WEBVTT 1 00:00:00.120 --> 00:00:03.799 Welcome to the deep dive. Today, we're really getting into 2 00:00:03.839 --> 00:00:09.919 the invisible world of radio frequency identification are FID. That's right. 3 00:00:09.919 --> 00:00:12.640 I probably use it all the time, maybe without even 4 00:00:12.640 --> 00:00:15.960 thinking about it, you know, tracking packages, tapping your payment cards. 5 00:00:16.000 --> 00:00:17.280 Absolutely, it's everywhere. 6 00:00:17.679 --> 00:00:20.320 So our plan today is to peel back the layers 7 00:00:20.359 --> 00:00:23.719 on how these systems actually work. We're focusing specifically on 8 00:00:23.920 --> 00:00:30.039 the ultra high frequency UHF and super high frequency SAHF bans. 9 00:00:29.760 --> 00:00:32.560 Exactly, and our mission really is to cut through some 10 00:00:32.640 --> 00:00:36.079 pretty dense technical stuff, the documents that define this field, 11 00:00:36.359 --> 00:00:37.960 and pull out the key insights for you. 12 00:00:38.159 --> 00:00:39.600 Yeah, make it makes sense, right. 13 00:00:39.799 --> 00:00:42.320 We want you to understand the core ideas, some surprising 14 00:00:42.320 --> 00:00:46.439 physics involved, the real world engineering hurdles, and frankly, the 15 00:00:46.479 --> 00:00:50.560 clever solutions that make contactless communication possible, so you can 16 00:00:50.600 --> 00:00:52.039 be genuinely well informed. 17 00:00:52.200 --> 00:00:55.799 Okay, let's start with the absolute basics. Then we're moving 18 00:00:55.840 --> 00:00:58.159 past things like barcodes, right, things that need line of 19 00:00:58.200 --> 00:00:59.320 sight or actual. 20 00:00:58.960 --> 00:01:02.759 Contact, correct, stepping into using radio frequency waves. It's all 21 00:01:02.759 --> 00:01:04.280 about contactless communication. 22 00:01:04.719 --> 00:01:07.519 So how does that start. What are the essential pieces 23 00:01:07.760 --> 00:01:09.040 of an RFID system. 24 00:01:09.239 --> 00:01:12.000 Well, at its heart, you've got four main parts. First, 25 00:01:12.000 --> 00:01:14.280 there's the remote element, that's what we usually just call 26 00:01:14.359 --> 00:01:16.480 the tag. It holds the data the little sticker or 27 00:01:16.560 --> 00:01:20.159 card precisely. Then you have the fixed element, the base 28 00:01:20.200 --> 00:01:23.400 station or interrogator that sends out the signal and listens 29 00:01:23.400 --> 00:01:24.200 for the tags reply. 30 00:01:24.560 --> 00:01:24.879 Okay. 31 00:01:25.159 --> 00:01:28.200 Third is the communication medium, which most of the time 32 00:01:28.280 --> 00:01:31.680 is just the air, right. And finally there's the host system. 33 00:01:32.079 --> 00:01:34.519 Think of that as the main computer that takes the 34 00:01:34.640 --> 00:01:37.439 data from the reader and actually uses it for whatever 35 00:01:37.480 --> 00:01:38.359 the application is. 36 00:01:38.599 --> 00:01:44.920 Got it tag reader air computer system? Sounds straightforward, But 37 00:01:45.959 --> 00:01:48.519 how far apart can the tag and reader be? You 38 00:01:48.560 --> 00:01:51.840 hear about warehouse tracking, Yeah, that must be pretty far. Ah. 39 00:01:52.000 --> 00:01:54.840 Yes, the operating range it varies a lot. Actually. We 40 00:01:54.879 --> 00:01:57.840 talk about proximity, which is really close, like centimeter tapping 41 00:01:57.840 --> 00:02:00.599 your card exactly, then vicinity, which be up to a 42 00:02:00.640 --> 00:02:04.519 meter or so, and long range, which is anything beyond that. 43 00:02:04.959 --> 00:02:08.560 So what kind of distances are we talking typically. 44 00:02:08.080 --> 00:02:10.879 Well, for those little passive tags, the ones without batteries, 45 00:02:11.120 --> 00:02:15.080 it could be anywhere from say tens of centimeters up 46 00:02:15.120 --> 00:02:18.879 to maybe several meters in good conditions. Okay, but if 47 00:02:18.919 --> 00:02:22.159 you have battery assisted tags, now those can really reach 48 00:02:22.199 --> 00:02:25.520 out maybe fifteen meters, sometimes even up to two hundred meters. 49 00:02:25.560 --> 00:02:28.840 Oh, two hundred meters, So warehouse tracking is definitely feasible then, 50 00:02:29.280 --> 00:02:31.960 but probably needs those battery tags. 51 00:02:31.599 --> 00:02:34.159 Often yes, or very specific system setups. 