Millimeter Wave (mmWave) Communication Part 1

good morning everybody so the my talk will be a little bit more technical sight keeping this sterilization activity little bit aside let us focus for some coming 45 minutes in their technical beauty of this upcoming technology the joy of doing research in this new area and this remarkable transformation that has happened in the history of communications from there so er age long narrowband communication research to a very wide band high frequency research and where we all the fundamental researchers are feeling that the communication Theory also needs to be revisited for a lot of the cases and so the information theory people and on one side the signal processing people on other side and also the device people on the third side along with the network people are struggling all together all over the world so I will try to show you all those challenges and the beauties and the new stuff that have come up for us if we this in their next it is up to us the way the researchers will utilize this and the way it will be really used for the benefit of the mankind for that next generation see we've had in the microwave over the long ears and this millimeter wave communication the range starts from these 3 gigahertz to 300 gigahertz literally speaking these days after hundred gigahertz we have started talking about the terahertz communications also because lot of the industrial products already in the prototype version came also in the 70-72 one gigas 60 gigahertz products are also little bit available for short range communications low range 28 30 gigahertz ford outdoor communications the renders are up to coming up with the stuffs so people the research has already started in the area of hundred conversion so before entering a little bit about the school of telecommunications National School of telecommunication is a Erika record initiative which happened two decades back to dedicate the research in the field of telecommunications only and today where we stand we will highlight so that they when we separated out from the communication department of communication the focus was only to do the research in the telecommunication domain only whereas the main department does research in lot of other areas also sorry so today's talk will be briefly on the understanding of the growth of communication so while he came in the area of this millimeter-wave few applications that are coming up already we understood and we have not understood expected to come so the way the propagation happens actually in the physical world the way the wireless channel changes and leave the challenge to us the existing channel models where people have struggled worldwide where we are standing for which 3gpp channel model to some extent they could get some shape but that's not the exhaustive one there are little research output in this channel modeling beamforming is the heart of this key technology that is coming up for this 5g analog it's not about 5 G's about the millimeter wave communication if you wish to do you need to have a beamforming technology and very strong beam tracking beam refinement technology have to be understood and devised without which this technology cannot come up so channel modelling and beam forming was my preferable area of little research on the last 6 years so I love these two areas I will have to be here a little bit more time and then towards the millimeter wave networks where my recently my tivities have started their side and what we understand for tomorrow's 6e communication also so I will highlight as a list of application wise this is a whole team in the school of telecommunications and that me only so we are 7 researchers working several faculty members working in the school of telecommunication we have above 60 PhD scholars in the school of telemetry communication dedicated for the researcher of this millimeter wave and above 4G 5g optical communications and the merging of also the wireless over the optical communication towards that research also we are there so in my team I would love to say that core area is the millimeter wave communications where the postdoc and PhD scholars are there one or two of them are older graduated from the team and they are doing their postdoc abroad then there is the massive MIMO so massive MIMO come very wonderfully getting coal awarded with millimeter wave because you have highly high frequency so your lambda value diminishes there is a possibility that is really you can build up a system where a massive my more antennas can be fabricated over a very small area that's where actually the real applicability of the massive MIMO comes into picture going ahead to our terahertz that town that has come up with a super massive - structure it's up to the device level as well as the engineers who are optimizing this design from the antenna side integrating actually and completely understanding that total optimization required 9 now integrating the IEF as well as that it more intelligently we will be able to handle that impedance that is happening there because we're very high frequency each and every smaller copper rack that will be fabricating of near PCB will act as an antenna so the job is how nicely we will be optimizing that radiation compared to your antenna there is a whole beam pattern is expected to be damaged you will try to turn the antenna beam pattern to some direction it will be continuously going to other direction for Satish was telling because of these issues that we are not able to handle 8 so next is the area is the optimization that's why the optimization was another interesting area that came into picture optimization starting from the circuit level to the signal level to the signal processing level to the network so network optimization for your coding modulation your impotence handling joint i f RF antenna design to the network is another very nice area where slowly we are looking into visible light communication is an area that where the some part is going on because we slowly understand for the short-range