Skip to content
Best Place To BuildBest Place To Build

How IIT Professors created India's own 5G Tech | Prof. Radha Krishna Ganti, EE, IITM on BP2B S2 Ep.6

In this episode of The Best Place to Build Podcast, we sit down with Dr. Radha Krishna Ganti, Professor of Electrical Engineering at IIT Madras and Vigyan Yuva Shanti Swarup Bhatnagar awardee, to uncover the untold story behind India’s first 5G phone call using indigenous technology. From building base stations from scratch to powering India's digital future, this episode reveals how IIT professors, students, and researchers laid the foundation for India’s 5G revolution. And what comes next with 6G research? ✨ What you’ll learn in this episode: • The inside story of India’s first 5G call with homegrown tech • Why 5G networks, GSM architecture & wireless communication are critical for India’s digital growth • How Electrical Engineering, Electronics & Communication (EE, EEE, ECE) at IIT Madras drives telecom breakthroughs. Plus what’’s the difference between all 3 branches? • What’s beyond 5G—the roadmap to 6G innovation 🔔 Subscribe to the channel for more conversations with builders, professors, and entrepreneurs shaping India’s future! And if you’re looking for a specific part, jump to it from the directory here: 0:00 Introduction 0:40 Welcome to The Best Place to Build 1:15 Meet Dr. Radhakrishna Ganti: IIT Professor & Bhatnagar Awardee 2:45 How Does Wireless Communication Work? 5:20 What exactly is 5G & Why It Matters for India’s Digital Future 7:05 From GSM to 5G: The Evolution of Wireless Networks 9:00 Building Indigenous 5G Technology from Scratch 12:10 The Nerve-Wracking Moment of the First 5G Call 13:40 Challenges in Scaling 5G Networks Across India 14:20 4G v/s 5G v/s 6G 18:50 Use-Cases of 5G 20:00 Verticals of 5G Technology 21:05 Who defines a country’s 5G needs? 25:05 The Story Behind BSNL’s 4G Deployment 29:05 Why Should Youngsters Be Motivated to Join EE? 32:08 Why Do People Say Electrical Engineering is Hard? 34:35 Professor Bhaskar’s (Ex-Director of IIT Madras) Legacy in Wireless Communication in India 36:00 GSM to 2G to 5G 40:10 Professor Ganti’s Insti Story 42:55 The Shift From a Theoretical to a Practical View of Systems Studies 46:10 Is IIT Madras Too Obsessed With Theoretical Learning? 47:00 What is The Difference Between EE, EEE & ECE? 51:00 How is AI/ML Affecting the World of Electrical Engineering? 52:40 Closing Thoughts & Reflections #bestplacetobuild #iitmadras #electrical #5g #techpodcast #makeinindia

Dr. Radha Krishna GantiguestUnknown Hosthost
Aug 29, 202554mWatch on YouTube ↗

EVERY SPOKEN WORD

  1. 0:00 – 0:40

    Introduction

    1. RG

      See, the math that you learn in electrical engineering is the basis of most of the AI ML that you do.

    2. UH

      Okay.

    3. RG

      Okay? So that way, electrical engineering people learning AI ML is actually very easy. We have, I mean, have seen this whole cycle of, uh, 5G. It's a big team effort, but this requires dedication of, like, the timescales are ten years. We felt that if we don't attempt to build this, it, it's always going to be, like, a distant dream.

    4. UH

      We do a lot of work with students who come in to campus through this program called Ask IITM. I have heard a lot of parents say that IIT Madras is supposed to be a theoretical campus, uh, and they're quite worried about it, or they're like... But when I talk to professors, when I talk to alumni,

  2. 0:40 – 1:15

    Welcome to The Best Place to Build

    1. UH

      they seem to have been doing a lot of building. Hi, this is Amrit. We are at IIT Madras, my alma mater, and India's top university for people who like to build. We are here to meet some builders, ask them: What are you building? What does it take to build? And what makes IIT Madras the best place to build? [upbeat music]

  3. 1:15 – 2:45

    Meet Dr. Radhakrishna Ganti: IIT Professor & Bhatnagar Awardee

    1. UH

      Hello, and welcome to The Best Place to Build Podcast. Today, we are sitting with Professor Radhakrishna Ganti. He's a professor of electrical engineering at IIT Madras in the field of wireless communication. He's also the twenty twenty-four winner of the Shanti Swarup Bhatnagar Prize. Young Scientist Prize, is that the right word?

    2. RG

      Yes. [chuckles]

    3. UH

      Congratulations, Professor, and welcome-

    4. RG

      Thank you.

    5. UH

      -to the podcast.

    6. RG

      Thank you. Thank you for having me here.

    7. UH

      Professor, we also had, uh, Professor Prabhu Raghu Gopal last year on the podcast, and I think the two of you got the prize together. Must have felt really good, right?

    8. RG

      Yeah, it, it, it was a absolutely fabulous feeling, but an eerie feeling also. I never expected [chuckles] the prize to come to me, and there was a huge gap of, like, one year before the... Um, not one year, about eight months before they announced the prize, and it was a sudden surprise one day, I see an email saying that, "Yeah, you got the prize." Yeah.

    9. UH

      You mean, eight months from the-

    10. RG

      Actual-

    11. UH

      -nomination?

    12. RG

      -date of nom... No, no, uh, from the time the prize was supposed to get announced. It got delayed because of elections last year.

    13. UH

      Okay.

    14. RG

      So they couldn't announce the prize, and we- almost of us forgot that we have submitted this, [chuckles] and, uh, suddenly it was a boom, we got this, uh-

    15. UH

      Nice. Congratulations, Professor.

    16. RG

      Thank you.

    17. UH

      There aren't too many prizes for professor... academic life, right?

    18. RG

      Uh, I guess so, yeah.

    19. UH

      Nice. Um, Professor, your work is in the field of wireless communication, a field that I understand really nothing of. [chuckles] Uh, I do remember, uh, in my first year, I was in mechanical, and, uh, my friends in electrical engineering had this project that they would carry one of these Nokia phones around campus with a notepad, and every twenty meter, they were supposed to stop and note down the

  4. 2:45 – 5:20

    How Does Wireless Communication Work?

    1. UH

      signal strength.

    2. RG

      [laughing]

    3. UH

      That's, that's my last, uh, [chuckles] briefing with electrical engineering and wireless communication. So if I, if I may ask, how does wireless communication work?

    4. RG

      Okay. So just let me first explain you, uh, a little bit about where do you see wireless communication. I'm, I'm pretty sure all of you know this. Your cell phones, all of you must have looked at 4G, 5G, at least you must have heard this-

    5. UH

      Yeah

    6. RG

      ... uh, yeah, abbreviations at some point. You're using Wi-Fi. You're using Bluetooth. Yeah. So all of these are wireless communication. Of course, you're, you're getting Tata Sky, you're listening to FM radio. So these are all examples of wireless communication. Essentially, wireless communication deals with: How do you transfer data between two points, either in sp- uh, in space, uh, in a reliable fashion, okay, when they are not connected by, of course, any physical wire, wired, uh, medium? Okay, so it's a, it's a very, very challenging thing, okay? Uh, so for example, nowadays, uh, people want to go on high-speed trains that are traveling at three hundred kilometers per hour, or a hundred and fifty kilometers per hour, and, uh, want to watch Netflix or, um-

    7. UH

      Right

    8. RG

      ... YouTube at those speeds. So how do you even enable this? Okay, it's not, it's not such a trivial thing, because when you're transferring data and things are moving so fast, uh, you see a lot of, uh, channel effects, what you call channel effects. Essentially, they, they distort your signal that you're sending. So our goal is, how do you overcome these distortions? How do you overcome noise, and but reliably transfer as much information as possible in the shortest possible of, uh, shortest amount of time?

