Important Announcement
PubHTML5 Scheduled Server Maintenance on (GMT) Sunday, June 26th, 2:00 am - 8:00 am.
PubHTML5 site will be inoperative during the times indicated!

Home Explore Spectrum_Draft_1 - July 2021

Spectrum_Draft_1 - July 2021

Published by Viraj Mankad, 2021-07-16 07:20:40

Description: Spectrum_Draft_1 - July 2021


Read the Text Version

Preface Learning to learn and learning by Dr Dilip Kothari sharing are two fundamental aspects of the self-learning process. Going forward through this process one can direct his/her learning and can navigate through a wide range of topics, areas of interest, find out good and explicit learning stuff as per his/her career goals, and in turn, sharing it with others help the interested group of people. Sharing the knowledge gives an opportunity to identify and create a peer group for generating and propagating good ideas. In the process, one can get the expertise to pursue it for professional growth. As part of the Department of Electronics Engineering, we not only provide a strong platform to the student and faculty to showcase their latent talent to spread the knowledge and disseminate information but at the same time explore it further by arranging expert lectures by technical specialists, senior alumni members working in different fields and invited talks by entrepreneurs, and arrange a battery of other activities. We are happy to bring out the current issue of the department newsletter. It focuses on the articles from the students, alumni, and faculty on emerging areas and technology, accounts of activities arranged by the students and the department, sharing of student achievements, and planning of forthcoming events. I understand that this is an attempt to produce the reflection of the standing of the department to cater to the needs of the student and faculty alike. Dr Dilip Kumar Kothari Professor, Department of Electronics and Communication Engineering, Institute of Technology, Nirma University.

Article by Faculty 1 Voice of Alumni 3 Expert Lectures 16 Publications by Faculties 17 Publications by Students 18 Events of ECO 20 Events by EC Department 24 Interview Mantra 35 Khoj – Exploring Hidden Talents 39 1. Student Articles 56 2. Poetry 58 3. Photography 59 4. Short Stories 62 5. Paintings 63 The Insta – Techies 66 Students' Achievements 68 Upcoming Events by ECO 70 Credits Section

Article by Faculty Global Broadband Internet - Starlink “Starlink” is a satellite network Dr Manisha Upadhyay that the private spaceflight company SpaceX is developing to provide low- cost internet to remote locations. SpaceX plans to have as many as 42,000 satellites, making the mega- constellation. Due to reusable launch rockets, these low-orbit satellites cost a fraction of the price of typical satellite launches, making it easier and more affordable to launch satellites on such a large scale. SpaceX’s satellite internet system is designed to offer blazing fast speeds up to 1 gigabit per second. Starlink expects to provide satellite internet to the entire planet, including remote locations where internet services are challenging to deliver. The plan involves launching a vast constellation of mass-produced satellites into low- altitude orbit. The satellite network is designed to transmit internet signals to earth-based hubs, delivering superfast connection speeds. The plan for delivering SpaceX satellite internet became a reality after the Federal Communications Commission (FCC) permitted SpaceX to launch satellites. SpaceX plans to expand coverage worldwide in 2021-22.” Additionally, SpaceX founder Elon Musk has declared that Starlink’s internet services will have a “latency below 20 milliseconds. The performance of Starlink far surpasses that of traditional satellite internet and a ground infrastructure global network. Starlink will deliver high-speed broadband internet to locations where access has been unreliable, expensive, or completely unavailable. The estimated cost of the decade-long project to design, build, and deploy the constellation was about US$10 billion. Product development began in 2015, with the first two prototype test-flight satellites launched in February 2018. The second set of test satellites 1

Article by Faculty and the first large deployment of a piece of the constellation occurred in May 2019, when the first 60 operational satellites were launched. SpaceX aims to deploy 1,584 of the 260 kilograms spacecraft to provide near-global service by late 2021 or 2022. In April 2020, SpaceX modified the architecture of the Starlink network. SpaceX submitted an application to the Federal Communications Commission (FCC) proposing to operate more satellites in lower orbits than the FCC previously authorized. The first phase will include 1,584 satellites orbiting at 550 kilometres in planes inclined 53.0°. The first 60 Starlink satellites were successfully launched on May 23, 2019, aboard a SpaceX Falcon 9 rocket. Members of the Starlink team plan to launch up to 60 more per Falcon 9 flight, with launches as often as every two weeks in 2021. Despite the promise of high-speed broadband internet, SpaceX has taken criticism within the astronomical community for its Starlink satellites due to their brightness and potential to disrupt observations of the night sky. Courtesy: 1. 2. 3. 4. provide-global-coverage-around-september-2021-06-22/ 2

Voice of Alumni Deep Learning and Robotics When the term AI is brought up, the Vaibhav Thakor first thing that would come into mind would (2021 Batch Passout) be a humanoid robot who can speak, walk and do tasks like humans. Using rule-based conventional algorithms to program a robot to perform tasks like picking up an object, and finding paths in an environment with many obstacles would not be possible. Fortunately, recent breakthroughs in the field of Deep Learning have opened up some scope of improvement in the field of robotics. Deep Learning is a subfield of Machine Learning, which uses Neural Networks to build a pattern recognition model from the data provided to it as an input. The power of Deep Learning lies in its ability to learn patterns within the data and to generalize these to the actual application. Deep Learning is divided into three paradigms: (1) Supervised Learning (2) Unsupervised Learning and (3) Reinforcement Learning; according to how the data is used for the learning process. In Supervised Learning, the data and the output to be predicted are provided and Neural Networks learn a function that can map from the data to output. Some practical applications are disease detection from medical images, finding objects on the road for self-driving cars, face recognition in mobile devices, predicting stocks, OCR for digital documents, etc. In unsupervised learning, only the data is provided without any output label, and a model is required to learn patterns and distribution of the data. Some good examples are generating human faces, bringing colour to black and white images, super- resolution, generating older and younger versions of the face. The third paradigm Reinforcement Learning differs from the first two in the way the types of data are used for learning. In Reinforcement Learning (RL) the data is not directly provided but instead, the algorithm interacts with the environment and its goal is to maximize 3