52 00:02:34.400 --> 00:02:37.960 So that brings up the power question. How does a 53 00:02:38.039 --> 00:02:42.080 tiny tag without a battery get any power? And how 54 00:02:42.080 --> 00:02:44.400 does it send its information back? Doesn't have a transmitter? 55 00:02:44.479 --> 00:02:46.439 Right, This is one of the coolest parts I think 56 00:02:46.759 --> 00:02:51.080 the energy transfer and communication modes. Those remotely powered or 57 00:02:51.120 --> 00:02:55.319 passive tags, they literally harvest energy from the reader's radio signal. Right, 58 00:02:55.360 --> 00:02:58.319 we air exactly like a tiny radio wave solar panel. 59 00:02:58.639 --> 00:03:01.680 They scavenge just an power to turn on their little chip. 60 00:03:01.759 --> 00:03:03.000 That's amazing, it really is. 61 00:03:03.080 --> 00:03:05.479 Now battery assist to tags, they have their own power, 62 00:03:05.479 --> 00:03:07.520 which is why they get that much better range. But 63 00:03:07.599 --> 00:03:10.520 here's the clever bit about sending data back. Whether it's 64 00:03:10.599 --> 00:03:13.879 passive or battery assisted, the tag usually doesn't have its 65 00:03:13.879 --> 00:03:15.919 own dedicated transmitter. 66 00:03:15.680 --> 00:03:16.759 So how does it talk back? 67 00:03:16.800 --> 00:03:21.159 Then it uses a technique called backscattering. The reader sends 68 00:03:21.199 --> 00:03:24.719 out a continuous, steady radio wave like a constant beam 69 00:03:24.759 --> 00:03:28.960 of light. Okay, The tag then changes its antenna characteristics electrically. 70 00:03:29.479 --> 00:03:33.240 It sort of flickers its own reflectivity to that incoming wave. 71 00:03:33.120 --> 00:03:35.039 Like a mirror flashing kind of. Yeah. 72 00:03:35.520 --> 00:03:39.159 By changing how much of the reader's signal it reflects back, 73 00:03:39.319 --> 00:03:43.000 it encodes its data onto that reflected wave. So it's 74 00:03:43.120 --> 00:03:46.039 riding the reader's wave back, but with its own message 75 00:03:46.120 --> 00:03:46.919 imprinted on it. 76 00:03:47.560 --> 00:03:50.599 Ingenious, Yeah, just modulating the reflection exactly. 77 00:03:50.719 --> 00:03:51.439 Backscattering. 78 00:03:51.479 --> 00:03:55.439 Okay, so that power harvesting and backscattering, that's pretty mind bending. 79 00:03:55.719 --> 00:03:59.520 But what about the specific physics. Why these higher frequencies 80 00:03:59.599 --> 00:04:02.039 UAH and SAHF. What's happening invisibly there? 81 00:04:02.400 --> 00:04:04.400 Right? This is where we get into some interesting physics. 82 00:04:04.400 --> 00:04:08.080 At UHF and SAHF. RFID systems mostly operate in what's 83 00:04:08.159 --> 00:04:11.039 called the far field. Yeah, think of it as the 84 00:04:11.199 --> 00:04:14.400 zone where the electromagnetic waves have sort of sorted themselves 85 00:04:14.400 --> 00:04:17.439 out and behave in a more predictable plane wave like manner. 86 00:04:17.720 --> 00:04:19.879 You have the electric field, the E field, and the 87 00:04:19.920 --> 00:04:23.319 magnetic field, the each field working together. The basic model 88 00:04:23.399 --> 00:04:27.120 comes from the Herzean dipole, but we don't need to 89 00:04:27.160 --> 00:04:28.120 get lost in that math. 90 00:04:28.360 --> 00:04:28.680 Okay. 91 00:04:29.000 --> 00:04:32.720 What's useful conceptually is the pointing vector. It basically tells 92 00:04:32.759 --> 00:04:35.839 engineers the direction and the density of the energy flow 93 00:04:35.879 --> 00:04:38.720 in those waves super important for figuring out if enough 94 00:04:38.839 --> 00:04:40.480 energy is actually getting. 95 00:04:40.240 --> 00:04:43.800 To the tag. So like aiming the energy precisely. 96 00:04:43.519 --> 00:04:47.800 And that leads to antenna gain. Antennas aren't just dumb radiators. 97 00:04:47.839 --> 00:04:51.319 They're designed to focus that energy, like a flashlight reflector 98 00:04:51.439 --> 00:04:53.439 focuses the light bulb's output into. 99 00:04:53.279 --> 00:04:55.720 A beam instead of just spreading it everywhere exactly. 