    9. UH

      Okay, so what I understand is that, say, the mics we are using-

    10. RG

      Yeah

    11. UH

      -they're wireless mics. So there is some information from our voice which is being captured by the mic and wirelessly transmitted.

    12. RG

      Yes.

    13. UH

      But the challenges increase when there's more data and when the person is moving very fast, as in a train.

    14. RG

      So I, I can give a few examples. So if, uh, so suppose you are like, uh, you are in a party, you are talking to each other, but when your, uh, friends who are, who are beside you start talking, you, you have difficulty in hearing.

    15. UH

      Yeah.

    16. RG

      And now imagine this: there's so many wireless devices around. Yes? So everyone, every wireless device wants to talk to its receiver or the base station, okay? So how do you ensure that all these things go through without any, uh, echoes? Okay.

    17. UH

      Sure. We've all had this experience where we are at a cricket match or a stadium or a concert, and too many cell phones, and nobody's getting any signal.

    18. RG

      Nobody. So how, how do you enable that kind of thing? So these are precisely the problems that we try

  5. 5:20 – 7:05

    What exactly is 5G & Why It Matters for India’s Digital Future

    1. RG

      to solve in wireless communication.

    2. UH

      Okay. So, Professor, um, we have heard that IIT Madras has a 5G test bed, and it's a very prestigious project. There's a lot of press around it, uh, but can you explain to us what the 5G test bed means? What it really... Why is it important?

    3. RG

      Okay. So before we get into the 5G test bed, let me take some time and explain first what 5G is, okay? What are the components that go into it, and then I will tell you why 5G test bed, uh, was necessary and what we did in it. Okay? So there are two parts of any 5G network- 5G system. One is the 5G network, and the other is the 5G, uh, cell phones. Okay? So the 5G network, when you drive around, you see all these base station towers that are there on the, uh, on, on the top of the, uh, towers, you have the small equipment that are put on the top of the towers. Okay? So these are called the 5G base stations. So these consist of radios which transmit and receive signals. They consist of signal processing units, which take these signals, digitize them, process them, convert them into bytes and IP packets, and-... and on the back end, you have something called the 5G core network, okay? Now, this data is transmitted from the, uh, 5G radio to the 5G core network, okay? So, uh, i- if you look at the 5G base station, uh, the 5G base station that's there on the top of a tower, it's a very complex piece of equipment, okay? So there are antennas in it, there are RF chips in it, uh, there are, uh, processing units in it. There's are lots and lots of s- software, which is signal processing oriented, okay? Uh, on- that's on the base station. If you look at the core network, so i- if, if you, I mean, uh, so for the entire south zone of India, there'll be one core network or two core networks, okay? So each core network has to actually handle, like, uh, hundreds of gigabytes of data

  6. 7:05 – 9:00

    From GSM to 5G: The Evolution of Wireless Networks

    1. RG

      per second, I mean, per second. Hundreds of gigabytes of, uh, data per... GBPS of, uh, data, okay? So this is a very, very complex network, okay? There are maybe like, uh, four to five companies across the world who can build this kind of infrastructure.

    2. UH

      These are the Samsung, Ericsson, Huawei.

    3. RG

      Samsung, Ericsson, Huawei, Nokias, and, uh, ZTE. I mean, I think there are a few smaller companies, but these are the bigger companies that most of you... And, and they spend billions and billions of dollars on getting this kind of, uh, equipment, R&D- on R&D and building this kind of-

    4. UH

      And also, they've spent thirty, forty years building that.

    5. RG

      Abs- you're absolutely right. They have like, uh, Ericsson, uh, prizes itself, saying that it's a hundred-year-old company. I mean, from the time of Marconi, it built radios, uh, uh, that's its place.

    6. UH

      Is it Marconi's company?

    7. RG

      It's not [chuckles] Marconi's company-

    8. UH

      Okay.

    9. RG

      - but they said from that time they have, uh, uh, built it. Okay, so, uh, so around, uh, five years back, uh, when 5G was getting standardized, uh, uh, we started thinking about how can we build this kind of infrastructure in India. Unfortunately, at that point, there were not too many... There were not almost any companies who are doing this. Uh, so people who... All the faculty, most of the faculty who are working in this kind of, this, uh, space got together, and we wrote a big proposal to, uh, Department of Tele, MIT initially, and then to Department of Telecom, saying that, "We will build this infrastructure." Okay, so it's a very, very ambitious project. I mean, uh, building this kind of whole infrastructure is like, uh, as I was saying, it's a billion-dollar projects with, like, hundreds and hundreds of PhDs working on it. But nevertheless, we felt that if we don't attempt to build this, it, it's always going to be a distant dream.

    10. UH

      If I can just, uh, tell, if I can just say it in my way, what I've heard you say so far, so a network contains, uh, base stations and core network.

    11. RG

      Core network.

    12. UH

      There are lots of base stations, one core network for many base stations.

    13. RG

      Absolutely. Absolutely right.

    14. UH

      Uh, and the base stations and the core networks, the technology that goes into

  7. 9:00 – 12:10

    Building Indigenous 5G Technology from Scratch

    1. UH

      them is very sophisticated, and really large multinational telecom companies sell these. We don't have a homegrown company.

    2. RG

      Yes.

    3. UH

      So the challenge from your team was to build that homegrown technology?

    4. RG

      Yeah, to build this homegrown technology. So, uh, eh, so essentially, eight institutions got together, and, uh, uh, we started building components of this. I mean, uh, each of, each of the institute took, uh, one part of this. So for example, Madras took the, uh, took the onus of building the radio part of things, okay?

    5. UH

      Mm.

    6. RG

      And also, we are the final integrator of all the things.

    7. UH

      Okay.

    8. RG

      So this required huge teams working together, software teams, hardware teams, embedded systems teams, RF teams, antenna teams, to working together to get this thing going. Uh-

    9. UH

      Fair enough. The number of patents that these companies would have filed just to get that, that will be in the thousands.

    10. RG

      Few thousands, easily.

    11. UH

      And it's all protected technology. So if you want to rebuild it, you have to rebuild every-

    12. RG

      Everything of this, okay? Uh, so it took us about four years. In between, we had, of course, the COVID, we had to go through that whole cycle, and, uh, uh, we built that. So a- at this po- so when the project started, it was started as a testbed project. So what was a testbed? Now we'll answer this question.

    13. UH

      Yes.

    14. RG

      So when we started this project, we, we, we thought that, uh, we'll build something in the lab scale.

    15. UH

      Mm.

    16. RG

      Okay? So we'll build these components, uh, not something that you can put in a field, but you can put in a lab, test them out, and also allow outside people to come and test their products, algorithms, and ideas on this hardware and software platform that we built.

    17. UH

      Mm.

    18. RG

      Okay? So this was what the government funded us for.

    19. UH

      Mm.

    20. RG

      Uh, so, uh, during the course of the project, we realized that, okay, we don't have to build a testbed. We can almost build a product that is close to TRL level seven or 7.5.

    21. UH

      TRL is technology readiness level.

    22. RG

      Technology readiness level, seven.

    23. UH

      And se- four is lab-ready.

    24. RG

      Four is lab-ready, so we can go, like, three notches up. That's something almost close to a field deployment. Uh, we said that we'll build something of that order.

    25. UH

      Okay.