Voice of Alumni the rewards by learning to take good actions. A simple example would be to teach the robot to walk on the terrain surface; the rewards can be the speed and distance travelled by the robot. Another paradigm known as self-supervised learning is becoming popular recently, which learns the representation from the data without any labels that can be used with Supervised learning tasks. The breakthroughs of Deep Learning in the field of computer vision can directly be applied to robotics applications, as the algorithm can adapt to new tasks by learning from the data. Convolutional Neural Network (CNN) is one of the most used models for computer vision-based tasks. CNN uses small rectangular receptive fields over the image to find local features and combine them to form the global features in grid form. These features can be lines, curves, different shapes, and more complex objects present in the image data. This simple process is repeated continuously with different receptive field sizes to create Deep CNN models. Deep Learning researchers have already developed powerful CNN models like ResNet, VGG, and InceptionNet pretrained on the Image datasets which can be fine-tuned to support robot-specific applications. Self-driving cars are the best example of applying computer vision to the field of robotics. They use Lidar or RGB-based cameras to collect visual information from the environment. This visual data is then used by the CNN-based models to localize and identify various objects like pedestrians, traffic signs, and other vehicles in the environment. Robotic arms can also be equipped with RGB or infrared cameras to create visual features like the colour, shape, and size of the objects using CNN. These features are then used by another learning algorithm like reinforcement learning to pick up the object using the robotic arm. One specific example is a chess-playing robot where RGB-based cameras can be used to collect images of the chessboard and features are learned using a CNN to keep track of the position of various chess pieces on the board. Drones with computer vision algorithms are also becoming popular nowadays. Here, LIDARs or Infrared based cameras are used with CNNs to create feature maps and another algorithm learns to create trajectory maps to avoid collisions with objects in the environment. Similarly, drones equipped with RGB cameras and CNN- 4

Voice of Alumni based object detection algorithms can be used to track objects from the air. The computer vision-based algorithms are not directly involved with learning the robot’s movements but rather assist it in a hazardous environment. Deep Reinforcement Learning (RL) based algorithms are usually used to learn robotic locomotion and movements. In terms of reinforcement learning, the learner is called an agent who interacts with the environment through actions and receives observations and rewards. The goal of the agent is to learn a function mapping from the known states and rewards to the future rewards and this function is used to decide on actions that can maximize the rewards. Here Neural network-based models are used to learn this function and it guides an agent in the decision-making process. These concepts of Reinforcement Learning with Neural Networks also have been applied to learn gait patterns in quadruped robots which allow them to run like a cheetah or horse. In the context of a quadruped robot where the goal is to learn walking gait pattern: the agent is the quadruped robot, the environment is the surface where the robot will learn to walk, the observations received from the environment can be the actual motor angles measured by the sensor and the torque of the motors, action is the motor angles to be used and rewards can be the walking speed and distance covered by the robot. At each time step, the robot will use motor angles predicted by the model using the observations from the previous time step to maximize the distance travelled. The reason it works so well in practice is that it’s roughly inspired by the working of the human brain. The release of dopamine- like chemicals in the brain acts as a reward for humans to perform various tasks in their day-to-day life. The Deep Reinforcement Learning based algorithms have been shown to work well with robotic applications within a restricted environment only. The important requirement for RL-based learning is to have an exact environment; as the agent learns through interaction with the environment and no other data is provided. Training a real robot in a real environment with RL-based algorithms would be unrealistic as training can take days and as time passes mechanical wear and tear will introduce noise in the data. To solve this issue 5

Voice of Alumni researchers have created many physics engine-based simulators to train robots in the virtual environment which later on can be transferred to the real environment. Some notable examples of these simulation libraries are OpenAI’s Gym, DeepMind Labs, and the CARLA Simulator for autonomous driving. We as humans take the simple skill of picking up various objects for granted, while robots are still having a hard time getting it done right. Even though robots have replaced humans in some industrial applications, the gap between them is visible in those fixed robotic movements instead of the human hand’s natural movements. Researchers have achieved state-of-the-art results using RL- based algorithms in the problem of grasping objects with a robotic arm. Using the visual information from the camera as observation, the robot is trained to pick up objects from one place and put them in another place. The Neural Network based model learns to predict motor angles which maximize the rewards for grasping decided by pressure sensors. One notable example is the robotic arm trained by OpenAI to solve the Rubik’s cube puzzle. In this project, the interesting thing was the robot’s ability to learn similar hand movements as a human without providing any explicit knowledge. The arm was trained in a simulation environment with an accurate physics engine, and the model was transferred to a real-world environment by randomizing the physical parameters. Even self-driving cars also use advanced RL-based algorithms to control the acceleration, brake, and steering angle of the car using CNN’s visual features from the environment. Deep Learning being a powerful magical toolbox has its limitations too. Unlike the visual cortex in the human brain, Neural Networks are not good at creating high-level abstract features which can describe the object independently of its local properties. The chair rotated at different angles with various colours is still perceived as the chair by the human brain due to the understanding of abstract concepts describing the chair object. But if a similar type of chair image is not present in the data used to train the algorithm, then it will fail or give an unpredictable response. Also, algorithms like Reinforcement Learning are still in the research phase and restricted to limited environments only. This type of unpredictable response can cause failure and irreparable damage to the robots. One good example 6