communication for some specific application this will have also come up as a big wave as the service-based research for example your automotive radar driven some applications and also the very high speed in Bakke communication within a train and all that and your AI is overrun everywhere I mean starting from the physical layer merging the joint fire and Mac optimization to your complete transceiver design driven by AI some visionary person in the yard of 2013 showed this in the EPFL one workshop that by 2020 automatically worldwide we will be targeting and demanding and 103 PS Valles should be our target beyond 2019 or beyond 2020 everybody will start thinking that how we will transfer hundred Gbps over wireless communications he didn't say that how long how far the distance why we should be able to but dreams should be something like that if that is the dream coming up later on then only at the back end the network will be also coming up hundred times higher so seeing that feature itself today's this whole millimeter wave to that era has communication and the huge bandwidth release everything so automatically slowly it happened and it gave the indication that today yes we have this dream we will be going ahead with this one in future so question starts from here what will we do with this huge bandwidth given and this huge amount of the data rate given to us so the service is there is any requirement there a simple example I have given there that if you go on for the very HD kind of the video transmission minimum data rate requirement tomorrow is coming fight for 60 bps today you are compressing the video and you were transferring if I do not want any compressed video if I wish to have uncompressed video transmission so you need its relieved the huge bandwidth so that clarity this is a demand that we have created for ourselves who told that you every day morning you we should have a very nice kind of the video to be transfer captured and to be transferred capturing camera technology yes it has been developed transmission know where it is because it is a power-hungry it is a bandwidth hungry situation other cases like these valleys data centers all the data centers today it is either having optical fiber cable connection or they are having at least a optical cable connection through the LEDs from the one section to the next what I would actually FICA wish to do it completely wireless fashion why can't we give away their their very high amount high very penciled in kind of the Valles communication from one server to the next there's a server room you can really control all those stuff because nobody's much interference also will not be there service will be static there is no mobility even if you are mobility is there that will be for the human movements and then you can also model and nicely control there so why can't we do and he really is all those connection second thing is actually from base station to best station outdoor also the people are talking about the same thing that's what he was trilling the back whole communication back to do it has not been framed currently in a proper fashion but what actually was dream as something different all your information sours the very dense network with the IOT is via cooler applications so I hope you all know that government has already targeted the bullet train applications and all things are coming up right who where is the telecommunication network when you are sitting in superb the Express train is running at 100 kilometer per hour you don't get any connection so what about see 50 kilometer per hour we are really not ready worldwide we are not ready with the telecommunication network with a extreme high speed of the believed communique bullet trains ok so there we have to start from the scratch forget about whether black network will be integrated with the 5g or 60 networks or not separately if I try to think that for the vehicular communication I will develop some new network the technology needs to be completely different the way we are thinking about the cellular communication will be running cellular communication will never see a speed of 350 kilometres per hour suddenly even if the Volvo cars are coming up they will never run on the road at 350 kilometer or 650 kilometres per hour sweet okay another case is in the regular communication the communication should not only happen between the pilot to the central pilot to everybuggy within the buggy all the users will love to have their access their internet as well as the social networking you have the same speed when they are static that is the dream so can you do that so it is actually trained to train communication we're in the buggy communication pilot to the Central Station communication pilot to the been the cellular communication in stranded infrastructure network communication so like the smart city it is a smart vehicle communication coming up can you now in understand that sitting there you can change your total travel plan you are seeing actually somebody is giving you the information that whenever you will be reaching they are some kind of the jamming all staff are there some services are broken can we immediately change sitting in the train your whole travel plan that was the question so seeing all this huge large channel bandwidth spectrum scarcity and the data rate requirement all are the services that are we are demanding each other millimeter wave communication was the obvious way there we should move and I should we should start from this point 100 years back a person from our country we shall she was a visionary person who showed us look one day 60 gigahertz communication will be in our life nobody Jie actually claim that and appreciate to that level but see they have did self 100 years back he demonstrated 60 gigahertz communication maybe with not a very finite kind of the devices horn antenna phase shifters all those power amplifiers LNA everything he himself manufactured sitting in that Bose Institute laboratory in Calcutta and that setup you read this material and