    26. RG

      Uh, and, uh, so what we did was, we followed all industry practices, the interfaces and everything we kept what other people did. So we went to a lot of trade shows. We, we talked to people. We figured out what people wanted. We did, uh, uh, actually, we did a customer survey, effectively, you have to put-

    27. UH

      Mm

    28. RG

      ... uh, during these three years, and we were able to build that, okay? So at the, at the end of this project, we had, we had base stations, we had core network that can be deployed anywhere. So for example, in our current campus, we have our own 5G deployment that is there. Not many people know this.

    29. UH

      Okay.

    30. RG

      We used to, we use this to test our, a lot of algorithms that we develop, a lot of, uh, new products that we develop, we, we test it, okay? From a testbed perspective, also, we were very successful. This is a product perspective. From a testbed perspective also, a lot of startups from India, a lot of companies from India, have come and used this facility to test their products-

  8. 12:10 – 13:40

    The Nerve-Wracking Moment of the First 5G Call

    1. UH

      and-

    2. RG

      Smaller deployments

    3. UH

      ... I, I think, uh, before this podcast, you told me that India's first 5G call was on this network.

    4. RG

      Oh, yeah. So one of the exciting things that happened was, uh, almost close to the end of the project, uh, uh-... uh, uh, honorable Minister of Communication at the time was, uh, Ashwini Vaishnaw Ji. He, he was-- He came here and made the first, uh, phone call, 5G phone call in India, uh, in our lab. Uh, it was a very exciting and a tense time for, uh, all of us. I mean, uh, I remember that time. So we were doing this in the research park in a small room. Uh, we had this base station put up and cell phone put up. We are expecting that only Ashwini Vaishnaw Ji will come into the room and make a call, but suddenly this huge rush of, uh, reporters and everyone in the room, they started blocking the base station. And, uh, uh, my student, my student was sitting at the base station. Like, so, so he said that he was whi- uh, vigorously fighting me. So he said, "The signal is not going through because people are blocking it, people are blocking it." So he made the decision that I will increase the transmit power at that time, and-

    5. UH

      Okay

    6. RG

      ... uh, Minister was able to make the call, and after that whole system went down for some time. [chuckles]

    7. UH

      [chuckles]

    8. RG

      It was a nerve-wracking moment, but nevertheless, it was a very exciting moment, uh, that the first 5G phone call was being made at, with the homegrown technology.

    9. UH

      That's amazing.

    10. RG

      Yeah.

    11. UH

      Uh, I mean, now, of course, uh, uh... But just, just to clarify this, we have 5G deployment in-

    12. RG

      Yeah

    13. UH

      ... India by Airtel and Jio.

    14. RG

      Yes.

    15. UH

      Do-- They still use the Ericsson. I don't know which company they use-

    16. RG

      Yeah

    17. UH

      ... but they still use foreign equipment?

    18. RG

      They still use foreign equipment. Okay, uh, that, that's unfortunate, uh,

  9. 13:40 – 14:20

    Challenges in Scaling 5G Networks Across India

    1. RG

      thing. Uh, see, the thing is, whatever we have done is up to TRL level seven point five eight, okay? We require companies in India to take this kind of technology. Also, develop their own technology. I'm not saying that they have to take, uh, uh, take this technology and take it to TRL level or whatever is required, uh, TRL. That also requires significant investment.

    2. UH

      Yeah.

    3. RG

      There are some companies which are now doing it, C-DOT, Tejas Networks. Uh, these are some of the companies who are trying to take these kind of technologies to, uh, uh, those thing. Even Reliance has its own, um, uh, does build its base stations, so I think in the future they may deploy it.

    4. UH

      Fair enough. You're saying that from an academia point of view, this is the farthest we can go?

    5. RG

      Yeah.

  10. 14:20 – 18:50

    4G v/s 5G v/s 6G

    1. UH

      And then some companies have to take that and push it into-

    2. RG

      Absolutely

    3. UH

      ... uh, product. Fair enough. Uh, Professor, one thing that we glossed over, um, what is 5G, 4G, 3G-

    4. RG

      Okay

    5. UH

      ... 2G, 6Gs, and, and what is G?

    6. RG

      Okay, so this is a very interes- this is a very interesting question. So who defines 5G, or who defines, for that sake, 4G or 6G, okay? So, uh, not many people know this. People, we just generally use, uh, 5G in a very colloquial sense, saying, "Oh, my phone is upgraded from 4G to 5G." What exactly is 5G? So every, every cycle of G, generation, G stands for generation, takes about ten years, okay? So twenty twenty to twenty thirty is the 5G, fifth generation. Twenty thirty to twenty forty would be the sixth generation, okay? So, so what makes a technology or a device 5G technology? So this, this is where International Telecom Union comes into picture. ITU, what we call ITU. ITU is a UN body, so, so the, its main, uh, job is to regulate spectrum. This is so where... So I think you also heard of spectrum allocations, no?

    7. UH

      Yeah.

    8. RG

      I mean, so there's a huge amount of money that government-

    9. UH

      This is the microwave, uh, segment, which cell phones can use commercially.

    10. RG

      Commercially, okay? So for example, in 5G, uh, Reliance uses, uh, three point four to three point five gigahertz band. And so they bought it, they bought it from the government, okay? So before they bought it, the government has to allocate the spectrum. So this spectrum has to be harmonized across multiple countries for lots of technical reasons, okay? This harmonization is done by ITU. That's its main, uh, one of the main jobs to do.

    11. UH

      Sure, this is easy to understand because it's not like I bought it, means I can lock it up-

    12. RG

      Yeah

    13. UH

      ... it is in the air. So if I bought a license, it means that somebody has to make sure nobody else is using it.

    14. RG

      Absolutely.

    15. UH

      So-

    16. RG

      Absolutely right, okay? And when you're close to borders of other countries, I mean, there may be interference, so the things have to be aligned in some, uh, broad sense, okay? So in addition to that, ITU also defines what e- every generation has to be. What does it mean by define? Means it tells that for a technology to be 5G, it has to have this much data rate, okay? It has to support these many devices, okay? It, it has to have this much latency.

    17. UH

      Okay, so a 5G base station will be a 5G base station if it adheres to certain standards.

    18. RG

      Certain KPIs, key performance indicators.

    19. UH

      KPIs.

    20. RG

      There are about twenty, almost close to twenty KPIs.

    21. UH

      Suddenly, it sounds very corporate.

    22. RG

      [chuckles] It's KPI. It's called KPI, key performance indicator. That's the term partly also that, uh, ITU uses, okay? So it's also very interesting. So we had a huge journey on, uh, for 5G in ITU, okay? Let, let me explain you a little bit of this, uh, in this, uh-

    23. UH

      Professor, before that.

    24. RG

      Yeah.

    25. UH

      Can you give me one example, like for instance, say, number of, uh, connections?

    26. RG

      Okay.

    27. UH

      How has it moved? How has the standard moved from, say, 2G to 3G to 4G to 5G?

    28. RG

      Right. Uh, so, uh, okay, so, uh, before the, before the number of connections, I have to tell you one technical point. Uh, so let's look at the bandwidth first. So in the bandwidth of 2G was about, uh, couple of megahertz.

    29. UH

      Okay.

    30. RG

      About two mega- close to two megahertz or so, okay? 4G, it became twenty megahertz per carrier, so per, per operator, that's what I'm saying, okay? Four- 4G, it became twenty megahertz. 5G, it's hundred megahertz, okay? So now you can see that the bandwidth is increasing. As the bandwidth increases, you can see that the number of users can also increase, because you are having more space to accommodate them, effectively more space in the spectrum to accommodate. So you can see approximately a five X time... I mean, it won't be five X in the exact, uh, sense, but up close to five X number of users can latch on to 5G than compared to 4G, because the spectrum is larger.

