Very. Good well it's a pleasure to be here good morning or, good afternoon everybody. I'm Ted Rappaport. We're. Messing with the Skype right here to make sure that it works and the audio is good. Are you able to see me and see, the slides. Was. That a yes okay. Great. And I want to give a big thanks to Adam, Holston RIT, and, web. Director, who's kind of put, this together and, has a pretty. High-tech system. Here, to. Project it and to project the slides so thank you Adam and. Thanks. To, a few of you in the room here I'd. Like to present some. Material. And express, some of the work we're doing at NYU Wireless. Thanks. To the many industrial, affiliate sponsors, who, support our, lab and have kind of followed. Us in the millimeter wave world, and who are also. Encouraging. Us to look at this great future. To terahertz, and beyond if you're. Aware, we have a great seminar series I'll, talk about that in a minute I'm, unable to there. We go yeah I'll talk about that now. We. Have a seminar, series we've put together we just had the second seminar. Speaker, just moments, ago finish. Joseph, Jornet. From. University, of Buffalo. It. Was supposed to be live but, apparently, they could not broadcast, it live today, at 11:00 it will be posted on the web shortly, they. Had some technical difficulty. I apologize. To some of you who were trying. To tune in for that. Hopefully. That won't happen again but that'll. Be up on our website shortly, but we've launched. This terahertz. And Beyond, series. Focusing. On circuits, to. Bring some of the world leaders to campus and we've, had a couple of great days already. With iodine Baba Connie, CLA and Joseph. Jornet who's here today. Really. Exciting. But, this gives you an idea of some of the topics and some of the fundamental. Breakthroughs, that are happening that will make terahertz, a reality, and I, think earlier. Than than many of us would have thought a few years ago. So. When we talk about millimeter. Waves and terahertz. It's. Really good to put everything in perspective on, where they are in the. Frequency. Range the physical, size and how. They how they look you, think. About the microwave, region where we're talking, about. Hundreds. Of megahertz, to. A few, gigahertz the. Wavelengths. Are on the order of. Tens. Of, millimeters. Centimeters, and. The. Frequency, range is like ten to the ninth. Jitter, Hertz. Really. The size, of butterflies, if you think of a physical size but, as we move up to the millimeter. Wave and, terahertz, regime, we. Start to look at frequencies. That are now on the order of hundreds. Of Jigga Hertz, ten. To the twelfth. In. Frequency, a thousand. Giga Hertz where, we're talking about the, size of the tip of a needle. So, these, wavelengths. Are very very small and what's. Interesting is the temperature, at, which an object radiates. The most. Intense, for. The particular, frequency, you look at the, human body at.
Room Temperature, at a hundred to 200 to 300, Kelvin say. 300, Kelvin it's, a smack dab in the terahertz. And even, just above the terahertz, region, moving. Into the. Infrared. So. We're. Really, terahertz. Is kind of this magical, frequency, we don't know a lot about it, but it's, when we start to get so small in size. Where. It. Temperature. Of objects. Where. This radiation is most intense the human body emits, terahertz. Kind of gnat really and and. It's it's so small and, that, has advantages. And disadvantages. Which I'll talk about in this talk, now. Those, of you on the call are very, familiar, with this, the. Spectrum horizons. Initiative. By the FCC, earlier this year is. Looking, to make spectrum. Available and to, help provide. Huge. Amounts, of spectrum, more than we've ever seen, before, in, the. History of the world over a hundred gigahertz of spectrum. For licensed point-to-point and looking. At 15 genders of unlicensed, devices, the. FCC, has proposed, this experimental. Licenses. Are now going to be there. Seeking comments, and looking to make, spur mental licenses, available from 95, gigahertz to, 3 terahertz. That, is HUGE because the FCC really. Has never looked before ninety. Five gigahertz, before. It's really kind, of been and off limits, NASA, has, a lot of the deep-space probes, a lot of probes up above the ionosphere, in space, a lot, of sensing, so, opening. Up the spectrum ninety-five, jitter Hertz the three terahertz, at least having that discussion is a very promising, sign. At the FCC by. The way on the experimental, license, system, FCC. Sought our advice in NYU was really a leader in creating. That fast, licensing. System, if. You look at the spectrum band from 95, to 300. Gigahertz and I. Have a chart here you can see that. Most of it is, occupied. By radio. Navigation and. By. Radio. Measurement. Satellites. Radio. Location, are. Really a lot of astronomy, a lot of radio astronomy, but. There's a tremendous, amount of spectrum here available. Now. The millimeter-wave coalition. Of course that's who we are on this phone call. We're. Looking at trying, to bring some momentum to, open up large, contiguous, blocks, of spectrum, and that is key large. Continuous, blocks. If you go back to this, chart. You can see that there aren't large. Continuous. Blocks. We're. Looking to open up some of these and be able to find some usage and make. Some. Spectrum. Available for, unlicensed, and licensed. Of. Course everyone on this call knows this but this is a slide, I'm starting, to present publicly, about.