the fast between that was published in the name of research admission promotion see the way he demonstrated it so if that can happen hundred years back from the same country today we should leave the world that we can do it with this demonstration if it is possible no it should happen from this country only so motivation what will you get if you move to millimeter wave not only the huge bandwidth see once something is coming with very huge bandwidth and you understand that beamforming is the only way to reach your stuff for given the question that we will be having a nice localization problem if we can do that see we are actually getting a very nice application of frequency reusing so space reusing also it's happening so that one frequency I am transmitting to somebody in the same room the same frequency I can transmit to somebody else that freedom in it is a another dimension see over the transmission we have got the freedom over time time division multiple Texas frequency division multiple access code division multiple access happened s division multiple axis we have never done it so that's his I that you call it a beam division multiple axes or you call it a space division multiple access now you join all the combination of all these four dimensions can you imagine the level of the flexibility that can happen since this is the way where they normal densification in the network can only happen the merging of all these four techniques out of this whenever I am talking about a beam forming what am i doing you know simply in the angle domain I am designing my filters we are very good in designing those in the frequency and time domain only thing that we have to understand now that when I am doing the filter design in the space domain can I do it very nicely with the speed with which the series moves thing is moving so it is the joint Doppler as well as miss frequency as well as your space domain filter design we have done it for the time frequency because delay Doppler domain we understood nicely we have not understood how we will do it for the Doppler versus angle domain or angle versus my frequency time domain angle versus my delay domain I need a modulation new complete new modulation technique I knew a complete new coding technique apart from that I need actually this kind of filter design in the angle domain right but see the special spatial resolution is a resolution that actually at every moment wise actually it can be resized and dynamically it can be changed because your location is every moment getting changed so that's where actually I comes into play because it is a very tedious job that manually you will be sensing and doing it by that time the stuff will be out so that's where actually you need to put the reinforcement learning all the machine learning stuffs to actually understand and train the whole system in such a way that this can be optimally utilize and implemented very good point is in order to do that millimeter wave has already enabled us to use the huge antenna array design understood so till today whatever the antenna design people have done and what exactly is the switching that people could do I will tell you but where the things has to be go has to go and how much time we are having for that that also we need to understand yeah the antenna arrays are also becoming a heart of your whole system design and the transceiver design it will be complex but it is doable these are the worldwide the difference frequency ranges and the application wise people are using currently I would love to show you one area see this 77 gigahertz this automotive radar it is not a dream now from Ti Bangalore they have already done it transceiver is ready they have put it on the video on the Volvo cars the whole seven seven gigahertz four cross for my mote receiver they have already gifted us for the further improvement of the spectral estimation technique with eradicable I've shown it to and Schumann so if you guys come get some time I really wish to invite you all I mean whenever you are towards that direction please visit our lab to see the kind of the actual work going on there so going ahead if we are really demonstrating something in 77th achieve and they are also via the massive - structure they have demonstrated and it is actually on the field here they have demonstrated in front of me I we can actually detect something at the gap of up to 30 meters enough actually to detect your bumps to detect your curvature that is happening to detect weak anybody who is there surrounding you any car within 30 meters of yours okay so if that and the time of the detection that is happening with that is having extreme high resolution because it is a 77 to 81 if you do that bandwidth one by that vanity will be in the nanoseconds level so question is whether signal-processing geyser that much intelligent to capture their signal and do the processing technology has the capacity to do it so as they have done it so I'm pretty sure that going ahead Crossing hundred gigahertz is a matter of few years gap fine now we cannot question ourselves whether the devices are ready or not yes the devices are ready how can they do it their ABCs are working actually over the 4 gigahertz samples per second to gigas wide bandwidth they have utilized so they are all here DC's built in the Ti is they are working actually greater than four mega Giga samples per second so if it is working so why not for the cellular or the other communications we can utilize those as it ABCs and then make ourselves smarter for the signal passing based on processing stuff so some applications we all know this yes I will pass this the very fast link setup can be done actually for all these down links the train station accession and all this verification wherever you go this with very less latency and jitter less than 5 milliseconds earlier it was but now it has even come down to the less than 1 milliseconds kind of the stuff for all your ultra-high device service download and all that this is actually the application that you wish to ensure when