Voice of Alumni of this is how Tesla’s car failed when it saw an advertisement board saying “STOP” and considered it as a traffic sign of stopping. As this edge case was not considered while training the model for detecting traffic signs, the car incorrectly identified the billboard as a stop sign and the car continuously stopped on that road. Despite all these limitations, Deep Learning still holds a great promise for building smart robots which can assist humans in daily life tasks and industrial applications. Common Wireless Bands In Wireless Networking Wireless Networking can be accomplished Aditi Bhatnagar by transmitting radio frequencies over two (2020 Batch Passout) spectrum bands, i.e., 2.4 GHz and 5 GHz. These bands are unlicensed. No permission or license from the FCC (Federal Communications Commission) is required. The bands that require a permit restrict the use and prevent interference. The companies pay a licensing fee to reserve the exclusive right to transmit on the assigned channels (frequency bands) within a particular geographical area. For example, it prevents two FM radio stations in an area from transmitting on 101.5 FM. Unlicensed doesn’t mean uncontrolled usage. A user has to comply with the rules of unlicensed bands, i.e., max transmitter output power = 1 watt and max EIRP = 4 watts. Interference is common in unlicensed bands. Basics of 2.4 GHz This is the most commonly used band in wireless networking and the most crowded band. Interference is a common problem in this band because most applications, such as microwaves, cordless phones, Bluetooth, baby monitors, etc., use this band. It is referred to as the ISM band (Industrial, Scientific, and Medical). It was originally 7

Voice of Alumni designed for use in non-communicative services but became useful in radio communications introduced in 1985. Why is the 2.4 GHz band so popular? The first reason is definitely “Range”. Longer wavelengths of 2.4 GHz penetrate obstructions better than the shorter 5 GHz wavelengths. It provides a more extended range using the same output power as two FM radio stations in an area from transmitting on 101.5 FM. Unlicensed doesn’t mean uncontrolled usage. A user has to comply with the rules of unlicensed bands, i.e., max transmitter output power = 1 watt and max EIRP = 4 watts. Interference is common in unlicensed bands. Basics of 2.4 GHz This is the most commonly used band in wireless networking and the most crowded band. Interference is a common problem in this band because most applications, such as microwaves, cordless phones, Bluetooth, baby monitors, etc., use this band. It is referred to as the ISM band (Industrial, Scientific, and Medical). It was originally designed for use in non-communicative services but became useful in radio communications introduced in 1985. Why is the 2.4 GHz band so popular? The first reason is definitely “Range”. Longer wavelengths of 2.4 GHz penetrate obstructions better than the shorter 5 GHz wavelengths. It provides a more extended range using the same output power as designed for use in non-communicative services but became useful in radio communications introduced in 1985. Why is the 2.4 GHz band so popular? The first reason is definitely “Range”. Longer wavelengths of 2.4 GHz penetrate obstructions better than the shorter 5 GHz wavelengths. It provides a more extended range using the same output power as The 2.4 GHz band utilizes frequencies from 2.412-2.472 GHz. It provides 13 overlapping channels spread equally over the frequencies. These overlapping channels can cause interference between the devices operating on these channels. But there are 3 non-overlapping channels 8

Voice of Alumni (1,6 and 11) that can be used for operation. 2.4 GHz is used by Wi-Fi standards 802.11b, 802.11n, and 802.11g. Channel 14 is banned worldwide except in Japan. Basics of 5 GHz This is another free wireless networking band with two main differentiating characteristics: Higher data transfer speed and lesser range. The 5 GHz band is an unlicensed ISM band (hence free). It offers frequencies in the range of 5.180 – 5.825 GHz. The 5 GHz band is further divided into sub-bands called U-NII (Unlicensed National Information Infrastructure) bands. There are four U-NII bands: i.) U-NII 1 Most widely used band. It has channels 36-48 (increments of 2) ii.) U-NII 2 It has channels 52-64 (increments of 4) iii.) U-NII 2e It has channels 100-140 (increments of 4) iv.) U-NII 3 Not available worldwide. It has channels 149-161 (increments of 4) The most important thing to note is the 5 GHz band offers 23 non- overlapping channels. These non-overlapping channels are a major operational aspect of Wi-Fi Standard 802.11ac’s incredible data speeds (theoretically 7 Gbps). 802.11ac groups, the channels to provide the higher bandwidth. The benefits of a 5 GHz band are reduced interference, less signal congestion, more non-overlapping bands, faster data rates, and fewer disconnects. The limitations of the 5 GHz band are increased costs, decreased range, and lesser DFS (Dynamic Frequency Selection) channels (52-140). These may interfere with the radar systems. If a radar signal is detected on the channel the AP (access point) is operating at, then that AP will be moved to another channel to avoid interference with radar. If this, in turn, limits the available channels for grouping, then the magic of higher data rates will be gone! 9