  11. 18:50 – 20:00

    Use-Cases of 5G

    1. RG

      enhanced mobile broadband. Okay? So that's what 5G does, you know, you have higher bandwidth.

    2. UH

      Oh, I just realized that in the beginning you said something about a person in a train watching Netflix.

    3. RG

      Yes.

    4. UH

      That sounds like a use case.

    5. RG

      That sounds like a use case. So this comes from the requirements that ITU defines. So ITU says that you will, uh-- the technology should support up to five hundred kilometers per hour mobility.

    6. UH

      Okay.

    7. RG

      Okay, so that's one of the KPI.

    8. UH

      Okay.

    9. RG

      Okay, so that's the difference. So in, uh, in fi- 4G, it was about hundred and twenty kilometers per hour. In 5G, it's five hundred kilometers per hour. In 6G, they're talking about twelve hundred kilometers per hour, because people moving on the airplane should get, uh- [chuckles]

    10. UH

      I just want to say that no train in India moves to five- [chuckles]

    11. RG

      I know, I know this. [laughing]

    12. UH

      Sounds like a Shanghai problem.

    13. RG

      E-exactly. I mean, you're, you're, you're, uh, on the spot. So this requirement came from our, uh, neighboring country.

    14. UH

      Okay.

    15. RG

      Yeah.

    16. UH

      Okay, what does that mean? Like...

    17. RG

      So what does that mean? So, so every country gets their requirements to ITU.

    18. UH

      Okay.

    19. RG

      Okay, saying that, okay, uh, for 5G... Uh, I-I'll take the example of 5G, okay? We want, uh, these many devices to be connected. We want the speeds to be this. Okay, we want this particular use case for te-telemedicine to be enabled.

    20. UH

      Possible.

    21. RG

      Okay.

  12. 20:00 – 21:05

    Verticals of 5G Technology

    1. RG

      Uh, so one of the 5G, 5G has multiple verticals. I was saying, you know, one is broadband space, the other is something called IoT, massive machine-type communication. Essentially, thousands of millions of devices should be allowed on the network, which transfers small amounts of data, sensors. Okay? There's something called, uh, ultra-reliable communication in 5G. This is to have 5G in fa-factory floors.

    2. UH

      Hmm.

    3. RG

      Okay? Or do remote medicine.

    4. UH

      Hmm.

    5. RG

      Okay, so where the reliability of your link will be like fine science. Okay, and you have very, very low latency, so you can play video games. I mean, ess- essentially, it's like if you play video games on a 4G network, you see a lot of jitter.

    6. UH

      Okay, so I hear you. Uh, you are saying that there are a wide variety of use cases.

    7. RG

      Wide variety.

    8. UH

      It's not just about how fast you move or how many connections are there-

    9. RG

      Absolutely

    10. UH

      ... but sometimes there's a reliability issue, sometimes you need to have millions of small devices, maybe like EVs-

    11. RG

      Yeah, absolutely

    12. UH

      uh, need to be online.

    13. RG

      Or your smart meters.

    14. UH

      Smart meters, yeah.

    15. RG

      Yeah.

    16. UH

      India has a smart meter-

    17. RG

      Smart

    18. UH

      uh, transition happening that's going to add, like, two hundred million-

    19. RG

      After, you can see the numbers-

    20. UH

      Hundred million

    21. RG

      IoT

  13. 21:05 – 25:05

    Who defines a country’s 5G needs?

    1. RG

      devices. So...

    2. UH

      Oh, so those IoT devices will be connecting directly to a 5G-

    3. RG

      5G network. Okay.

    4. UH

      Okay.

    5. RG

      So-

    6. UH

      But you said that the country presents. So who in a country presents?

    7. RG

      So let me tell you this example from India. So one of the things that we felt, uh, from Indian perspective was that, uh, rural coverage is very important. Okay, so because if you look at Indian, Indian population, still a significant amount of population is in the rural area. Okay? The question is: how do you cover them? How do you, how do you provide the 5G coverage to them? Okay, there are multiple ways of doing this. One thing you should understand is we are a price-sensitive economy, and on top of that, the, uh, the revenues that you get from rural areas is on the bit on the lower side. So there's not too much incentive for companies to place too many base stations in the rural areas to cover everything. Okay? So we came up with this idea that we want to place one base station in the Gram Panchayat, and the Gram Panchayat, uh, from there, the base station should cover a large area, the neighboring villages.

    8. UH

      Okay.

    9. RG

      Three, four villages around it.

    10. UH

      Area of the base station itself becomes a specification.

    11. RG

      Specification. That's one thing that ITU specifies. Okay?

    12. UH

      Okay, so India would push for a higher-

    13. RG

      Higher radius-

    14. UH

      Mm

    15. RG

      ... of coverage. So we went. Okay, so essentially, what happens is the Department of Telecom participates in ITU. So, uh, uh-

    16. UH

      When you say "we," you mean Department of Telecom?

    17. RG

      Department of... We, as representatives through Department of Telecom, go there.

    18. UH

      Oh.

    19. RG

      Okay, so we took this idea. We, I mean, uh, myself and a bunch of other folks, uh, uh, took this idea, and we went to ITU saying that we want this kind of large coverage, okay? And, uh, as maybe you can expect, there was a huge pushback towards it, saying that, uh, we will not allow this kind of a thing. Okay? Uh, be-because from the Western side, where-- from the Western notion of things, this coverage is not a very critical thing. For them, it's higher speeds, lower latencies. These are the things that make things exciting. But from an Indian perspective, I mean, if you don't have coverage, what is the point of having all of, uh, other, uh, fancy things?

    20. UH

      Is it fair for me to say that if a country which doesn't need this feature allows their country to have those features, then they have to invest extra for an extra feature?

    21. RG

      Maybe you have to invest, or, or there may be because there's... See, other countries have companies in it, uh, they call-

    22. UH

      Oh, so it becomes like a political thing.

    23. RG

      It's a, it's a political, commercial, uh, issues. So but we persisted. I mean, we-- took us about three, almost close to two and a half years of fight with lots of countries, I mean, engagements with a lot of countries, to push this specific idea into ITU.

    24. UH

      Okay.

    25. RG

      It's called LMLC, a low la- li- low mobility, large cell. It's called low mobility, large cell for coverage. Large cell, you understand, uh, when-

    26. UH

      Yes.

    27. RG

      Okay.

    28. UH

      Because cell, cell-

    29. RG

      Yeah

    30. UH

      ... phone tower, got the cell.

  14. 25:05 – 29:05

    The Story Behind BSNL’s 4G Deployment

    1. UH

      [chuckles] Um, professor, uh, there was a point where, uh, we were talking about... I think maybe in our preparation call, we were talking about BSNL and, uh, the deployment of the BSNL network.

    2. RG

      Okay.

    3. UH

      Um, and I found, I found it very interesting. Can you just, uh, share that?

    4. RG

      Okay. Uh, so, uh, uh, let me start with this. A lot of people don't know what is happening with BSNL 4G, okay? Uh, people, we have this, uh, habit saying that BSNL, "Why hasn't BSNL come up with 4G?" It has come up with 4G, but there's a nice, interesting story behind this whole thing. So BSNL, uh, BSNL was delayed in deploying 4G. We have to agree with that, because a lot of other issues which are not technical in nature.