you are traveling in the bullet trains kind of your communication virtual reality you should be able to see whenever you are moving at a very high speed all the data centers the connectivity Internet connectivity should be valueless there is no question of putting anywheres anywhere there because everybody is static you come a with a dining hall exhibition flight high-speed rail pass classroom this is the information shower that should be actually making you up-to-date any moment anywhere you go the network should be utility you that's a booth upon a do you remember that dog eggs are moving and then there is 2 they used to actually this has come to a reality you know do you realize that that time I don't know who made that ad but today that is a reality that whenever you go the whole information I should not say the network I should say the information with the network is with you so offloading actually this part I should explain a little bit see you we are we traditionally we are really very good we need the 4G and we have achieved a very nice data rate for now don't come to the indoor actually whole data it is gone we cannot deny that fact you go to the age age age of the whole cell then also your quality of service in terms of data rate is out that was the major problem and we were trying to improve that and for enfold the LTE advanced we have tried to do that and we have done some solution also for it but for indoor the same problem we left so now what the 5g onwards it is the it should be weak you just kind of for example whatever the data rate you are getting outside that same data rate people are demanding inside the high speed rail same data rate we are demanding when we are actually below the mines and the chilleen's inside your indoor environment below mines I don't know but at least indoor environment to the outdoor environment at least it should be be cutest because indoor is if you are even in a static zone so it should happen yes so for that actually people are thinking actually if outdoor is working at the 28 gigahertz with a 1 gigahertz bandwidth hinder should be with the 60 gigas 1180 specification is already frozen so it should be offloading to Wi-Fi and then 60 gigas means AP will be working inside so he will be taking the load of that exactly and AP should be talking to the base station at least that much emu is done between your Wi-Fi versus the outdoors so a piece will be smart enough to talk to the west station's 5g via stations and away if we are not going to do that if that handshaking does not happen he be good as transmission is not what soon that is a multi rat has to be inside it should be the 60 gigahertz communication how do these 28 gigahertz so offloading should be happening from 28 to 60 gigahertz then also there is a mismatch in the bandwidth also so how will you in the indoor that data rate you will be maintaining and all that that is a question of your plan specification talks about actually kept it open that how to do this handshaking now the a piece of the de l'avenir d they will be actually open enough intelligent enough protocol he has been already given to share and handshake ad is actually single user X has come with my move a y also you have seen that in China they have already demonstrated protocols are there question is actually whether technology handshaking has you see actually physically finally you have to transmit that has to happen signal processing guys have to decode otherwise this whole problem of the standardization and everything is then lost I mean somebody sitting in the ground level has to do the research to make it happen then only the meaning of standardization comes into picture so to make it happen again people realize that the way we have designed traditionally the network side that has to be diminished see everywhere we have done some redundant design every base station is having its own I F and gaze ban section so the amount of the cost that is involved the amount of the processing power that is involved and also the latency that we are targeting that is not going to work if you are going like that way fine so if we try to also handshake the base bar multiple base stations exchange the information multiple base stations people came up with the fact let's tomorrow we will have only the antenna and some part of the antenna circuitry with the best ations that will be the definition of base station rename this as the radio resource and whole base band pool let us cool actually to a common cloud very big step according to my understanding because I have seen actually these days in tents all the projects are behind how to manage their high processing speed service how to do optimization resource sharing everything there itself and biggest question how to provide security because multiple such base station signals sharing their baseband data there and multiple day where they have to do the resource allocation that also with that minimal way minimal latency and all that and another biggest question the people like Aussies for example I have captured my signal data may username to our irate one channel is already troublesome for me to do the channel estimation for massive mine alright then I am sending after that the ADC samples from the Phi but is directly going to my cloud and then cloud is doing all that isness so my channel from user into me and my resource head to the cloud they are not same people are telling this side will be wireless but your are arranged to the basement pool that will here either your Ethernet port or will be optical channels are only changed who will be doing channel state estimation for me and where is it going on currently nobody knows fine if I do it sitting on the other H I need to put some general estimation Block in a RF because my signals after all it got sampled and it is out so I thought I will save the power so I have not give me any baseband processing

Channel: ACCS India

Views: 3,112

Keywords: Science & Technology

Id: Rt09KWCEUGg

Length: 26min 5sec (1565 seconds)

Published: Sat Oct 12 2019