Our Coalition. Now. We're not the only game in town this, millimeter-wave. Coalition, a lot. Of other countries are out in front of this millimeter, wave and terahertz. Regime. In fact in, Europe there, really been much more out in front and are much more open about the dialogue of having frequency. Bands available from, 50 to, 300 gigahertz, and they've. Already earmarked, a number, of frequency, bands that they're looking at, the. Traditional, bands which are in the millimeter wave range, as long as a band, V, band and this. Is on the agenda at ITU. Are for WRC. 19, in. 2019. Agenda, item 115, is going to be considering, land. Mobile and fixed services, in the, 275. To. 450. Jitter Hertz range and. Both. Asia and, Europe have. Really been kind of unified, in this so, it's appropriate that the millimeter, wave coalition. Get, active and try to give some guidance to the FCC, to see if we can get some global, harmonization. One. Of the first millimeter, wave 2 terahertz, band transitions. That will likely occur, is moving. Up to D band while. Demand, is not a terror Hertz frequency. Range technically. Terahertz, is 30, to 300 gigahertz. Terahertz. Is being. Used commonly. For things above 100. Gigahertz so, if. You look at the. D band 110. To 170. Gigahertz this is really you. Can think of it as the first terahertz. Band even though terahertz, technically, is only. 300. Gender. Hurts two to, three terahertz. But. There's a lot of spectrum, in Deeb and you can see that D band is used by radio astronomy. Radio. Navigation, satellite. But. There is spectrum, here for, mobile indie. Band now one of the nice things about D band and the. Frequencies, at, 140. Gigahertz and up. To 240, gigahertz. Is. That, the relative, attenuation. The attenuation, in air. Due. To the. Molecules. The water molecules, air vapor there's. Very little attenuation beyond. The, normal freeze of free space path loss so. There are regions. In the sub. Terahertz, region. Where we can do really. Mobile communications. And as I will show shortly, we. Can do mobile communications. Even better than, we do today, and. I'll get into that. Now. There are a couple of really good papers, the overview papers, on terahertz, some. Of the experiments, that have been done over the last decade, and on, the left you can see the water vapor chart, the, attenuation, chart, due to oxygen in air, different. You. Can see at. Different levels, of humidity and you can also see for, different distances. And there. Have been both indoor, and outdoor measurements. Made. Experimentally. All, the way up to, 700. 800. Gigahertz you. Can see that we're talking about distances, that are either sub-meter. Or. Distances. Up to several. Kilometers. In. Outdoor environments, up--it, carrier frequencies, of, 100-200. Even, up to 400, jinder hertz outdoors. And of. Course as I'll show in a minute when you go up to these higher frequencies, these. Kind of distances, should not be surprising, even. With the difficulty, of making power sources, because. There's huge advantage, to high gain antennas, and we'll, see that how, those high gain antennas, can, actually, overcome, the path loss. And if. You notice on the bottom left curve the. Attenuation. The. Gaseous, attenuation. In the atmosphere, indeed. Has many peaks and spikes but, there are some, good zones in the 200, to 300. Gigahertz. Region, where. You'll only lose 10 DB more per kilometer, due. To free. Space. Due, to the oxygen. Meaning. That over 100. 200, meters, you're only going to lose a couple of DB extra, over. Normal, propagation. So, there's. Some amazing, places that, we can operate up to several hundred gigahertz. For. Cellular. Mobile. Point-to-point. Wireless. Replacement. For backhaul, very, encouraging, and the, data rates you get when you go up to these higher frequencies naturally.