100 00:04:56.240 --> 00:05:00.600 A theoretical isotropic antenna would radiate equally in all directions, 101 00:05:00.639 --> 00:05:04.439 which isn't very efficient. Real antennas are directional. They have 102 00:05:04.879 --> 00:05:06.879 gain focusing power where you need it. 103 00:05:07.360 --> 00:05:10.240 Now. I always kind of assumed higher frequency meant, you know, 104 00:05:10.319 --> 00:05:13.439 stronger signal, better range. Yeah, but you're saying that's not 105 00:05:13.480 --> 00:05:14.360 necessarily true. 106 00:05:14.519 --> 00:05:17.920 That's a really key point, and it's counterintuitive. It's about attenuation. 107 00:05:18.120 --> 00:05:22.120 In free space. Signals naturally get weaker over distance, but 108 00:05:22.240 --> 00:05:25.439 this loss is much much worse at higher frequencies. How 109 00:05:25.519 --> 00:05:28.279 much worse Well, the source material gives a striking example 110 00:05:28.920 --> 00:05:32.800 the signal loss. The attenuation is about seven times seven 111 00:05:32.839 --> 00:05:35.680 point four times actually greater. At two point four to 112 00:05:35.680 --> 00:05:38.600 five geta hurtz compared to nine hundred megaherts over the 113 00:05:38.639 --> 00:05:39.680 same distance. 114 00:05:39.399 --> 00:05:41.680 Seven times weaker just because the frequency is higher. 115 00:05:41.800 --> 00:05:45.720 Yep, it's a fundamental physics thing. So at higher frequencies 116 00:05:45.759 --> 00:05:48.720 you often need more transmit power or much higher antenna 117 00:05:48.720 --> 00:05:51.040 gain just to achieve the same range you'd get at 118 00:05:51.079 --> 00:05:51.959 a lower frequency. 119 00:05:52.240 --> 00:05:54.800 Wow. Okay, So if the signal is getting attenuated so 120 00:05:54.879 --> 00:05:58.560 much more, how on Earth does a tiny passive tag 121 00:05:58.839 --> 00:06:01.879 manage to screep together enough power to even turn on, 122 00:06:02.680 --> 00:06:04.759 especially at those higher sahf frequencies. 123 00:06:04.839 --> 00:06:08.000 It's genuinely impressive engineering. It's called power recovery. At the 124 00:06:08.000 --> 00:06:11.120 tag antenna, inside the tags chip, there are rectifire circuits. 125 00:06:11.160 --> 00:06:13.600 Often there's simple things like voltage doublers. 126 00:06:13.160 --> 00:06:14.240 And they convert the radio waves. 127 00:06:14.519 --> 00:06:17.600 They convert the tiny alternating current picked up by the 128 00:06:17.639 --> 00:06:20.920 tag's antenna from the radio wave into a usable direct 129 00:06:20.959 --> 00:06:23.560 current to power the chip. And these chips are designed 130 00:06:23.560 --> 00:06:26.519 to work on incredibly small amounts of power, like how 131 00:06:26.600 --> 00:06:29.959 small well. Examples given are like thirty five milliwatts needed 132 00:06:30.000 --> 00:06:32.600 at nine hundred megahertz or maybe one hundred and twenty 133 00:06:32.639 --> 00:06:35.439 miliwads at two point four to five gigahertz for certain 134 00:06:35.519 --> 00:06:38.560 chips to wake up and operate. It's minuscule. It demands 135 00:06:38.600 --> 00:06:39.720 incredible efficiency. 136 00:06:39.800 --> 00:06:42.079 That really puts the harvesting concept into perspective. 137 00:06:42.079 --> 00:06:45.040 They're sipping power absolutely just enough to do their job. 138 00:06:45.319 --> 00:06:47.879 Okay, but the real world isn't a perfect physics lab. 139 00:06:48.199 --> 00:06:52.160 It's messy. What happens when these carefully aimed, attenuated radio 140 00:06:52.199 --> 00:06:57.279 waves hit you know, stuff, walls, boxes, people, Ah. 141 00:06:56.920 --> 00:07:01.000 Yes, the environment, it's a huge factor. Materials interact with 142 00:07:01.079 --> 00:07:03.399 r F waves in all sorts of ways. Absorption is 143 00:07:03.399 --> 00:07:03.879 a big one. 144 00:07:03.920 --> 00:07:05.759 Things just soak up the signal pretty much. 145 00:07:05.879 --> 00:07:08.360 Water is a classic example. It's a major absorber and 146 00:07:08.399 --> 00:07:11.639 it also detunes the tag antenna. They did experiments putting 147 00:07:11.639 --> 00:07:14.920 tags in water and the resonant frequency shifted way down, 148 00:07:15.079 --> 00:07:17.000 like from nine hundred mitle herds to seven hundred and 149 00:07:17.000 --> 00:07:19.040 fifty middle herts, basically useless. 150 00:07:19.079 --> 00:07:20.399 And people are mostly water. 151 00:07:20.279 --> 00:07:24.399 So exactly reading tags on people or near liquids can 152 00:07:24.439 --> 00:07:27.519 be really challenging. Even things you wouldn't expect, like certain 153 00:07:27.560 --> 00:07:31.240 types of car windshields, leaded glass or heat absorbing glass 154 00:07:31.240 --> 00:07:34.240 can significantly block or detune the signal. 