Voice of Alumni Recent Biggest Upgrade in Wi-Fi – 6GHz Band We have all been dealing with the problem of congestion in Wi- Fi networks with lots of interference, low throughput, slow transmissions, and finally a drop-off. This new spectrum targets this problem and provides spacious channels for communication. It provides 7 super-wide 160 MHz channels to support high throughput connections and lower latency and it is unlicensed, so you don’t have to pay to use this band (as a developer, that’s bliss)! The 6GHz spectrum is the recent revolution in wireless networking providing a lot more bandwidth, a lot less interference, and a lot more space for data transmission. As Kevin Robinson, marketing leader for the Wi-Fi Alliance says – “This is the most monumental decision around Wi-Fi spectrum in its history, in the 20 years we’ve been around,” The brand name is Wi-Fi 6E, implying devices that support this standard will have support for 6GHz band along with all the cutting-edge technologies of Wi-Fi 6 standard (802.11ax). So, definitely, this new spectrum is going to make lives happier as it’s all good until you find the Wi-Fis down! The Big Transformation in \"Healthcare\" by Computer Vision Preface: Jaisheel Chhatrawala (2020 Batch Passout) As we all know that, nowadays the “Medical & Life Science” business has grown up rapidly even during the pandemic, and revenue is increasing till today. So, to look into the larger picture of the key business driver for the Medical and Healthcare segment, computer vision can play a vital role and it can be responsible for leveraging revolutionary changes in the growth of the medical and life science business. 10

Voice of Alumni Implication: Today, computer vision is supporting an increasing number of clinics in better diagnosing their patients, monitoring illness progression, and prescribing appropriate therapies. Not only can technology assist medical personnel in saving time on regular activities, but it also allows them to devote more time to patients. Computer vision's implications for commercial diagnosis, based on activities such as radiography analysis, predictive analysis, and healthcare monitoring, offer many of the opportunities for the healthcare business. Computer vision is a new discipline that aims to teach computers how to see and interpret images in front of them in the same way that humans do. Computer vision does this by utilizing image-processing artificial intelligence algorithms. Physicians who use computer vision technology to examine a variety of health and fitness data will be able to make better medical judgments. Today, such instruments are utilized by healthcare facilities to monitor blood loss during surgery, such as C-section’s procedure Prospect Areas for Implication: A. Timely Detection of Illness: The majority of severe disorders, such as cancer, must be detected in their early stages. Because of its highly honed pattern- recognition capabilities, computer vision allows for the accurate diagnosis of early symptoms. This might help save many lives in the long run by allowing for proper diagnosis. B. Medical Imaging: For decades, computer vision-assisted medical imaging has been a reliable source of information for doctors. It not only creates and analyses photos but also acts as an aid to doctors, assisting them in their interpretation. The software reads and converts 2D scan photos into interactive 3D models, allowing medical providers to have a better idea of a patient's health condition C. Nuclear Medicine: Nuclear analysis is a field of clinical medicine that deals with 11

Voice of Alumni the use of radionuclide medication in diagnosis. Remote radiation therapy techniques using computer vision are sometimes referred to as nuclear medicine. It mostly uses single-photon emission computed tomography and positron emission tomography in diagnosis. Conclusion: To summarise, computer vision has a lot of potential in healthcare since it can increase the quality and level of life all over the world. Doctors frequently use photos and scans to establish diagnoses, which need to be faster and more accurate to provide better healthcare, and computer vision technologies can help. Making Safer Chips for Automobiles Introduction: Nilkanth Pathak As the automobiles of today's world (2012 Batch Passout) get incrementally electrified and autonomous, a great deal of efforts is being made by the semiconductor companies to design and develop chips that carry out a required operation in a safe and predictable manner. This article briefly talks about the standard practices followed in the industry in this regard while highlighting the importance of 'Functional Safety' - the term used for the same. Continuous Evolution of Automobiles: You might have come across recent news articles highlighting the issue of ‘chip shortage’ [1] in the context of cars and the automobile industry in general. This is an interesting situation given the automobile industry is deemed to be a product of mechanical engineering and not electrical/electronics engineering. Given the evolution of technology, the increasing no. of vehicles on the roads of the world and the environmental concerns, “Smarter, Safer, Greener” is today’s mantra of automotive development worldwide. 12

Voice of Alumni An average car of today’s world uses nearly ~1000+ chips to accomplish various functions, including but not limited to Engine management, HUDs, Infotainment, Security, braking management, creature-comfort, vehicle stability, connected-car features, ADAS (Advanced Driver-assistance Systems, e.g. lane-keeping assist, auto- emergency braking), RADAR and LiDAR in case of semi-autonomous or fully-autonomous vehicles, BMS or battery-management systems in case of EVs. The list just keeps getting longer. Some of the chips enable the features that enhance the travel experience e.g. infotainment systems, while some chips play essential roles of electrically managing some of the fundamental features like steering control and vehicle stability management. Semiconductor Chip Development for Automotive Markets: The Semiconductor or VLSI market is rapidly growing to keep up with the above-mentioned demands of the automotive industry and many times the semiconductor companies leading the innovation and development front, just as they do with personal electronics, healthcare and communication industries. Although, the chip development for the automotive market comes with some stringent requirements on quality and failure rates (or FIT rates – Failure in Time rates) on the hardware given the nature of the use of such ICs or chips. An IC, being used to control the steering assembly of a car has human lives riding on it for it to fail . Whereas, an IC being used for home electronics such as an air- conditioner doesn’t generally have life-threatening consequences when it smokes out. The IC/chip development process with the target of automotive applications comes with a certain set of added rules and standards to follow such as the ISO-26262 Road vehicles – Functional Safety standard [2]. Decoding Functional Safety (FuSa): By definition, Functional Safety is Absence of unreasonable risk due to hazards caused by malfunctioning behaviour of E/E systems. Functional Safety or (FuSa) hence means that if the chip malfunctions due to a hardware error or a design error, the hazard caused by it while being 13