Increase. Because. Of the greater bandwidth, savais labelled we, go up to terahertz, and. These. High hundreds. Of Jigga Hertz frequencies, because there's more bandwidth, and you can, see that the achievable, data rates first, carrier, frequency, on the right are, enormous. We're talking, hundreds. Of gigabits, per second, and you. Think about a fiber-optic, cable we're starting to talk about true fiber cable. Replacement, over-the-air. What's. Interesting is to implement these high, data rates we can have an opto electronic approach. Using, optics or we. Can have an all electronic, approach. Some. Of the optical, approaches, involve. Using, lasers. Or optical. Signals, and beating. Them mixing, them with another laser signal, very. Close in frequency, at the optical layer which. Is very, wide apart, in frequency, down. At conventional. And this, is how we can make several, hundred jitter Hertz sources. There's. A lot of power inefficiency, in going optical, to electronic, but, we could actually build. Oscillators. And modulators, using, optical, electronic. Components, and by, mixing two optical frequencies. That are 50 or 100 or 200. Apart we, can create very stable very. Powerful, modulators. And mixers. And. Signals. To, enable these hundreds, of gigabit, per second. Transmissions. Now. At NYU Wireless, we've built a broadband, channel sounder you, can see it on the left thanks. To our industrial affiliates. Who have supported this work, namely. Keysight, and National, Instruments. They've. Both given, us a tremendous amount of equipment have, supported, the development of this gear you, have Virginia, diodes is on this call as one of our vendors you, can see we have Virginia diode, Radiohead's. On. The, left and right and. These. Are, basically. Front, ends and we mix down to about 25. Gender, Hertz local oscillator, this. Slide keeps wanting to go forward, for some reason there's some timing set in and I guess. Adam. I don't, know why it wants to jump ahead and. Three. Of my graduate, students are on the right you. And Xiao Xu hanwen ojas who are working on various, applications. Of this. 140 gigahertz band that I'll present, in a little bit we've just published the paper, it appears, publicly, in the global. Communications, conference, and all of our industrial affiliates. Of course see this work well before it's ever published. But. Here's some details about the. Channel. Sounder that we have we basically have a wideband transmitter. And receiver, we're. Using. Basically. A simple. Dual conversion, i f, 2i. F system, with. 4 gigahertz RF, pass band and we. Have really a commercial, grade 4, gigahertz wide RF. Transmitter, RF, receiver with rotatable, horn, antennas, and we're, using this now to conduct, channel measurements, indoor and outdoor we. Could also modulate, it so we basically have a true transceiver. And. This is a very powerful capability. Because, we're able to take it into the streets of Brooklyn the, streets of New York City deploy. Towers, and measure. Measure. Transmitter. Transmit. Power, path loss, multipath. Angle. Of arrival and build, the channel models that will be needed for terahertz. Wireless. Communications. And here's, some measurements, we made and by, the way we are behind, on schedule on the measurements, we have a lot of measurements to do and the students have started doing it every weekend we're, hoping to get outside before. It snows we're doing indoor now but we've really got to get outdoors this, fall there's, a tons of data we're starting to collect. First. We're doing indoor because as you'll see there's some very exciting applications. Indoor. With position, location and we're, also trying to understand, how, the environment would, work up, at 140, degrees how different, is it is how, different is it from 28, and 73, which will be commercialized. Soon. Well. First here's a look at the free, space. Measurement. In a closed area within a few meters and, I, want to point something out that's very important, to realize in, the, industry, in academia, people are, always saying the, path loss is worse as you go higher in frequency it's.
What You always hear. Things, are lossy, and then you see a curve like this, on the right which, shows that over, short distances, indeed. You, have more path loss at. 140. Jigga Hertz than, you do at 28, Jigga Hertz right everyone. Thinks that and that's, because I'll show the math in a minute because there's this free space loss component. In freeze free space equation. That. When you remove the antenna gains, that. Is when you assume the antenna gains r1. It's. True. Things. Are more lossy, but. In all communication. Systems, we use antenna, gains. And. Here's the math that. Shows in fact the opposite, is true when you use the antenna gains when. You use the area of the antenna and keep. The area the same and go up in frequency you, actually, get gain as you, go higher in frequencies, if, you look at the top here's, the freeze, free space free, space equation and, what, people always do is they peel off the lambda, weird over. 4 pi d squared, term on the top they. Peel that off and, that's. Why they say things are lossy, as you, go higher in frequency but. If you look at a physical, antenna with an aperture. ACB. And you. Replace, the antenna gains at both the transmitter, and receiver with. The aperture, what. You see is the, lambda squared in the. Free space loss, cancels. Out with a lambda by 4 in, the, antenna gains and at. The bottom equation you. See that it's not lambda. Squared in the numerator, but. Rather it's, lambda, squared in the, denominator you. See. This I don't, know if you're catching this atom but it's, the lambda, squared, in, the denominator here. It was lambda, squared in the numerator and, as, lambda gets smaller, what, this means is for the same physical. Area. Your. Receive power goes up basically. Your, gain is increasing, for, a constant, physical area, that's. Very powerful and, the. World really doesn't, quite. Understand, this yet you. See what that basically means, is you, can operate at higher bandwidth. Scaled. By, the area, of the transmitter, and receiver antenna. Higher. Bandwidth, for, a given SNR, as you, go higher in frequency so, you're basically getting more, bandwidth, for free, as we. Go up and keep the physical area the same over frequency, in fact. To show that here on the Left what, you can see is if. You. Assume on, these bottom, three, curves. That. You have omnidirectional, antennas. At the transmitter and. The receiver which. A lot of people assume. Well. Yes indeed. Your received power is going to be stronger, at. 28. In Hertz than it is in 140. Gigahertz. You. Can see there's about 14, DB difference from 28, to. 40, 140. Gigahertz, there's. About 14, DB difference here. However. If you assume one end of the link, has. An antenna, area, that stays constant with, frequency, and then. You just assume, the other end of the link has. A directional. Antenna in other words it gets smaller with. Frequency, then. It turns out frequency, is independent, you get the same received power in all three bands.