155 00:07:34.319 --> 00:07:35.920 Okay, so absorption is bad. 156 00:07:36.240 --> 00:07:40.920 What else reflection? R F waves bounce off surfaces, especially metal, 157 00:07:40.959 --> 00:07:46.079 but also walls, floors, everything, really and that causes problems. 158 00:07:46.120 --> 00:07:50.000 It can reflections can combine in unpredictable ways. Sometimes they 159 00:07:50.000 --> 00:07:54.079 add up. That's constructive interference, creating signal hot spots. 160 00:07:54.279 --> 00:07:55.959 It works really well and expectedly well. 161 00:07:56.040 --> 00:07:59.160 Yeah, but they can also cancel each other out destructive interference, 162 00:07:59.199 --> 00:08:02.319 creating signal black holes or nulls where the tag just 163 00:08:02.360 --> 00:08:03.399 can't be read at all. 164 00:08:03.680 --> 00:08:06.560 So the signal strength can vary wildly just by moving 165 00:08:06.600 --> 00:08:07.519 a few inches. 166 00:08:07.319 --> 00:08:10.879 Absolutely makes consistent reading really difficult. And then there's diffraction 167 00:08:11.079 --> 00:08:15.439 bending around objects and refraction bending through materials. It's complex. 168 00:08:15.519 --> 00:08:18.439 And what if the tag itself isn't facing the reader perfectly? 169 00:08:18.920 --> 00:08:20.879 Does the angle matter hugely? 170 00:08:21.319 --> 00:08:25.920 That's tag polarization loss. The electromagnetic wave from the reader 171 00:08:25.959 --> 00:08:30.600 has a specific orientation, a polarization. The tag's antenna is 172 00:08:30.600 --> 00:08:34.000 designed to receive that best at a certain angle it's tilted. 173 00:08:34.120 --> 00:08:36.919 If it's misaligned, the amount of power the tag antenna 174 00:08:36.960 --> 00:08:41.559 effectively captures drops significantly. The source material points out tags 175 00:08:41.559 --> 00:08:45.080 are hardly ever positioned physically at the optimal angle. 176 00:08:45.240 --> 00:08:46.279 Makes sense in the real world. 177 00:08:46.440 --> 00:08:49.519 Yeah, and this relates to the tag's radar cross section 178 00:08:49.720 --> 00:08:52.799 or rcs. That's a measure of how effectively the tag 179 00:08:52.840 --> 00:08:56.120 reflects that power back to the reader, and crucially for 180 00:08:56.240 --> 00:09:00.000 backscatter systems, it's the tag's ability to change its rs 181 00:09:00.000 --> 00:09:03.799 A dynamic RCS or DRCS that allows it to encode data. 182 00:09:03.919 --> 00:09:08.320 So, with all these challenges absorption, reflection, dead spots, polarization issues, 183 00:09:08.759 --> 00:09:11.600 how do companies claim they can read like hundreds of 184 00:09:11.600 --> 00:09:13.840 thousands of tags per second on a palette? That sounds 185 00:09:13.840 --> 00:09:14.639 almost impossible? 186 00:09:14.679 --> 00:09:16.879 Well, this is where you often see a gap between 187 00:09:17.159 --> 00:09:20.159 the paper simulations you might see in a presentation and 188 00:09:20.200 --> 00:09:22.480 the daily reality of operations. 189 00:09:21.960 --> 00:09:23.480 The marketing versus the physics. 190 00:09:23.720 --> 00:09:27.320 Kind of Yeah, theoretical maximum read rates like up to 191 00:09:27.519 --> 00:09:31.039 one thousand or sixteen hundred tags per second sound amazing, 192 00:09:31.679 --> 00:09:34.720 But in a dense environment like a palette loaded with 193 00:09:34.840 --> 00:09:38.440 taged items close together, yeah, the physical effects, especially that 194 00:09:38.519 --> 00:09:42.279 destructive interference from all the tags re radiating signals back, 195 00:09:42.600 --> 00:09:45.120 can drastically reduce the actual performance. 196 00:09:45.360 --> 00:09:46.480 So what are the real numbers? 197 00:09:46.519 --> 00:09:50.159 Like measurements mentioned in the sources show maybe one hundred 198 00:09:50.159 --> 00:09:53.120 and sixty tags reliably read per second in the US 199 00:09:53.159 --> 00:09:56.320 on a uniform palette, but potentially only eighty or ninety 200 00:09:56.360 --> 00:10:00.000 in Europe under their regulations significantly lower than the theory 201 00:10:00.120 --> 00:10:00.840 radical highs. 202 00:10:01.120 --> 00:10:04.840 So managing those signals avoiding collisions is critical. Absolutely. 