    5. UH

      Yeah, they're political issues.

    6. RG

      Political issues in nature, okay? Uh, then, at some point, the government decided that, uh, BSNL has to get into 4G, and, uh, so the BSNL tender was floated, and, uh, at that point, what happened was, uh, not many Indian companies were able to get into the tender for a variety of reasons, okay? Because-

    7. UH

      We were companies, less of-

    8. RG

      Le- they had-

    9. UH

      Proof of technology.

    10. RG

      Proof of technology was not there effectively, okay? Uh, so at that point, uh, our neighboring countries, I mean, a lot of our networks come from our neighboring countries. Uh, I, I don't want to name the country. You know who it is. [chuckles]

    11. UH

      I mean, there is only-

    12. RG

      One country

    13. UH

      ... one neighboring country which has the capability of-

    14. RG

      Yes. So at, at that point, what happened was Galwan attack happened.

    15. UH

      Okay.

    16. RG

      Okay, so during that time, Galwan attack happened, and, uh, uh, the government made, made a very right, right decision, saying that, uh, uh, our network should be secure, which is very, very important, and we should build this completely homegrown.

    17. UH

      Mm.

    18. RG

      Okay? So actually, it's a, it's a very, very bold decision by the government. I mean, uh, y- you understand these networks cost, like, tens of billions of dollars. These are not like, uh, a few hundred crores also. It's like tens of thousands of crores. They made a very bold decision, and because of that, what happened was, uh, they said that Indian company, Indian c- only... whatever we deploy in BSNL network has to be coming from, completely from India.

    19. UH

      Okay.

    20. RG

      So this was the, uh, premise with which BSNL deployments have started.

    21. UH

      So these are two decisions. One is that no purchase from a country bordering with a land border to India.

    22. RG

      Land border to India.

    23. UH

      And second is that BSNL will be using only completely homegrown-

    24. RG

      Homegrown technologies.

    25. UH

      But this was at a time when there was no complete homegrown technology.

    26. RG

      There was no... Exactly, exactly. So what-- So they, they were also very patient. So what they... I mean, uh, we had some things, but it was not at the scale at which, uh, you can do a commercial deployment.

    27. UH

      Mm.

    28. RG

      Okay? So then they called for a proof of concept, okay? They, they formed a- so essentially, what companies also understood was that individually, none of the companies had the, uh, money power or the technology prowess to do the entire system. So they themselves formed, like, two consortiums consists of multiple companies, and these consortiums, actually... Uh, so one of the consortium, for example, had Tatas in it.

    29. UH

      Okay.

    30. RG

      So that, that gives you a lot of power, you understand?

  15. 29:05 – 32:08

    Why Should Youngsters Be Motivated to Join EE?

    1. UH

      bit. Um, we spoke about standards, we spoke about what it means for the country and all of that. Maybe we can dial back a bit. You are a professor of electrical engineering, and, um, I'm curious to know... All this sounds very fancy, but I'm curious to know, when, when I was a student, uh, we were attracted by, you know, rockets and planes and things that move, robots. Um, and so, you know, you would, you would see a plane and you'd think, "Oh, I'll join aeronautical engineering." What is the... For, for young people, and maybe if, if you can reflect on your own experience, what is the driving force in electrical engineering? What is those things which are really exciting about elec?

    2. RG

      Uh, I mean, uh, so electrical is a very big field, okay? So electrical engineering covers all the way from motors, transformers, power generation, all the way to wireless communication, quantum mechanics. There's so many. I'm gonna focus, um, since I'm in-

    3. UH

      Yes

    4. RG

      ... I work on wireless communication-

    5. UH

      Yeah

    6. RG

      ... I'm going to answer this question from that perspective, okay? See, uh, I mean, every day, we use our cell phone, we use our laptops. I mean, we are in the digital world. Yes? Anywhere, any, anytime a bit gets consumed and a bit gets generated, you have some, uh, you- it's because of some engineer from electrical engineering.

    7. UH

      Mm.

    8. RG

      Okay, there's no, no, it's no second guessing. It's not mechanical, [chuckles] it's not aerospace, it's electrical, okay?... So here you talk about information. I mean, electrical engineering, specifically wireless communication, talks about information propagation, information storage, information retrieval.

    9. UH

      Mm.

    10. RG

      This is exactly the world we live in, information. Okay? Uh, so, so, so what, what, what excites me, okay, is so if you take two chips, so you have processors in your thing, okay? You have a USB stick, okay? Every day you put something in the USB stick, and then you want to transfer your photos from USB stick to this, and you have gone across the USB 2, USB 3. Speeds are going up. How does this happen? Well, how does speed go up, and what is the thing that we are doing? Why, why does the speed go up?

    11. UH

      Mm.

    12. RG

      Okay. Uh, it's because, like, there's a thousand engineers who are working on that particular technology to increase your speed, okay? Uh, you're, you're talking about image processing, for example. Image processing is not wireless, okay? Uh, so all this stuff happens in electrical engineering. So electrical engineering is the field to be. [laughing] Not computer science, not AI. [laughing] So... Okay, saying that, uh, see, there's a lot of fundamental work we do in electrical engineering, okay? Uh, so a lot of LLMs, a lot of, uh, what do you say, neural networks that you see-

    13. UH

      Mm

    14. RG

      ... what people are attracted. The basis, the foundational math is actually done in electrical engineering. Okay, so w-what-whatever... S-signal processing is the foundation of, I think, the modern economy. I mean, you cannot deny it.

    15. UH

      Yeah.

    16. RG

      And the, and the basis for this happens in electrical engineering.

    17. UH

      Mm. Come here-

    18. RG

      Yeah

    19. UH

      ... I think, as you're saying it, I remember this term called SNR, signal-to-noise ratio-

    20. RG

      Yes

    21. UH

      ... which is now used in the business world, but it's actually a very engineering term.

    22. RG

      It's a very communication term.

    23. UH

      Yeah.

    24. RG

      Signal power by noise power. Yes, so, uh, it, it comes from us. Us. [chuckles]

    25. UH

      So

  16. 32:08 – 34:35

    Why Do People Say Electrical Engineering is Hard?

    1. UH

      remove the noise and join elec.

    2. RG

      I hear you.

    3. UH

      I hear you. Uh, but it's also considered, uh, as, and especially at IIT Madras, it's considered to be a difficult course. Why does it have that reputation?

    4. RG

      Okay, so [chuckles] I mean, uh, but, but I, I was-

    5. UH

      It's true. [chuckles] Is, is it difficult?

    6. RG

      It's no longer true. Let me put it that way. [chuckles]

    7. UH

      Okay.

    8. RG

      So it, it's not difficult. It's interesting. I'll put that. So, so, see, this is a difficult technologies. These are, these are difficult things to learn, but the journey is interesting, and that's more important. It's not that as professors we make it tough for the students to pass the exams or, uh, or, uh, make their life difficult by giving difficult question papers. That's not what we do. See, electrical engineering is unique in that way. It's just not practice. There's a lot of theory associated with electrical engineering. So to do wireless communication, you have to under- have a very deep understanding of probability theory. You have to have an understanding of linear algebra, calculus. So when you understand all these things, then you start appreciating what wireless... So for the first two years of when I did course in wireless, for the first two years, I was reading, doing theory after theory. I, I mean, at some stage I thought, "Where is wireless in this?" It took me another four or five years to understand how wireless and all these things fit together, okay?

    9. UH

      Oh.