Voice of Alumni used in-field shouldn’t result in an unreasonable risk to humans or properties. Let’s take an example. There is a chip designed to control and actuate the motors of the steering assembly of a car. The intended function of the chip is to sense the angle and turn-count based on the steering rotation by the user and actuate the motors accordingly to steer the set of wheels of the car precisely by so much amount. Now, imagine the car travelling at 80 kmph on a highway, while the driver is holding the steering straight. Suddenly, the chip malfunctions and it sends a wrong signal (wrong angle value) to the steering assembly to turn the wheels left (despite the driver holding the steering straight)! It’s needless to say that such mishaps can be life- threatening. This example highlights the need for ‘Functional Safety’. As per the ISO-26262 Functional Safety standard, applications for which semiconductor chips and sub-systems are used can be categorized into different Safety Integrity levels or ASILs (Automotive Safety Integrity Levels) with ASIL-D being the most stringent safety level (aimed at chips used for ‘serious’ automotive applications, like the one we discussed in above steering assembly example) and ASIL-A being a lesser stringent level (applications like infotainment systems). While it’s beyond the scope of this article to detail the differentiated flow of functional-safety mandated chip/ASIC development, we shall look at some of the methods in brief: 1. Redundancy of Hardware: Chips with FuSa requirements often have two redundant paths to perform the same function so that if one of the paths fails due to a fault, the chip still performs the required function. 2. Failure-Mode Effects and Diagnostics Analysis (FMEDA): A review process where the design of the chip is analysed in simulation, in by concept to understand its behaviour with respect to its mission-mode function under various fault conditions. 3. Hardware/Software Diagnostics: Methods in the hardware design of the chips or in software codes to flag or highlight in faults that occurred during the operation. 14

Voice of Alumni 4. FIT rate Analysis: Analysis of the chip production yield data from the foundry. 5. Fault-Injection Testing: Simulations to check the chip’s behaviour under faults. Takeaway: It takes quite a bit of process and quality checks at the semiconductor vendor level, subsystem manufacturer (Tier-1) level and the automotive OEM level, done in maximum mutual understanding of requirements and risks. Next time you see the airbag sign or the check-engine sign on your dashboard, please take a moment to appreciate this evolution in the automotive and semiconductor industry that is trying to keep us safer on the roads. References: 1. industry-110-billion-in-2021.html 2. 15

Expert Lectures Expert Lectures Arranged During January 2021 to June 2021 The Department of Electronics and Communication aims at the overall development of the students. As a tradition, the department invites experts from the industry and academia to deliver lectures to the students. The very purpose of this activity is to encourage the students to interact with the industry. The details of such lectures conducted from January 2021 to June 2021 are as follows: ▪ Dr Dhruv Dave, CPU Verification Engineer, ARM Embedded Technologies Pvt. Ltd., Bengaluru delivered a talk on “Tools, Technology, and Techniques for CPU Verification” to the students of Semester II, M. Tech EC, on January 08, 2021. ▪ Mr Bhavin Patel, Software Engineer II, Cadence, Ahmedabad, delivered a talk on “Universal Verification Methodology” to the students of Semester VI, B. Tech EC, on February 11, 2021. ▪ Mr Manoj Parmar, Program Director – AI Shield at Robert Bosch Engineering and Business Solutions Private Limited (RBEI), Bangalore, delivered a lecture series on “Real-Time Operating System” to the students of Semester VI, B. Tech EC and Semester II, M. Tech EC (ES), during February 23, 2021, to March 11, 2021. ▪ Dr Dhruv Dave, CPU Verification Engineer, ARM Embedded Technologies Pvt. Ltd., Bengaluru delivered “Memory Organization” to the students of Semester VI, B. Tech EC, on February 26, 2021. ▪ Dr Xingquan Zhu, Professor, ECE Department, Florida Atlantic University, USA, delivered a lecture series on “Neural Networks” to the students of Semester VI, B. Tech EC, during March 02, 2021, to March 18, 2021. ▪ Dr Deepak Mishra, Scientist, SAC/ISRO, Ahmedabad, delivered a lecture series on “Information Coding Theory” to the students of Semester VI, during March 02, 2021, to March 23, 2021. 16