203 00:10:05.200 --> 00:10:08.679 Robust collision management algorithms weighs for the reader to sort 204 00:10:08.720 --> 00:10:11.600 out replies from hundreds of tags talking at once are 205 00:10:11.600 --> 00:10:14.159 incredibly important and complex engineering challenges. 206 00:10:14.240 --> 00:10:18.120 Okay, So let's say the reader manages the power, handles 207 00:10:18.120 --> 00:10:21.240 the messy environment, sorts out the collisions. How does the 208 00:10:21.279 --> 00:10:24.080 actual data getting coded and sent? How does zeros and 209 00:10:24.120 --> 00:10:25.519 ones become radio signals? 210 00:10:25.639 --> 00:10:28.039 Right? There are two main steps here, bitcoding first, then 211 00:10:28.159 --> 00:10:31.679 carrier modulation bit coating. That's how you represent the binary 212 00:10:31.799 --> 00:10:35.559 zs and ones as electrical voltage levels or transitions before 213 00:10:35.600 --> 00:10:38.840 they go onto the radio wave. There are different schemes 214 00:10:38.879 --> 00:10:42.519 like RZI and RZ Manchester Miller coding. 215 00:10:42.840 --> 00:10:45.000 Why so many do they have different advantages? 216 00:10:45.080 --> 00:10:48.159 They do. The choice affects things like the data rate 217 00:10:48.440 --> 00:10:52.159 energy efficiency, which is super critical for passive tags, how 218 00:10:52.240 --> 00:10:55.200 well it handles tag movement, and how easy it is 219 00:10:55.240 --> 00:10:57.840 for the tag to decode the signal and stay synchronized. 220 00:10:58.120 --> 00:11:01.200 Synchronized yeah, keeping the tags in internal clock aligned with 221 00:11:01.240 --> 00:11:05.120 the incoming data stream. Some codes like Manchester or Miller 222 00:11:05.200 --> 00:11:08.879 have guaranteed transitions like a built in heartbeat, which really 223 00:11:08.919 --> 00:11:11.679 helps the tags stay locked on even with low power 224 00:11:11.720 --> 00:11:15.240 and noise. NRD, which can have long strings of the 225 00:11:15.279 --> 00:11:18.000 same bit level, can be harder for a simple tag 226 00:11:18.080 --> 00:11:18.919 receiver to handle. 227 00:11:19.039 --> 00:11:20.480 Okay, so that's the bit pattern. 228 00:11:20.799 --> 00:11:24.120 Then what then comes carrier modulation. This is where you 229 00:11:24.159 --> 00:11:27.240 take that coded baseband signal, the pattern of zeros and ones, 230 00:11:27.399 --> 00:11:30.200 and use it to modify the main radio frequency carrier 231 00:11:30.240 --> 00:11:31.600 wave that the reader is sending. 232 00:11:31.399 --> 00:11:34.519 Out, like making the flashlight beam blank exactly like that. 233 00:11:34.879 --> 00:11:38.559 You can change its amplitude brightness that's amplitude shift keying 234 00:11:38.799 --> 00:11:43.320 or ASK. You can change its frequency color pitch, frequency 235 00:11:43.320 --> 00:11:47.360 shift keying FSK, or change its phase a more subtle 236 00:11:47.399 --> 00:11:49.720 timing shift phase shift keying. 237 00:11:49.519 --> 00:11:51.799 PSK and SK is common. 238 00:11:51.840 --> 00:11:54.399 Very common, especially for the reader to tag link. You 239 00:11:54.440 --> 00:11:57.200 can have ASK one hundred percent where the carrier is 240 00:11:57.279 --> 00:11:59.559 turned completely off for a zero and on for a 241 00:11:59.600 --> 00:12:03.279 one for example. This sends maximum power during the on bits, 242 00:12:03.279 --> 00:12:05.200 which is great for powering passive tags. 243 00:12:05.240 --> 00:12:06.240 Makes sense, Or. 244 00:12:06.200 --> 00:12:08.759 You might use ASK x percent where the amplitude is 245 00:12:08.799 --> 00:12:11.759 only reduced by say thirty percent or fifty percent. This 246 00:12:11.799 --> 00:12:14.039 can allow for faster data rates, but gives the tag 247 00:12:14.120 --> 00:12:17.240 less energy to work with tradeoffs again, always trade offs. 248 00:12:17.320 --> 00:12:20.320 Now, you mentioned something earlier about spread spectrum and a 249 00:12:20.360 --> 00:12:23.320 connection to classic movies. This sounds interesting. 250 00:12:23.840 --> 00:12:26.879 Yes, it's a fantastic story. We're talking about spread spectrum 251 00:12:26.919 --> 00:12:31.399 techniques and specifically frequency hopping, which has this incredible, almost 252 00:12:31.480 --> 00:12:32.720 unbelievable origin. 