    10. RG

      So there's a huge math component associated with electrical engineering, okay? That is the reason why people feel, "Oh, it's tough." But it's not tough, it's actually interesting. These are exactly... So you do a lot of problems for JEE, all those sophisticated, uh, math problems, calculus problems, combinatorial problems, probability problems. Where exactly do you use that?

    11. UH

      Mm.

    12. RG

      You use that in electrical engineering. Every day, I use that.

    13. UH

      Very interesting. So you're saying that before you come to the more visible parts of electrical, uh, there's a lot of math that you have to-

    14. RG

      Rigorous training, that's what I put it as.

    15. UH

      Rigorous training. [chuckles]

    16. RG

      Yes. So, but it's fun.

    17. UH

      Okay, I hear you. I mean, I, I, I do remember reading about information theory and things like that. Like, it was not in my coursework. I did some graph theory, which I was very bad at. So, uh, but I get what you're saying. It's, it's tough, it's fun, it's, uh, it's a little long journey, and then it becomes very exciting very quickly.

    18. RG

      See, one good thing is, when you do all this, getting into any other field is also very easy. Okay, so I... I mean, people who understand, uh, this kind of all these, uh, things that we do in wireless, can easily jump into machine learning, can easily jump into fig- uh, any other, I mean, computer science, for that sake, I mean.

    19. UH

      Students today only want to ask, "Can I go to quant finance after this?"

    20. RG

      Even that, because you learn probability.

    21. UH

      Mm.

    22. RG

      I mean, most of the quant, I mean, uh, probability, statistics, are

  17. 34:35 – 36:00

    Professor Bhaskar’s (Ex-Director of IIT Madras) Legacy in Wireless Communication in India

    1. RG

      something that fundamental to our wireless community. We teach them. [chuckles] So-

    2. UH

      And, uh, I, I, I also, I think, at least when I was a student here, Professor Ashok Junjamala was, uh, teaching, and Professor Bhaskar Ramamurthi was teaching, and there were a lot of... Like, there's a lineage of really great, respected electrical engineering professors. Uh, I mean, the department has this... I, I don't, uh, they must have been professors for you, right?

    3. RG

      Yes, yes. Both of them have taught me. Professor Ashok taught me Communication Systems, and, uh, Pro- Professor Bhaskar also taught me communication. We both took that course.

    4. UH

      Nice.

    5. RG

      I remember that, yeah. Uh, uh, so e-essentially, the work that we are doing now, in some sense, uh, uh, uh, the mindset has been established by Professor Ashok and Professor Bhaskar tw- twenty years back, I mean, maybe three decades back, okay?

    6. UH

      Mm.

    7. RG

      Uh, so they were the first ones actually to build this wireless in local loop. I mean, essentially, uh, they took the bold, uh... They made this bold attempt of building stuff during twenty years back, when things were more restrictive in India, when things were not available in India. They took that thing, and they were so very successful also in that.

    8. UH

      In the late '90s.

    9. RG

      In the late '90s, yeah.

    10. UH

      Mm.

    11. RG

      And, uh, because of that, the group that they have built and the people whom they have nurtured also have this, um, uh, uh, thought of building stuff. It's just not only theory, but building stuff is very, very important right now, and I think the genesis of this came from,

  18. 36:00 – 40:10

    GSM to 2G to 5G

    1. RG

      uh, what, uh, Professor Ashok and Professor Bhaskar established, like, two decades back.

    2. UH

      I remember hearing stories about how WLL as a concept, we were far ahead. Of course, from a technology point of view, uh, WLL was competing with CDMA, right?

    3. RG

      Uh, with GSM.

    4. UH

      With GSM.

    5. RG

      Yeah.

    6. UH

      And GSM is the technology that eventually, uh-

    7. RG

      Became 2G

    8. UH

      ... became really big.

    9. RG

      Yeah.

    10. UH

      Uh, and I guess WLL, I don't know, is it still in use now?

    11. RG

      See, W- see W- at that point, uh, uh, there were, I mean, uh-... uh, WLL and, uh, 2G came roughly around the same time, roughly around the same time. So, uh, we bet on WLL, and WLL was a very, very nice technology. It worked well. It, it got, it scaled up very well. There were some countries that have, have actually deployed WLL also. But from a global perspective, 2G took over at some point in, uh, in the, uh, decade. 2G took over, and because it had a lot of big backing from bigger, deeper pockets, ultimately it won. But nevertheless, what happened was that also proved that as a... I mean, we, uh, we were able to build world-class technologies at that time even, okay? With very, very little resources, we were able to do that, and we were able to show commercial deployments.

    12. UH

      Mm. Of course, the team- teams gain a lot of competency, and they move on.

    13. RG

      Yeah. So these are the-

    14. UH

      And-

    15. RG

      - same people who now join companies and who are building-

    16. UH

      Yeah

    17. RG

      ... 4Gs and 5Gs right now.

    18. UH

      5G, yeah.

    19. RG

      I mean, uh, yeah.

    20. UH

      Yeah, yeah. Actually, in the, in the technology world, there are a lot of these competing technologies that have come at the same time. I remember Blu-ray versus, uh, DV- DVD. I... That was a-

    21. RG

      Yeah, that was a, yeah, yeah

    22. UH

      ... big thing at, in data storage and, um, of course, USB-C versus... So anyway, it's okay. It's fine. I got it. Um, um, I remember that Professor Bhaskar taught... I didn't take any electrical engineering class. I think we had one, EE 110. I have a very hilarious story I'd like to share. Uh, I, I tell it to Professor Bhaskar whenever we meet. He was, of course, the director. Uh, but, um, and, and so he, uh, he- when, uh, I joined, uh, Professor Anand was the director, and Professor Bhaskar was this celebrated professor who doesn't have time for undergrad students. [laughing] Uh, but he did take eight classes for us, a small section of a course, and, uh, the story I have, his classes, everybody came. There were no proxies, no- nobody missed the class, and it was full house and, uh, everybody was super quiet because he's speaking, and he speaks in a very, uh, a, a voice that you want to listen to. [chuckles] Uh, and, um, what happened was, [chuckles] uh, we are sitting in this class, and there's a parliamentary fan which fell [chuckles] on a student. There's utter silence, and everybody's looking at the fan. And then someone, that student had to be carried to the hospital. And Professor Bhaskar looked at the class after that student left, and he said that when he was a child, he used to be worried [chuckles] that a fan would fall on his head. [chuckles] It's a hilarious memory I have of Professor Bhaskar.

    23. RG

      So I mean, uh, B- Professor Bhaskar, I mean, he's one of the, uh, person who understands wireless and communication in general in a very, very deep manner. So I mean, uh, we have been teaching classes with, uh, Professor Bhaskar. So I'll just tell you an anecdote. So we both were setting the question paper, okay? Professor Ba- uh, one of, one of the exam, uh, DSP, actually. So Professor Bhaskar said, uh, "RK Vishwa, let's set up this exam paper a little bit, uh, less, I mean, not so tough this time. Let's make it easy." I said, "Fine, Professor Bhaskar, why don't you give me your questions, uh, okay?" So he sends me a bunch of, uh, questions that, uh, he has... He, he was using, uh, when, when we were students, okay? That's when I realized that, that his, uh, operating point is at a very, very, very high level. [laughing]

    24. UH

      [laughing]

    25. RG

      His easiest questions are about more like ten times more tougher than my tougher questions. [laughing] So I had to recalibrate the whole thing and, uh, see, I, I,... I mean, even now, I learn a lot of things from him. I mean, any conversation I have, it's like he's a mentor. He's a mentor for me, and, uh, so much technically, and, uh, even otherwise also I learn from, uh, him.