Expert Lectures ▪ Mr Jignesh Panchal, ASIC/FPGA Verification Engineer, Arm, Sweden, delivered a talk on “SOC Verification” to the students of Semester VI, B. Tech EC and Semester II, M. Tech EC (VLSI), on March 12, 2021. ▪ Dr Dipankar Nagchoudhuri, Professor, DAIICT, Gandhinagar, delivered a talk on “Analog CMOS Integrated Circuits” to the students of Semester VI, B. Tech EC, on March 26, 2021. ▪ Dr Shruti Oza, Professor, BVU, Pune, delivered a talk on “Research Talk on Flash ADC” to the students of Semester IV, B. Tech EC, on March 30, 2021. ▪ Dr Anna Vizziello, Senior Research Fellow at the University of Pavia, Lombardy, Italy, delivered a talk on “A New Frontier for Telecommunications: Intra-Body Network” to the students of Semester VI and VIII, B. Tech EC, on April 17, 2021. Publication by Faculties ▪ Gupta Jay, Dhaval Pujara, and Jorge Teniente. \"Profiled Horn Antenna with Wideband Capability Targeting Sub-THz Applications.\" Electronics 10.4 (2021): 412. ▪ Sharma Rachna, and Yogesh N. Trivedi. \"Performance analysis of dual-hop underwater visible light communication system with receiver diversity.\" Optical Engineering 60.3 (2021): 035111. ▪ Patel Rohit B., and Dilip Kumar Kothari. \"Design and evaluation of 2.3 Tb/s (23 ch.× 100 Gb/s) multi-carrier WDM optical transmission systems.\" Optical and Quantum Electronics 53.4 (2021): 1-13. ▪ Gajjar Ruchi, et al. \"Real-time detection and identification of plant leaf diseases using convolutional neural networks on an embedded platform.\" The Visual Computer (2021): 1-16. ▪ Pandya Ankur, et al. \"Investigation of Recycling and Impurities Influxes in ADITYA-U Tokamak Plasmas.\" Plasma and Fusion Research 16 (2021): 2402055-2402055. 17

Publications by Students ▪ Riddhi Pandya, Sambandh ▪ Patel Priyam and A. Raina. Pradhan. “Software Validation “Comparison of Machine for Safety System based on Learning Algorithms for IEC61508.” IEEE First Tumor Detection in Breast International Conference on Microwave Imaging.” 2021 Emerging Trends in Industry 11th International Conference 4.0, 2021. on Cloud Computing, Data ▪ Ankit Dhama, Vijay Savani. Science & Engineering “Design and Simulation of (Confluence) (2021): 882-886. Clock and Data Recovery.” ▪ KV Sai Akhilesh, Meet Singh 2021 International Journal of Chauhan, Rajiv Mishra. Research and Analytical “Optimized Speech Recognition Reviews (IJRAR), 2021. and Separation System Using ▪ Rushabh Mehta, Tanish Deep Learning.” 2021 Zaveri, Piyush Bhatasana. International Conference on “Face Recognition based Smart Technology, Artificial Attendance System.” Intelligence and Computer International Journal of Engineering (ICSTAICE), 2021. Research and Analytical ▪ Keshav Kasat, Kunwar Reviews (IJRAR), 2021. Vaibhav. “Wavelet Analysis of ▪ Hitesh Dadlani, Prof. Vaishali Brain Signal Data.” 2021 Dhare. “Design of Reversible International Conference on Technology based Parity Electronics, Communication, Generator.” 2021 International Information Technology and Conference on Management (ICECIM), 2021. Multidisciplinary Research & ▪ Rohan Manish Malhotra, Patel Development (ICMRD-21), Himang Harishbhai. “COVID- 2021. 19 Detection using Chest ▪ Saumya Borwankar, Dhruv Xrays with CNN.” 2021 Shah, Jai Prakash Verma, International Conference on Sudeep Tanwar. “Automatic Electronics, Communication, Speech Emotion Recognition Information Technology and using Cochleagram features.” Management (ICECIM), 2021. 2021 4th International ▪ Meet Naik, Pahal Chhapiya. Conference on Recent “Secure End-to-End VoLTE Innovations in Computing, based on Ethereum 2021. blockchain.” 2021 International 18

Publications by Students ▪ Conference on Substantial ▪ using basic algorithm Development in the field of processes” International Engineering, Management and Conference on Humanities, 2021. Multidisciplinary Research & ▪ Anju Katarmal, Akash Development (ICMRD-21), Mecwan, Manish Patel “RF March 2021. CMOS Double Balanced ▪ Vaishali Dhare, Usha Mehta, Gilbert Cell Mixer for 5G Dhruvil Koshti “Cell Application”, 3rd IEEE Displacement Defect Analysis International Conference on of QCA Majority Voter”, Signal Processing and International Journal of Communication (ICSPC’21), Nanoscience and Arunya Institute of Nanotechnology (IJNN). Technology and Sciences, ▪ Rhea Biji, Vijay Savani Coimbatore, India, May 2021. \"Performance Analysis of ▪ Jinalee Jayeshkumar Raval, Vedic Mathematics Algorithms Dr. Ruchi Gajjar. “Real-time on Re-configurable Hardware Sign Language Recognition Platform\", Sādhanā, Springer Using Computer Vision” 3rd Journal. IEEE International Conference ▪ Aneesh Khanusiya, Dr Vijay on Signal Processing and Savani \"Retinopathy based Communication (ICSPC’21), Diabetes Recognition using May 2021. Convolution Neural Network\", ▪ Rishabh Buddhdev, Dilip Journal International Journal Kothari “Smart Mirror - A of Scientific & Engineering New Way to Life”, IETE Research. Mumbai April 2021. ▪ Bhasha Shah, Dr Usha Mehta ▪ Ankit Kumar Yadav, Dr Usha “Development of Static Mehta “Design and Simulation Timing Analysis Tool in of I2C Protocol” International Perl”, 5th International Conference on Conference on Recent Trends Multidisciplinary Research & on Electronics, Information, Development (ICMRD-21), Communication & Technology March 2021. (RTEICT-2020), November ▪ Parth Thobhani, Dr Usha 2020. Mehta “Development of ATPG 19

Events by ECO Abhivyakti – A Declamation Competition The Electronics and Communication Students’ Organisation organized an online Declamation competition named ‘Abhivyakti’ on the occasion of National Science Day on 28th February 2021, from 1:00 pm to 3:00 pm via Cisco WebEx. As part of this competition, participants were to deliver a speech through oration, a presentation, or a video on the topic “India’s Contribution to the World of Science and Technology”. A total of 11 students participated from the first and second years of the EC engineering department. More than 60 students from the first, second, and third years, along with esteemed faculty members, attended the event. Dr Ankur Pandya and Dr Chetna Chauhan, Assistant Professors (Physics) from the Institute of Technology, were invited to be the judges for the event. The event began with all the participants delivering their elocution, one after the other. Topics like the Indian Space Research Organization, India’s semiconductor industry, the contribution to fibre optics, and the life of the scientist J. C. Bose. The presentations were backed by solid research and a deep understanding of each topic. With the completion of all speeches, the board presented a documentary on the lives of the great Indian scientists, Sir C. V. Raman and Dr Vikram Sarabhai. Finally, the judges gave their valuable remarks on all the presentations and announced the results. Palak Kapuriya (20BEC076) was declared as the winner and Upamanyu Dixit (19BEC140) and Sneha Jain (20BEC120) were declared as the first and second runners’ up respectively. 20

Events by ECO Code Cook – A Coding Event The Electronics and Communication Students’ Organization organized an online Coding competition named ‘Code Cook’ on 30th March 2021, from 8:00 pm to 10:00 pm on an online platform CodeForces and Cisco WebEx. This competition was designed exclusively for First-year students to promote the importance of coding among young minds and instill in the habit of developing strong analytical skills and logical problem- solving skills for further application in programming. The problems were designed based on the day-to-day college activities like interaction with seniors, exams, grading, etc. More than 50 students of the first year from various departments participated. Second-year ECO members helped in organizing the event. Dhruv Patel (20BCE201), Smitraj Rana (20BCE281) and Utsav Vasoya (20BCE312) stood first, second and third respectively. To encourage students, prizes were distributed to all the winners. 21

Events by ECO STTP on Research Methodology – In association with IEEE ITNU The Electronics and Communication Students’ Organisation in association with IEEE ITNU, arranged STTP on Research Methodology. This training program held place under the guidance of Dr Dhaval Pujara, Professor and Head of Electronics and Communication Department, Institute of Technology, Nirma University. This webinar was held on continuous two weekends in March 2021, with the first session on 06th March 2021. In the first week, students were introduced to research methods and processes and literature review techniques and sources. In the second week, students were taught how to formulate a research problem and research ethics. Also, they were briefed about the documentation process. In between various fun and networking session took place. More than 130 students participated in this training program. 22

Events by ECO Webinar on Embedded Systems and Internet of Things The Electronics and Communication Students’ Organisation organized a live webinar on the topic of “Embedded Systems and the Internet of Things” on 11th June 2021, from 11:00 am to 12:00 pm via Cisco WebEx. The speaker of the session was Mr Kushal Vishnu Nesarkar, Embedded Systems Engineer at EdGate Technologies. As part of this webinar, the participants were explained the importance of Embedded Systems, their application-based study on microcontrollers, and utility of IoT with respect to Industry 4.0. Mr Kushal explained different applications of Embedded Systems and presented practical applications of IoT. From this webinar, one can get a clear idea about becoming a successful embedded engineer. The participants were also given useful information on various open-source tools for embedded system development. A total of more than 90 students from the second and third year of the EC engineering department attended the webinar. 23

Events by EC Department GUJCOST sponsored One-Week National Level STTP on “System on Chip Design: Trends and Tools” One-Week National Level Short Term Training Program (STTP) on “System on Chip Design: Trends and Tools” organised during January 4-8, 2021 from 9.15 am to 4.15 pm. This STTP was sponsored by Gujarat Council on Science and Technology (GUJCOST) and was conducted through online mode using the WebEx platform. The webinar was coordinated by Dr Manish I. Patel and Dr Nagendra P. Gajjar from Electronics and Communication Engineering Department, IT-NU. The inaugural session was chaired by the Director, IT-NU and the HoD, Electronics and Communication Engineering Department, IT-NU. The coordinators, Dr Nagendra Gajjar and Dr Manish I. Patel along with the following experts conducted sessions in the STTP: ● Mr Nandan Tripathi, NVIDIA, Bangalore ● Dr Amit Bhatt, DAIICT, Gandhinagar ● Mr Prakash Ganesh, CoreEL Technology, Bangalore ● Mr Kishore Siddani, Mathworks ● Mr TVS Ram, SAC-ISRO, Ahmedabad ● Mr Vidhu Mouli, Xilinx, Hyderabad ● Mr Pinal Patel, eInfochips, Ahmedabad ● Mr Kaushal Modi, eInfochips, Ahmedabad ● Mrs Amrutha Nair, eInfochips, Ahmedabad ● Dr Mithilesh Jha, Infineon, Bangalore ● Mr Himanshu Patel, SAC-ISRO, Ahmedabad The objective of the STTP was to provide a comprehensive overview of the design criteria, methodology, skills and knowledge which are needed for an SoC designer. Along with the theory sessions, hands-on sessions and recent trends in SoC designing and the related tools was the key focus of the STTP. 24