Because. You've kept one, of the physical areas. Steady. With frequency, and you've let the other antenna, be omnidirectional, but. If you assume both, antennas. Around the directional, that, is you consider the area, being constant, at both the transmitter and receiver you. Get 50/50, 14 DB stronger. Received, power at, 140. Jibra Hertz and then you do it 28 jigger Hertz this is very powerful, in other words you get 14, DB. 20. Times the bandwidth, for free, at the same operating. Signal-to-noise, ratio, so. You can basically operate, with 20 times, wider, spectrum, allocation. At no, cost, to the signal-to-noise ratio. When, you use directional, antennas. Now. You have to find the beam as the, area gets as the. Area stays the same and you go up in frequency it becomes. A much more focused beam, so you have to do beam finding, and be, able to lock those more, site antennas, together but. Once you've got the boresight aligned you, get this huge gain in the, link so going. To millimeter wave and terahertz doesn't. Degrade. Your, link it. Actually, improves, your link so, going. Up higher is better, it's. Better for. The future of wireless now, some things aren't better as you go higher in frequency and here we've started to look at different materials, we've, looked at glass we've looked at drywall, and what you can see is you go up higher in frequency. Common. Materials, like drywall. And, glass. They. Become more lossy, as we go higher in frequency, clear. Glass is 3 to 4 DB loss at 28 June Hertz it's, 7 DB at 73, June Hertz and it's 8 or 9 DB at 140. Giga Hertz so there's a couple DB, differ. We're looking at a lot of these materials, now we're, really doing lots of measurements and. Getting. More data for common materials, as we speak but. Understanding. This is important, so, that we can figure out range and coverage and, interference. In Wi-Fi. And other kind of indoor environments. And then we'll go outdoors and also, look at the channel in this way. So. It's going to be dependent on the material but. This is really great we can use antenna gain to overcome, the. Free. Space loss in omnidirectional. Systems. Now to. Make measurements to understand the channel it's important to have a good methodology, to make sure you're truly measuring, the, materials, under test and not, picking. Up multipath, or measuring. Other, items, so, we've developed. Some. Standardized, measurement, guidelines, that can be used by other researchers we've published them recently at. The I Triple E vehicular technology. Conference, in order, to do both cross polarization. Discrimination. Tests. To make sure that our antennas are working right and how to measure cross polarized. Discrimination. So that we can understand, what the channel does in, depolarizing. The transmitted. Signal and also, looking at how to do these. Penetration. Loss measurements, and it turns out there's some simple geometry, based, on being outside of the Fraunhofer, region so that we can assume a plane wave but, also being such that our transmitter, and receiver are not, too far away that, we're picking up multipath, from ceiling, or ground or walls and so. There are some very simple geometries. That if you follow you can be sure that you're, measuring, the true device under test or, measuring. The true, antenna, cross. Polarization, and. We. Verified these approaches, by, looking first at some cross polarization. Discrimination. With a number. Of horn antennas, this, could also be used for phased array patch, antennas, lens antennas, but. We looked at wide beam and narrow, beam 30. Degree 10 degree, different. Horn antennas, here's, at 73, jipner hertz we're also doing this at 142. Hertz. Where. You basically, take, measurements, of your antenna in kind. Of a open free space chamber, it could be a large room it could be outside and if. You make the measurements, over several, close distances. That are still in the Fraunhofer, region and you. Get the similar slope, that is you get a slope over distance that, behaves properly by, free space you, can be assured that you're measuring the antenna, and you're not picking up multipath, components, and here's an example at 73, jr. Hertz with three different antennas, where you can see we've measured pretty. Much a 28. 29, 30 DB, cross polarization. Discrimination. So, we know with good confidence, because we recreated, it over, several closing distances, that are in the far field that. We can be sure that we are measuring an accurate. Xpd, and not picking up ground, bounce or reflections, or artifacts.