253 00:12:33.000 --> 00:12:34.200 Okay, it was. 254 00:12:34.159 --> 00:12:37.159 Invented and patented during World War Two, not by a 255 00:12:37.240 --> 00:12:40.559 radio engineer, but by the famous Hollywood actress Hetty Lamar 256 00:12:40.720 --> 00:12:44.440 behavior Lamar the very same along with an avant garde 257 00:12:44.480 --> 00:12:46.519 composer named George Antheel. 258 00:12:47.159 --> 00:12:48.879 No way, what were they trying to do? 259 00:12:49.279 --> 00:12:54.159 Their goal was to create unjammable guidance systems for Allied torpedoes. 260 00:12:54.679 --> 00:12:57.519 They came up with the idea of rapidly changing or 261 00:12:57.559 --> 00:13:00.919 hopping the radio frequency of the guidance signal in a 262 00:13:00.960 --> 00:13:02.720 pseudo random pattern, so the. 263 00:13:02.799 --> 00:13:05.240 Enemy couldn't lock onto one frequency to jams. 264 00:13:05.000 --> 00:13:07.799 Exactly and the sequence for the frequency hops. It was 265 00:13:07.840 --> 00:13:11.720 based on player piano role technology. Anthel was experimenting with 266 00:13:11.759 --> 00:13:13.840 synchronizing multiple player pianos. 267 00:13:14.039 --> 00:13:17.879 That's wild and actress in the composer inventing secure military 268 00:13:17.879 --> 00:13:18.799 communication tech. 269 00:13:19.120 --> 00:13:22.360 It's an amazing piece of innovation history. The Navy classified 270 00:13:22.399 --> 00:13:25.600 the patent and variations were used secretly for decades. 271 00:13:25.720 --> 00:13:29.879 Wow. So how does frequency hopping spread spectrum FHSS work 272 00:13:29.919 --> 00:13:31.000 in RFID today? 273 00:13:31.519 --> 00:13:34.639 Well, the principle is similar. The reader rapidly hops between 274 00:13:34.679 --> 00:13:38.000 different frequencies within its allocated band. This makes it inherently 275 00:13:38.039 --> 00:13:41.159 resistant to interference on any single channel, and it allows 276 00:13:41.240 --> 00:13:44.480 multiple readers to operate in the same area without interfering 277 00:13:44.480 --> 00:13:46.879 with each other quite as much. And related to this, 278 00:13:47.080 --> 00:13:50.240 in Europe, there's often a requirement called Listen before Talk 279 00:13:51.000 --> 00:13:54.519 or LBT, part of the ETSI standard. That's it, before 280 00:13:54.519 --> 00:13:57.600 a reader transmits on a particular frequency channel, it has 281 00:13:57.639 --> 00:13:59.799 to quickly listen to see if someone else is already 282 00:13:59.840 --> 00:14:03.200 using it. If it's clear, it transmits. If not, it 283 00:14:03.279 --> 00:14:07.039 waits or hops to another channel. It helps manage spectrum sharing. 284 00:14:07.559 --> 00:14:10.399 It's probabilistic, though, and it limits how long a reader 285 00:14:10.440 --> 00:14:13.399 can stay on one channel, often just a few seconds max. 286 00:14:13.480 --> 00:14:17.080 Okay, And is FAHSS the only way to spread the spectrum. 287 00:14:17.159 --> 00:14:20.879 No, there's another main technique called direct sequence spread spectrum 288 00:14:21.080 --> 00:14:25.559 or DSSs. Instead of hopping frequencies, DSS takes the data 289 00:14:25.600 --> 00:14:28.720 signal and multiplies it by a much faster pseudorandom code 290 00:14:28.720 --> 00:14:30.519 called it chip sequence multiplies it. 291 00:14:30.679 --> 00:14:30.960 Yeah. 292 00:14:31.039 --> 00:14:34.240 Mathematically, the result is that the signal's energy gets spread 293 00:14:34.240 --> 00:14:37.080 out over a much wider frequency band, making it look 294 00:14:37.159 --> 00:14:40.919 more like low level noise. This also provides resistance to 295 00:14:40.960 --> 00:14:43.000 interference and allows multiple users. 296 00:14:43.159 --> 00:14:46.000 So FAHSS jumps around, DSS smears it out. 297 00:14:46.039 --> 00:14:46.799 That's a good way to put it. 298 00:14:46.919 --> 00:14:50.039 Yeah, okay. So bringing this all together, the bitcoding, the 299 00:14:50.080 --> 00:14:54.159 modulation like ASK, the spread spectrum techniques. What does this 300 00:14:54.279 --> 00:14:58.240 mean for how well a system actually performs, especially if 301 00:14:58.279 --> 00:14:59.120 things are noisy? 302 00:14:59.279 --> 00:15:01.879 It means the system designer has a lot of choices 303 00:15:01.919 --> 00:15:06.000 and each has consequences. For example, using ASK one hundred 304 00:15:06.000 --> 00:15:10.000 percent modulation, turning the carrier fully on and off delivers 305 00:15:10.080 --> 00:15:14.360 more energy per bit to the tag that helps maximize range. 