    26. UH

      Oh, but this is the kind of, uh, depth you

  19. 40:10 – 42:55

    Professor Ganti’s Insti Story

    1. UH

      need to be able to represent India at a global level and-

    2. RG

      Sure

    3. UH

      ... have that negotiation and build that technology. Amazing. Uh, Professor, um, can I just dial back? You were also a student here-

    4. RG

      Yes

    5. UH

      ... from, I think, uh, '98 to 2002. Is that right?

    6. RG

      '99 to 2004.

    7. UH

      Okay.

    8. RG

      I was dual degree.

    9. UH

      Yeah. Um, how come you joined IIT Madras? What is your story? Uh, how was your student years?

    10. RG

      Okay. So, [chuckles] I mean, while, uh, while-

    11. UH

      Which hostel were you in?

    12. RG

      Okay, first, let's start with that.

    13. UH

      Let's start. [chuckles]

    14. RG

      I'm, I'm in Saraswati hostel. [laughing]

    15. UH

      Okay, I was in Tapti.

    16. RG

      Tapti, okay. So I, I'll start with one more thing. I was the only undergrad who had a single room at that time.

    17. UH

      Is it?

    18. RG

      Yes. So... [laughing]

    19. UH

      Wow!

    20. RG

      [laughing] I was the only lucky guy. I, uh... So, uh, yeah, I, I joined actually Mechanical Engineering, okay? Then I, after, uh, the first semester, I did a branch change to Electrical. [chuckles] Then after four years, I did a, again, a, uh, a change to, from BTech to dual degree. Okay, at that point, I should have bought some land in, uh, [chuckles] Velachery-

    21. UH

      [laughing]

    22. RG

      ... I retired at that point, so, uh, yeah. But, uh, nevertheless, this is, uh, uh, I mean, uh, this was my academic year.

    23. UH

      Sorry, just, just to qualify that comment for everyone watching, uh, at, at the time when we were in campus in 2000 odd, so IIT Madras was in Guindy... Is, is, is still in Guindy, but it was considered like the south edge of the city.

    24. RG

      Exactly, exactly.

    25. UH

      Uh, and now the city has moved twenty kilometers the further south. So the-

    26. RG

      Now this is the center of the city.

    27. UH

      Uh, it's the center of the city, yeah.

    28. RG

      So it was fun. It, it, it, it was a great, uh, place to be. I mean, I had a lot of fun, so I played basketball. I used to play basketball. Even now, I play for staff team.

    29. UH

      Okay.

    30. RG

      Okay, [chuckles] so, uh, uh, the, uh, I, I used to maintain our IIT Madras website at that point.

  20. 42:55 – 46:10

    The Shift From a Theoretical to a Practical View of Systems Studies

    1. UH

      system side. So I think, can you explain that switch from, like, a theoretical view of the subject to, uh-

    2. RG

      So, uh, I, I was just telling you before, uh, uh, communications, communicate, wireless specifically, has a lot of math in it, okay? Uh, so you go on doing a lot of, uh, lots and lots of math. Only when you start building systems, you understand where this whole math fits in, okay? So, uh, uh, so, so what I did was, after my PhD, when I went to, as a postdoc to UT Austin, uh, for a couple of years. During that time, I started tinkering, a little bit tinkering with, uh, something called software-defined radios. These are, essentially, these are hardware where, with, um, where, where there's a lot of programmability in it. So in that, I started implementing, uh, a few of the wireless algorithms that we have designed into those, uh, systems. That is the first time I have started dabbling with, uh, systems, okay? But even after I came back to IIT Madras, primarily, I was doing a lot of theory, still a lot of theory, but the good thing with IIT Madras is that, uh, faculty work, uh, closely a lot, okay? So I was able to get some... I mean, uh, some other, I mean, I was able to get in touch with a few more faculty in the RF domain, and then we started exploring some problems where there's systems aspects of things and theory aspects of things, okay? So I started doing the theory, and he started doing the systems part of it, and then we started learning from each other, okay? It took me about two, two and a half years to get comfortable in building basic systems.

    3. UH

      Hmm.

    4. RG

      Okay. Of course, you should have it. So one thing I want to tell is, actually, once I started building the system, that's when I realized that actually, uh, systems is, uh, there's not too much difference between the two. So if you're good at one, you'll always be good at the other.

    5. UH

      Mm.

    6. RG

      People have this weird notion that I, I'm a guy who will only can do math and not build stuff. I, I mean, nowadays, I strongly disapprove that. It's just that you have not tinkered with stuff. One, the moment you start tinkering with it, then you realize that similar kind of thought processes hold in systems engineering, okay? Systems engineering is where you actually build stuff, understand the practical aspects of things, and from there, design your algorithms.

    7. UH

      Sure. So, for example, if I'm working on a error correction algorithm, that's a theoretical thing, but if I then try to put it into a, a, a 5G test bed or something like that-

    8. RG

      Ah.

    9. UH

      -that's the application of it.

    10. RG

      That's it.

    11. UH

      There's a lot of system thinking that-

    12. RG

      Yes.

    13. UH

      -will help.

    14. RG

      So I can tell you, there's a beautiful example. So, so you have designed an error correction code, okay? For example, you have come up with theory, okay. So when you start implementing it, when you start implementing in real time, you'll have a lot of constraints in terms of how you can implement it. Saying that, for example, your data has to be processed in point two milliseconds, for example.

    15. UH

      Mm.

    16. RG

      Okay? Maybe you have developed a theory which requires a longer time to process but gives you good results. Maybe you have to cut down some parts of the algorithm, or you have to modify some parts of the algorithm, so you meet the latency requirements or the processing requirements, okay? So when... So essentially, when you start doing this, you'll start getting a rich set of problems that you can work even in theory.

    17. UH

      Very interesting. Hmm. Okay, thank you for that. Um, uh, uh, as you were saying it, uh, I, I, we, we do a lot of work with students who come in to campus through this program called Ask IITM, and as you were saying it, I have heard a lot of

  21. 46:10 – 47:00

    Is IIT Madras Too Obsessed With Theoretical Learning?

    1. UH

      parents say that IIT Madras is supposed to be a theoretical campus, uh, and they're quite worried about it, or they're like... But when I talk to professors, when I talk to alumni, they seem to have been doing a lot of building.

    2. RG

      So, see, I, I, I'm-- Actually, that's a completely wrong statement. [chuckles] Okay, pardon me. Actually, IIT Madras is one place where we build so many stuff. No other IIT does anything close to this. In wireless, I'm very confident, but I'm also sure in other fields, we do lots and lots of building, actually, more than any other IIT, okay? Saying that, few, about, about a decade back, people... See, people were more focused on papers, people were focused on, uh-

    3. UH

      Right

    4. RG

      ... research and things. But now, the whole trend is changing. Most of the faculty are involved in some kind of systems work, either with industry or by themselves, okay? And we build a lot of cool

  22. 47:00 – 51:00

    What is The Difference Between EE, EEE & ECE?

    1. RG

      stuff here.

    2. UH

      Nice.

    3. RG

      Well-

    4. UH

      That's why we say, Best Place To Build.

    5. RG

      Yes, it's the best place to build.

    6. UH

      [laughing]

    7. RG

      I totally agree with you. [laughing]

    8. UH

      Professor, uh, because we are talking of Ask IITM, one question that we face from parents also is, what is the difference between EE-

    9. RG

      Yeah

    10. UH

      ... EEE, [chuckles] and ECE?

    11. RG

      Okay. EE, EEE, and ECE.

    12. UH

      EEE and ECE.