In, The system. Similarly. For penetration, loss measurements, the, same thing you followed the geometry, and you, make several distance, measurements. Close in but still far enough in the far field and you, make sure that the slope is correct got, a little bit of a slight, a couple. D a couple, tenths of a DB at the, five meter. Distance. This. Might be due to some fernell, effects, on the object, that we were measuring but still they're very very small within a fraction, of a DB so this is a measurement. Approach that you can use and we, use it for both polarizations. And for. Cross polarizations. The student didn't show horizontal, to horizontal, but we're making both horizontal, to horizontal, vertical horizontal horizontal. To, vertical and vertical vertical, and you can see that we have penetration. Loss measurements, and you. Take the average of these and you can get a very good estimate a, measurement estimate, of what, a particular material will do. Now. What. Are some of the applications of, millimeter, wave and terahertz. Once. You get up to this terrorist, regime where the where. The wave. Lengths start to become on the order, of microns, you, know hundreds tens of microns there's. A lot of things we can do now we've known a lot about, femtosecond. Optoelectronics. Since, the 80s, there's terahertz, time-domain, spectroscopy. Right. Now there's commercial, systems. Being. Used to, test plywood. To. Test wood, material, to see how the blue has been healed there on factory conveyor belts there's certain, spectroscopy, signatures. That you can get from, particular. Materials, from. Being. Able to detect fissures, cracks. Efficacy. Of coatings. To make sure they give, a proper. Spectral. Return, when hitting it with a broadband, signal, there. Are also, components. That have. Certain. Resonances. And by, using terahertz, imaging you, could basically exploit. Excite, the resonances, see, the energy produced, back from, the resonances the vibrations, so, you have this amazing spatial. Resolution. Due. To the short wavelength, you're, able to use that information, along with other methods. Say, visual. Imaging, and you, could basically, use, if. You will frequency, fusion, use terahertz, spectroscopy, along, with other known, things visual, or microwave. Response electrical, response, there's, a way to do. Identification. Of materials this way this. Is used commonly to detect, explosives. To, see how materials. Being used by terahertz. Imaging now. There's some disadvantages, relatively. Long acquisition, time given the huge bandwidth, given, the number of antennas, often used to try to get a sufficient link, it's. Hard to find the focal plane when you have such, tight. Resolution, how do you aim the, focus, because we're talking very very tiny wavelengths, very huge bandwidth, and active. Is active. Components, or a challenge, they're expensive, they don't really exist today so, you're dealing with passive. Signaling. Pass. Over monitoring. And that's very very, low-power. It's very hard to get the signal levels needed to, be able to get enough response, so, getting enough power out to, be able to do signal processing, is a challenge, right now. Well. In. Medical, imaging there's some very exciting, applications. And already a couple of our speakers, in the NYU. Circuits. Terahertz, and Beyond seminar, series have given us some insights into some of the very exciting. Medical. Imaging, capabilities. When, you get below the millimeter, wave lengths you, start, to see that you're talking about a. Place. Where we can basically. Put. Energy in, and see. What comes out looking. At the cells as lenses, looking, at the cells as structures, that give us insight, into what's, inside the, cell size is less than the wavelength and the, terror signals, pass through the tissue and basically. Can. Create, this scattering. The. Cells, are impacted. By the terror, Hertz radiation, but, the atoms, aren't broken up but. They're kind of shaken around there's, actually. Movement. And this. Movement. Gives. Energy. And gives various, ways. That we can detect what's happening. You. Can do time-domain spectroscopy, with, coherent, detection, and. Then, you can measure the absorption, coefficients. And the, refractive, indices and, you can basically based. On the categorization, of the refractive, indices make, decisions, about what's happening, tumors. Will have a different. Response than healthy. Tissue, different. Parts of the body have a different. Refractive index, they will reflect. And pass energy, differently, in, fact, if you look at tissue characterization. In polar, lips lip, polar. Liquids. There's. High, absorption, when, the.
Atoms. Are arranged, in a polar fashion. And fatty. Tissue has fewer polar molecules. And. Therefore, there's, going to be less exact, less, absorption, there's. A lower refractive index. Kidney. And liver tissues, have a much higher. Absorption. Coefficient. Absorb. The energy much, more and, in. Humors eudeco. Absorption. Coefficient, of refractive, endure, index. Are higher in tumors. They're, more solid, like solid, organs like, kidney or liver, so, we can get a lot of information out from. Terahertz. Transmissions. In the, body without. Causing. The. Breakup. Of atoms in the. Valence shells, which lead to unstable, atoms, which, can lead to cancer we're. Not at the energy, level yet, where the energy. From, the radiation is sufficiently, large that it can break an electron. Out which, leads to instability. It shakes, them but, does not break them. Now. One of the exciting applications. Of in the wireless world when you go up to terahertz. When, you're up at these frequencies where. The wavelengths. Now sub, millimeter, wave is, there's. A huge amount of bandwidth and there's. Also the ability to make very very directional, antennas, so. We can start looking at and we need the directional, antennas to be able to get the link margin, we saw how by, getting the directional, antennas we'll get more and more bandwidth, for free without. Losing the signal-to-noise ratio, so, you, can look at many applications. Of indoor positioning. Smart. Factories IOT. Sea in the dark using imaging, to basically. Look. Through walls, get. Pictures, almost. Like an x-ray but without using the mechanism, of x-ray looking, at the response, back from radiated. Energy to. See what's, happening, in the channel. These. Example, these benefits, wide bandwidth, sparse. Channels, with narrow beams we, can estimate the angle of arrival in, a. Handheld. Device or, in a base station so. We could start using this indoor we, could start and this doesn't have to be indoor positioning I should have just said mobile.
Positioning Because this applies outdoor, to and I need my students to think broadly they. Put it into orbit it's outdoor, indoor it's wherever it is and that's why the outdoor measurements, are so important, and while we're doing it because these same ideas can work out or what, we're doing it indoor here for the next view before, we move out door but, the bottom line is we've these wideband unless you can find the angle of arrival and, angle, of departure very narrow beams. It's sparse, so, this could be a feature, is. This showing up on the slide this. Thing okay, so. So. You, basically can see a. You. Can really add, this, kind of positioning, and it, will come for free in future, wireless, systems. One, of the ways to do. Location. Is to, use what's called angle, of arrival based, positioning, there are a number of methods used, where. Basically you figure, out the angle between beacons. Of known a priori, location. You figure out the angle and you can imagine spinning, a rotating, electrical. Antenna, or having, a phased array. And. As, I'll show shortly, imagine. Hat being able to receive energy from all angles of direction simultaneously. You, could be looking for sources and figure, out where you are. The, other way to measure, that's been looked at and. There are many many ways and we've published a paper at globe comm in, fact you can see down here some of the citations. Of the globe campaign well, now. I don't know where the globe come there is there's. A paper we have coming out of globe comm where, the student looked at signal. Strength positioning. Based on a simple, distance. Based signal, level law the farther away you go the signal gets weaker this is a very crude model but it turns out you can find the maximum likelihood, at any particular distance, for what the signal should be and you. Fuse, that with. Path, loss and angle, of arrival and, it, turns out you can do remarkably, well here. In the NYU wireless, center. We have about a 30, by 40 meter. Floor. With, cubicles, and looking. At simple. Fusion, just the first iteration we have a lot more work to do on this but a very simple iteration, you. Can see that we're able to get within a couple of meters accuracy. Using, a number of base stations and with, measurements, we made at 28 and 73. Jitter Hertz this is very encouraging because, we're not even trying to get position, we're just using the, measurements, we've made and just seeing how a couple, of different algorithms might work so. There's a lot more work we can do but, this gives us a lot of confidence, that position, location will be able to come for free as we get up to wider band and these directional, antennas, and one of the nice tools that we're developing is, a ray tracer which. Will allow us to use a physical environment, outdoor indoor, put. It in and it will help reduce. The number of measurements we make but. We, do have the measurement equipment here in our industrial affiliates. At NYU Wireless are constantly. Talking to us we're helping them they're asking us to measure things if, your company's interested in joining NYU, wireless we would love to have you provide. Data. All of our access, to our students, all of our research is given to our affiliates, before. The world ever sees it and we, do a lot of work for our affiliates, and, you're basically stockholders. In our in, our center so, the ray tracer will become a tool that our affiliates, can use as well and this. Will help us figure out what's, happening, there's. Another issue in terahertz. And even, a millimeter wave today that. Was, never a problem in earlier generations of, cellular and that is this. Idea of spatial, consistency. You. See in channel, modeling, the. World of Wireless has always considered, simulators. To figure out their path loss and their coverage, and their capacity by, generating, a channel, state and. Taking. That channel state and assuming. You, know over a local area that phases, change uniformly. And they, really haven't had, the problem, of doing the kinds of things that 5g. Will require which is beamforming, beam. Search. Diversity. Base stations, you. Know when we go up to these multi. Gigabit per, second data rates we're. Going to be have very, short time intervals, in which we have overhead. To do the control, and beamforming and then, we want to carry as much payload, as possible, before. We. Recalibrate. The channel so, the question becomes what, happens is a mobile at high speed say it 100 km/h. Is, moving. With. Tens. Of gigabits, per second data coming to it well. You, can travel over several meters. While. While, getting, you. Know huge, amounts of data and we, want to be able to have a very accurate channel, model that can model the transients as you, move over a local area and we. Call this spatial, consistency. Channel, models today really don't have the, ability to implement, successive.
Sample, Functions, that very accurately, recreate, the. Impulse response and, NYU Wireless, we have a track we've, measured the channel, we have the data and we're, implementing, that so, that you can really. Understand, the channel both from an omnidirectional and, directional, sense and directionals, more important, as we, go to terahertz how does the channel change with, a directional antenna does, the directional, antenna move out of the if you will spatially, consistent. Region, of the channel, or not, so. We're investigating, that so. That we can implement in, simulation. These, tools these, channel, models that can accurately allow, the industry, to, build. Algorithms. And build, the fire mac layer. Techniques. By having an accurate channel, model so there, are spatially, correlated, large-scale, parameters, the, light spread channel fading and there are also small. Scale parameters, that are spatially. Correlated, to a local area and. 3gpp. Has decided. Arbitrarily. It should be 10 to 15 meters in urban. Microcell, that's the standard body for the 5g cellphone new radio that's being implemented, today. But. It really is going to depend on the correlation distance, of large-scale parameters, and you only know those if you measure them, fortunately. We've measured a lot of this we have a huge amounts of data and one, of my master's students, is going through that data and we. Have this NYU sent channel simulator, that's a drop based channel simulator. We've. Got over 75,000. Downloads we get dozens. Of downloads a day all the major companies, research, universities. Are using, NYU sim and they're. Basically. Using. This to study the channel and we. Want to implement, spatial. Consistency. And here's an example of some of the many measurements we have where. We've measured in downtown, Brooklyn, of, the receiver moving, around a corner we. Ramble that we capture the impulse response at all these locations the, angle of arrival time. Of arrival the entire channel, and in. 3d. And then, we're able to see what happens and you could see at, the receiver locations, as you go around the corner you, go from a line of sight to an on line of sight the, multipath, extends, greatly, here, you can see the time. Delay becomes, greater as we go around the corner and so. We're using this data to build a spatial, consistent, model. I only. Have a couple minutes so let me finish off real quickly with, some other ideas that, some, other colleagues in NYU wireless are working on you. Know as we go up to terahertz, we've, got as I, mentioned huge distances. Potentially. And we saw that what's been measured to date and this, is where we think the bands are of interest are going to be, one. Forty, two twenty, three forty six, eighty and what's. Fascinating is, once, you start to get up to these huge band widths as you basically, can do edge computing, and cognition, we're. Basically having enough data where you could send, the. Information needed, to do decisions. Brain, information. You, know huge huge, decision, making can be sent over the air we kind of remote decision-making. For, major. Functions. Some. Of the big power, challenges. Are of course the RF sources, for power efficiency. Gascan. Indium. Gallium arsenide. Will, be potential. Solution. And the ADC and baseband processing, where we have low resolution. Signal, processing is what we want because if else show in a minute you, have lower number of bits in your analog to digital converter. You get much more power efficiency. Now, we're, going to need a adaptive, beam forming directional, sensing and. The. Other thing is we're, going to need to do is have, combat. These end-to-end effects, of channel intermittency, if someone walks in front of the channel, fortunately, the standard, is such, that this, end-to-end, channel, intermittency, is not going to be a problem unless you, really, have a deep shadowing, event unless, something really dramatic happens. I don't think and then. Intermittency. Is going to be that much of a problem it's pretty easy to solve but you, know getting the power efficiency. And getting, the beamforming. To be able to work is really important, now, the Holy Grail the, optimal, situation is, an, all-digital fully. Digital, terahertz, transceiver, we've. Done a lot of work on hybrid beamforming and, hybrid beamforming is likely to be used for the next decade. Hybrid. Beamforming is much more power efficient, you don't lose much at all and spectral, efficiency compared. To full digital but, eventually maybe, a decade maybe 15 years from now fully, digital, will be great to have and our job at nyu wireless is to be looking at things 10 to 15 years ahead so.
We're. Looking at the idea about having a fully digital beamforming. Array. Where you get a multiple, channels, simultaneously. Operating on the same antenna with, individual, beams formed, very, specifically, for each of those channels now. If you could do that with low resolution ABCs, as I'll show in a minute you, really, can, start doing better than, today's hybrid, beamforming systems. It, becomes more power efficient, but. That's tricky you could only do that if you use low resolution, and this. Formula, P. Equals, CF, to be is well known to be the power, its, ratio of power consumed, by an analog, to digital converter. There's, also many other challenges. But. What's. Really interesting is if you look at the 3gpp. Standard. They, give you about 20 milliseconds. In order, to form the beams and right now in 5g. New radio I wish they wouldn't have done this but they did it you basically are gonna look at one sector, and then another frame you look at another sector one frame you look at another sector, you, know if we implemented, you could have even done to its hybrid beam for me you could have done it all in one second, you, can in one sector so you could have done 120 millisecond, block and looked, all the way but they don't have that in the 5g new radio so, you're gonna have a little bit more latency, to find the antenna when you go into a deep fade, fortunately. 320. Milliseconds. Is not going to be that bad but it is a, it bad, for driving speeds actually, there's a lot bad for driving speed for pedestrian, 5g new radio will, be okay someone blocks in the way we'll still find another path but, you'd really if you could have done that all digitally, and had. It done in one 5g. New radio burst. You, could have done it in 26 milliseconds, and when you get to 26, milliseconds, to finding and fixing on the new beam you, can really start recovering. From any kind of deep fade, here's. What's really fascinating if, you look at the power consumed. In milliwatts. For. Analog, and hybrid. Analog. Is just one channel so this is just one channel hybrid. Would, be for multiple channels, and. If you do fully digital, at 8 bits and 4 bits look at the total power consumption, you. Can start getting fully, digital, for 16 channels, 16, receiver channels, less. Power consumed, than one analog. Channel that. Would be amazing, that. Would be amazing to get 16 channels less, than one analog, Channel today well. To do that we're gonna have to be able to make things work at four bits or less. And you. Could do sigma-delta. ADC x'. And actually get, even more equivalent, bits there are ways to do this but, what's exciting is if you look at the transmitter, even. At the transmitter, you get huge power advantages.
If We could ever go fully digital, with relatively, low resolution so. We. Looked at both high and medium this is some deep regen and some of his team looking. At power levels, and what's. Fascinating is, an all. Digital. Implementation. Actually could. Be more, power efficient, than, hybrid, if. We, can make it work with low resolution, and that's the challenge that's a great research area it's a real digital signal processing research area now some of the problems, with. Low. Resolution is, you'll never be able to exploit the, high SNR, s you, know that hot great resolution, would give you but you don't really need super, high SNR, s as long as you're getting the bit, error rate that you want another. Problem with low resolution is, you have relatively, low, I loves it. When you do filtering. Because you don't have a lot of resolution, to get deep side lobes however, you've got directive, antennas, so the directive antennas, will help along, with the low, resolution to, push this out of band below. What, digital signal, processing, alone. Will do these, are exciting areas, that's been published last, year in Asilomar, but. It's an important area for our industrial, affiliates understanding. Blockage we've done a number of works in blockage. A number, of students number of faculty, so we're understanding, the dynamics. But. That and we've also made an NS 3 module that a lot of our industrial affiliates, are using where. We're able to implement, kind of an open source, end-to-end, model, for, a betted apt of antennas, channel, models you can put in why you seem into this and. We're. Working on building a terahertz module, by the way if you look at the lecture today at the, NYU, circuits. Symposium. On circuits. Terahertz and beyond you'll. See that Joseph Jornet has an MS 3 module, in the, end 4 terrors fascinating. So, the you, know universities, were supposed to keep things open sorts we, like doing that and once. You have these link. Layer and physical. Layer models, the, kinds we're developing, here at NYU Wireless, you can then begin looking at end end performance. So. Let. Me conclude by saying, this. Millimeter-wave. Coalition, is important, I'm hoping. This presentation, can generate more interest more, involvement, from our industrial. Affiliate. Sponsors, this. Is a wide open area the spectrum, is huge. There's, a lot of work happening as you'll see in the NYU Wireless seminar series here. At NYU Wireless we have a couple of very large projects. We. Have a, comm. Center where, NYU, is the system's, lead with, Sunday Brandon, as part. Of a 21 million dollar. SRC. Project, to look at terahertz. Systems.
Devices, And then. Professor, ranga and myself are, on a NIST contract. Looking. At public, safety, communications. And using. Drones. And some, of these high bandwidth. Video applications. For Public Safety, in, conclusion, I want to thank all of you for being on the call thank you for being in the room great. Thanks, to our industrial affiliates. And I just learned that AT&T, is rejoining. So we'll have the AT&T logo up, there shortly. And if your company is not a member we, hope to have you involved in our Center soon thank. You very much.
2018-10-02