306 00:15:14.440 --> 00:15:18.000 For passive tags, good for range, but maybe not the fastest. 307 00:15:18.559 --> 00:15:22.759 Then using bitcoding like Manchester or Miller with those built 308 00:15:22.759 --> 00:15:25.840 in transitions gives the tag much better noise immunity and 309 00:15:25.879 --> 00:15:29.000 makes it easier to stay synchronized compared to something like ENERZ. 310 00:15:29.279 --> 00:15:31.039 More reliable decoding right. 311 00:15:31.080 --> 00:15:34.320 And then adding frequency hopping on top provides that robustness 312 00:15:34.320 --> 00:15:37.120 against interference and allows more systems to operate in nearby. 313 00:15:37.399 --> 00:15:41.480 So you're constantly balancing range, speed, power consumption for the tag, 314 00:15:41.559 --> 00:15:43.440 and resilience to noise and interference. 315 00:15:43.480 --> 00:15:46.279 It's a complex equation to solve for each application. Now, 316 00:15:46.320 --> 00:15:48.559 for all these different systems from different companies to actually 317 00:15:48.600 --> 00:15:52.159 work together, there must be rules standards, right. You cant 318 00:15:52.200 --> 00:15:54.080 just have everyone making up their own radio signals. 319 00:15:54.120 --> 00:15:57.360 Absolutely not, It would be chaos. The International Organization for 320 00:15:57.399 --> 00:16:01.919 Standardization ISO is crucial here. They develop standards like ISO 321 00:16:02.039 --> 00:16:04.960 eighteen thousand DASHER for UHF. 322 00:16:04.519 --> 00:16:06.519 Air Interface air interface meaning. 323 00:16:06.559 --> 00:16:08.799 Meaning the exact rules for how the reader and tag 324 00:16:08.840 --> 00:16:14.159 communicate over the air, frequencies, modulations, coding commands, anti collision protocols, 325 00:16:14.279 --> 00:16:18.360 everything needed for interoperability. There's also ISO eighteen thousand and 326 00:16:18.399 --> 00:16:21.279 four for the two point four to five Getta Hurtz band. 327 00:16:21.440 --> 00:16:23.840 Okay, and I've heard of EPC. Is that related? 328 00:16:24.080 --> 00:16:27.679 Yes? EPC stands for Electronic Product Code. Think of it 329 00:16:27.720 --> 00:16:30.399 as a system for assigning a unique serial number to 330 00:16:30.480 --> 00:16:32.240 every single physical object. 331 00:16:32.360 --> 00:16:34.559 Different from barcode, then very different. 332 00:16:34.399 --> 00:16:37.679 A barcode usually identifies the type of product, like this 333 00:16:37.799 --> 00:16:40.919 is a can of soup. An EPC tag identifies this 334 00:16:41.039 --> 00:16:45.559 specific can of soup distinct from every other can. EPC Global, 335 00:16:45.679 --> 00:16:48.600 now part of GS one, develops standards built on top 336 00:16:48.639 --> 00:16:51.720 of the ISO AIR interface, like the widely used EPC 337 00:16:51.840 --> 00:16:54.799 Class one Generation two or C one G two standard 338 00:16:54.840 --> 00:16:57.879 for UHF. It defines the data structure. 339 00:16:57.440 --> 00:17:00.639 For that unique ID, giving every item a day digital. 340 00:17:00.360 --> 00:17:03.279 Identity, precisely huge for supply chains. 341 00:17:03.360 --> 00:17:06.039 But going back to those performance numbers, you mentioned the 342 00:17:06.039 --> 00:17:09.880 difference between pay per specs and reality, especially with reading 343 00:17:09.960 --> 00:17:10.599 whole palets. 344 00:17:10.759 --> 00:17:13.240 Right, the standards define how things should work, but the 345 00:17:13.240 --> 00:17:16.079 physics in dense environments like a palette full of taged 346 00:17:16.079 --> 00:17:19.480 items still applies. Those claims of reading thousands of tags 347 00:17:19.559 --> 00:17:22.880 might come from simulations, but real world tests often show 348 00:17:22.960 --> 00:17:25.480 much lower numbers for getting a one hundred percent read 349 00:17:25.519 --> 00:17:27.039 rate on a full pallet, like those one. 350 00:17:27.039 --> 00:17:28.839 Hundred and sixty versus eighty ninety numbers you. 351 00:17:28.839 --> 00:17:33.119 Mentioned exactly that destructive interference from closely packed tags reflecting 352 00:17:33.160 --> 00:17:36.440 signals back is a real killer for performance in those scenarios. 353 00:17:36.480 --> 00:17:39.559 It's a persistent challenge and regulations. 