    13. RG

      Okay.

    14. UH

      Yeah, because colleges use these different terminologies.

    15. RG

      Okay, so at least in the conventional parallel, uh, uh, in terminology, EE stands for Electrical Engineering, which consists of mainly machines and, uh, transformers, power electronics, and this kind. EEE consists of the same thing, except that you add electronics to it, analog electronics to it.

    16. UH

      Mm.

    17. RG

      ECE is electronics plus communication.

    18. UH

      Mm.

    19. RG

      Okay? So this is a standard dif- differentiation that happens in, uh, most of the o- other institutions, um, other than IIT. So in IITs, we call everything under the one umbrella of electrical engineering. So, for example, in IIT Madras, electrical engineering consists of power electronics and power, power systems. We have VLSI, we have devices which work with solid state devices, we have circuits, we have optical, and we have communication theory. Okay, so all of this together is called electrical engineering in IIT Madras.

    20. UH

      Understood.

    21. RG

      So people can do anything they want.

    22. UH

      They could, I guess, as they grow interest, they would-

    23. RG

      Yeah. And also, one thing is, all the problems we solve are multidisciplinary in nature these days. Now, it's not like one person can solve, okay? So, for example, when we... In my team, I have mechanical engineers. Why? Because we deve- we develop these radios that you put on the top of a tower, okay? There's no cooling there. It has to be cooled by the wind, okay? And you cannot put fans there, because if fans get stuck with mud, uh, you can't go up on the top of tower every day and repair them. So how do you cool a system like a radio, which transmits a lot of heat? Because power amplifiers are there, which are, uh, I mean, which convert a lot of electrical eng- power to thermal power.... So how do you cool that? So a lot of our design has thermal-

    24. UH

      Management.

    25. RG

      -management on it. So it's a very interdisciplinary work.

    26. UH

      Uh, I have this question w-written here. I'm going to ask it as I've written it. [laughing]

    27. RG

      [laughing]

    28. UH

      Does life get too intense? Is there time to do anything else? Uh, I mean, maybe because you are an electrical engineering student also, so maybe you can give, give it from a, from a student and a faculty perspective.

    29. RG

      Okay. So, uh, from a faculty perspective, see, once you do something that you enjoy, even if it's, uh, it takes more time or is it, is it stressful, you can-- you still live with it, and you, I, I mean, at, at the end, you have fun with it, okay? So I, I, I enjoy what I'm doing. Okay, there's a lot of work that, I mean, enormous amount of, uh, work that I do, but I enjoy what I do. Okay? And I, and I see the outcomes of it also, which are very nice, u-useful. Hmm. So if you don't enjoy what you are doing, I mean, it does not matter, even if it's very light, you start hating it.

    30. UH

      Yeah.

  23. 51:00 – 52:40

    How is AI/ML Affecting the World of Electrical Engineering?

    1. RG

      teachers are so tough. Uh, they, they psych you, you know? [chuckles] By the time you go there, you'll be sweating, but nothing of that sort happens. It's actually very cool and nice. Yeah. [chuckles]

    2. UH

      Nice.

    3. RG

      Uh-

    4. UH

      Yeah, I guess if you don't enjoy it, then you'll feel this way.

    5. RG

      Absolutely. Yeah.

    6. UH

      Professor, last question, and question to everyone in the world today: How is AI and ML affecting... How is the-

    7. RG

      See

    8. UH

      ... the evolution in your field?

    9. RG

      Okay, I, so I, I, I told you before, so, uh, yeah, neural networks were the basis of AI ML. Yes?

    10. UH

      Mm.

    11. RG

      So neural networks have been done in electrical engineering from the time when I was writing the projects on neural network, uh, with Professor Chakraborty. I forgot the name of the professor. Chakraborty. Yeah, okay. So, uh, so, so we have... See, the math that you learn in electrical engineering is the basis of most of the AI ML that you do.

    12. UH

      Okay.

    13. RG

      Okay? So that way, electrical engineering people learning AI ML is actually very easy, okay? Saying that, AI ML is there everywhere. Nowadays, you cannot escape not using AI ML. So even in my field of wireless, AI ML is coming into picture. Okay, now we are looking at AI ML to build... You are talking about error correction codes. Earlier, people used to build error correction codes. Okay? So now we are trying to use AI ML to build error correction codes. We are trying AI ML to do decoding. Okay? So people are trying to use AI ML to do a lot of stuff, but saying that the domain knowl-- I mean, see, E- I mean, it's not like generic AI ML. It's not like I can use ChatGPT to do this, okay? How to fine-tune my networks and other things in such a way that it works to my problem. So their domain knowledge is absolutely, absolutely necessary. Okay?

    14. UH

      Okay.

    15. RG

      So it's a combination of domain knowledge and AI ML. I think

  24. 52:40 – 54:53

    Closing Thoughts & Reflections

    1. RG

      that is where things are headed, and, uh, that's a, that's a path to success, let's put it.

    2. UH

      So it's like a tool in a journey that's anyway going in that-

    3. RG

      Yeah, it's, it's a tool.

    4. UH

      Every, every specification in the 5G, uh, list would have a corresponding set of features, and every feature would be probably, uh, better. Maybe AI ML can improve it a better. Is that-

    5. RG

      A little bit, okay? So one thing that is very unique to wireless, okay, maybe I'll take a minute and explain what this is. In wireless, theory came before practice, okay? So I don't know, uh, I mean, there's a gentleman called Shannon.

    6. UH

      Mm.

    7. RG

      Claude Shannon, okay? So he founded this whole information, that you're talking about information theory, yeah?

    8. UH

      Yeah.

    9. RG

      So the information theory gave something very special. It said that it gave you bounds. It said that the maximum capacity of a channel is so and so. So what does this imply, okay, from a practical perspective? It says that if I'm communicating to you, the maximum date rate, data of rate of communication is maybe thirty bits per second per hertz, okay? So that is a goal. Thirty bits is my goal. So from the last fifty years, people have been working, working, working to reach that limit. You cannot beat that limit. It's impossible. Even with AI ML or with anything, you cannot beat the limit. That's the limit, okay? So people have been working to reach this goal. So unlike other fields, if you employ AI ML network here in wireless, you know exactly where you are with respect to the best.

    10. UH

      Hmm.

    11. RG

      So you'll be judged more critically in this field [chuckles] compared to other fields. It's just not... You cannot build something random and say it works. That's not sufficient here. You have to tell me how far are you from what Shannon has given.

    12. UH

      So mechanical equivalent would be Carnot's, uh-

    13. RG

      Carnot cycle

    14. UH

      ... efficiency cycle, yeah.

    15. RG

      If your AI ML beats that, uh, builds an engine that beats it, then that, that... It cannot beat it.

    16. UH

      It cannot beat it.

    17. RG

      It cannot beat it. So it's a question of how close you are.

    18. UH

      Yeah, fair enough.

    19. RG

      Yeah.

    20. UH

      Professor, this was fantastic. I think I have exhausted my questions. I'm really, uh, grateful that you spent the time with us. Um, uh, you've invited us to your lab, so we'll visit your lab later-

    21. RG

      Absolutely

    22. UH

      ... uh, and, uh, take a look at all the things that you have spoken about.

    23. RG

      Sure, sure, absolutely. I'll, I'll show you. I'm happy to show you around.

    24. UH

      Thank you.

    25. RG

      Thank you.

Episode duration: 54:58

Install uListen for AI-powered chat & search across the full episode — Get Full Transcript

Transcript of episode _KQT9GA_pHk

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome