you, the spectrum is being

2027
01:44:27,500 --> 01:44:27,900
rotated.

2028
01:44:28,020 --> 01:44:30,220
So this is the FFT of the output of the de-chirping.

2029
01:44:30,380 --> 01:44:33,580
And we've got the tone there, which is what the signal is going to be.

2030
01:44:33,580 --> 01:44:38,260
The system is hearing, it's the DC tone, it's hearing itself, basically.

2031
01:44:38,660 --> 01:44:43,280
Unfortunately, that's really high, and it shouldn't be that high, because we won't be

2032
01:44:43,280 --> 01:44:46,220
able to resolve anything otherwise.

2033
01:44:51,040 --> 01:44:51,600
Okay.

2034
01:44:58,320 --> 01:45:00,080
It might be...

2035
01:45:02,940 --> 01:45:04,060
Too much here.

2036
01:45:04,260 --> 01:45:05,280
Oh, there it is, see?

2037
01:45:05,800 --> 01:45:08,600
See how we've got the center line, which is actually offset?

2038
01:45:09,880 --> 01:45:12,080
I haven't figured out why that's offset yet.

2039
01:45:13,180 --> 01:45:20,640
But as I move my laptop toward and away from the receiver, you can see that even though

2040
01:45:20,640 --> 01:45:24,260
it's in the zero range and it's still overloading, you still get velocity information.

2041
01:45:26,140 --> 01:45:29,660
And so that's, you know, that's basically how RF...

2042
01:45:30,140 --> 01:45:30,780
Radars work.

2043
01:45:31,380 --> 01:45:33,660
Now, let's see if...

2044
01:45:36,380 --> 01:45:37,820
Because I promised you music.

2045
01:45:38,760 --> 01:45:39,200
Music.

2046
01:45:39,500 --> 01:45:39,940
Music.

2047
01:45:41,400 --> 01:45:42,840
Is it actually...

2048
01:45:43,380 --> 01:45:47,520
Yeah, let me plug in the speaker there.

2049
01:45:47,660 --> 01:45:55,780
So it's a little bit finicky in that I need to do that, and then I need to change the

2050
01:45:55,780 --> 01:45:59,040
offset here so that it ends up going into the zero bin.

2051
01:46:00,080 --> 01:46:02,160
Because that's the one that's being transformed.

2052
01:46:11,020 --> 01:46:13,020
Do I have the audio actually on?

2053
01:46:16,480 --> 01:46:17,080
I do.

2054
01:46:17,640 --> 01:46:18,280
Sorry, what was that?

2055
01:46:18,700 --> 01:46:19,840
Your speaker volume is down.

2056
01:46:19,940 --> 01:46:20,720
My speaker volume is down.

2057
01:46:22,480 --> 01:46:23,100
There we go.

2058
01:46:23,620 --> 01:46:24,340
Thank you.

2059
01:46:24,340 --> 01:46:26,040
Another audience participation night.

2060
01:46:26,300 --> 01:46:27,740
What would I do without you all?

2061
01:46:28,680 --> 01:46:29,120
Thank you.

2062
01:46:30,080 --> 01:46:30,380
Thank you.

2063
01:46:30,500 --> 01:46:30,660
Thank you.

2064
01:46:30,660 --> 01:46:30,940
Thank you.