354 00:17:38.960 --> 00:17:41.559 Must play a big role too, especially across different. 355 00:17:41.319 --> 00:17:45.440 Countries, oh massively. The regulatory landscape is very different globally. 356 00:17:45.720 --> 00:17:49.440 In the US, the FCC sets the rules. They might allow, say, 357 00:17:49.680 --> 00:17:52.799 up to four watts of power eerp in the main 358 00:17:52.920 --> 00:17:57.640 UHF RFID band and let systems transmit continuously one hundred 359 00:17:57.640 --> 00:18:00.920 percent duty cycle okay. But in Europe the TSI standards 360 00:18:00.960 --> 00:18:04.000 are generally stricter. Power limits are often lower, maybe two 361 00:18:04.079 --> 00:18:08.000 watts ERP, which is measured differently related to a dipole antenna, 362 00:18:08.440 --> 00:18:10.960 and they often mandate things like listen before talk and 363 00:18:11.079 --> 00:18:14.200 have much lower duty cycle limits, maybe only allowing transmission 364 00:18:14.279 --> 00:18:16.079 ten percent of the time in certain bands. 365 00:18:16.200 --> 00:18:18.359 So a system design for the US might not even 366 00:18:18.359 --> 00:18:20.759 be legal or perform nearly as well in Europe. 367 00:18:20.799 --> 00:18:25.279 Correct. These regulatory differences fundamentally impact system design, achievable range 368 00:18:25.319 --> 00:18:28.039 and red speeds. And we also have to consider human 369 00:18:28.119 --> 00:18:32.039 safety regulations right exposure to radio waves. Yeah, bodies like 370 00:18:32.240 --> 00:18:36.119 IC and IRP set guidelines for limiting exposure to electromagnetic fields, 371 00:18:36.359 --> 00:18:39.559 things like SAR limits specific absorption rate measure energy absorbed 372 00:18:39.559 --> 00:18:42.839 by tissue. System designers absolutely have to ensure their devices 373 00:18:42.880 --> 00:18:44.559 operate well within these safety limits. 374 00:18:44.640 --> 00:18:47.799 So given all this, the physics, the environment, the standards, 375 00:18:47.839 --> 00:18:50.880 the rules, right, what actually goes into making a tag 376 00:18:51.039 --> 00:18:53.240 or a reader and how do engineers make sure they 377 00:18:53.279 --> 00:18:54.440 work reliably well? 378 00:18:54.519 --> 00:18:58.720 Tag design is surprisingly intricate. You have the silicon ship itself, 379 00:18:58.920 --> 00:19:01.720 the antenna which to be carefully designed for the target 380 00:19:01.799 --> 00:19:04.519 frequency and the object it's going on, and the substrate 381 00:19:04.559 --> 00:19:07.359 holding them together. Even just how the chip is physically 382 00:19:07.400 --> 00:19:10.599 attached to the antenna substrate is critical. A bad connection 383 00:19:10.799 --> 00:19:13.160 can kill performance or detune the tag. 384 00:19:13.000 --> 00:19:15.599 Completely, and tiny details matter immensely. 385 00:19:15.759 --> 00:19:18.680 For the base stations the readers, they have sophisticated RF 386 00:19:18.720 --> 00:19:22.720 components inside things like circulators, which act like traffic cops 387 00:19:22.720 --> 00:19:26.039 for the radio signals, directing the strong outgoing signal to 388 00:19:26.079 --> 00:19:29.880 the antenna and routing the incredibly weak incoming backscatter signal 389 00:19:29.880 --> 00:19:32.640 from the tag to the sensitive receiver without them interfering. 390 00:19:32.960 --> 00:19:34.920 Directional couplers are used to sample. 391 00:19:34.640 --> 00:19:37.519 Signals too complex electronics definitely. 392 00:19:37.480 --> 00:19:40.240 And to ensure everything works us expected and plays nicely 393 00:19:40.279 --> 00:19:44.079 with other equipment. There are standardized tests. Conformance tests like 394 00:19:44.160 --> 00:19:47.079 isowainhead of zero four seven check if a device follows 395 00:19:47.079 --> 00:19:50.319 the air interface rules. Performance tests like iso winn has 396 00:19:50.400 --> 00:19:53.079 zero four to six measure how well it actually works, 397 00:19:53.200 --> 00:19:54.640 range sensitivity, etc. 398 00:19:55.160 --> 00:19:56.359 And where do they do these tests? 399 00:19:56.720 --> 00:20:01.079 Ideally in specialized anacoic chambers. These are rooms ligned with 400 00:20:01.160 --> 00:20:05.279 material that absorbs radio ways, preventing reflections. This creates a 401 00:20:05.279 --> 00:20:10.640