Events by EC Department The registration of the workshop at a nominal charge was open for engineering students of UG, PG and PhD programs, faculty members and industry personnel across the nation. Total fifty-nine participants attended the STTP. The feedback received from the participants was overwhelming. The content covered in the STTP were as follows: 1. SoC Processors and Architecture 2. IPs for SoC 3. Tools like Xilinx Vivado, IP Integrator, Vitis, Vivado HLS 4. SoC Design Flow 5. Design and Prototype FPGA/SoC using MATLAB and Simulink 6. Recent Trends in SoC Platforms on FPGA 7. SoC Testing and Verification 8. SoC Applications in Automobile and Indian Space Program Certificates were issued to all the participants who attended the sessions. 25

Events by EC Department National Conference on “Advancement in Communication, Electronics, Computers and Automation Technology” The Department of Electronics and Communication Engineering, Institute of Technology, Nirma University has organised a national conference on, “Advancement in Communication, Electronics, Computers, and Automation Technology” during February 12–13, 2021. The conference was exclusively for the students of UG and PG courses from the disciplines of Electronics & Communication, Instrumentation and Control, Computer Science and Engineering and Electrical Engineering. The conference was sponsored by the GUJCOST, Department of Science and Technology, Government of Gujarat. The IETE, Gujarat and IEEE, Antenna Section, Gujarat were the technical collaborators of the conference. The alumni members of the department were involved in the various organizing aspects of the conference. There was an Advisory Committee composed of several alumni members across the globe. Some of the alumni have provided partial support for the conference. Alumni members acted as reviewers for the conference as well as for the scrutiny of the papers. A few alumni members served as experts for the technical sessions. There were 62 papers from more than 90 authors that were received across the country. After the similarity check and twofold reviews, 36 papers were selected for the oral presentation. Total 54 authors have registered for the paper presentation and 236 students have registered as participants for the conference. The selected papers were divided into four technical sessions namely: (i) Signal and Image Processing, (ii) Computer Vision, AI and Machine Learning, (iii) VLSI and Embedded Systems iv) Communication Systems and 26

Events by EC Department Standards. All the sessions had plenary talks followed by oral presentation sessions. Looking at the current pandemic situation, the entire conference was organised in an online mode. To save time and to avoid technical issues, video-recorded presentations were collected from all the authors well in advance. The conference started at 1:00 pm on February 12, 2021, on the WebEx online platform. Shri Rajeev Jyoti, Dy Director, MRAS, SAC, ISRO, Ahmedabad inaugurated the conference as the Chief Guest. Dr R N Patel, Director, Institute of Technology, Nirma University presided over the function. In the beginning, Dr Dhaval Pujara, HoD – EC, welcomed the gathering. Dr Akash Mecwan, Coordinator, ACECAT, introduced the audience to the conference and the proceedings of the next two days. Mr Rajeev Jyoti advised the students to take maximum benefit of the conference and also appraised the efforts of the department towards the betterment of students. Dr Rajesh Patel also appreciated the efforts of the department. Dr Patel also unveiled the first issue of the second volume of department magazine-Spectrum. The first technical session on, ‘Signal and Image Processing’ started at 2:00 pm on February 12, 2021. Dr Chirag Paunwala, Dean, Research SCET, Surat was the expert of the session. Dr Paunwala delivered a plenary talk on, “Non-Invasive COVID 19 Detection Technique”. In the oral presentation, Dr Paunwala was accompanied by Mr Vishal Mehta, Head, Image Processing Section, Jekson Vision, Ahmedabad (alumni member of EC), and Dr N P Gajjar, Professor, EC, IT, NU as judges of the oral presentation session. A total of 8 papers were presented during the session. The second session on February 12, 2021, was on, ‘Computer Vision, AI and Machine Learning’. Dr Karunakar Kotegar, Head, Computer Department, Manipal University was the expert for the session. Dr Karunakar discussed “Computer Vision using Deep Learning” in the plenary session. The plenary session was followed by an oral presentation session. Dr Kotegar and Dr Samarth Brahmbhatt, Post-Doctoral Fellow, Intel Intelligence lab and alumni of EC judged the session. Mr Jaimin Patel, Manager, Palo Alto, USA other alumni of EC Department joined them for the session as a judge. A total of 9 papers were presented in the session. 27

Events by EC Department The third session was organised on February 13, 2021. The session was on, ‘VLSI and Embedded Systems’. Alumni member of EC Department – Dr Gaurang Upasani, Commuter Architect, Google, USA was the expert for the plenary session. He discussed the “Computer Architecture Challenges and Trends” in his session. The session was followed by the third oral presentation session. Yet another alumnus of the department Mr Nilav Choksi, Principal Engineer, Volansys, Ahmedabad served as a judge with Dr Dilip Kothari, Professor, EC, IT, NU. A total of 10 papers were presented in the session. The final session on, ‘Communication Systems and Standard’ started at 3:00 pm on February 13, 2021, with the plenary talk on, “Index Modulation for the Next Generation Wireless Communication” by Dr Yash Vasavada, Associate Professor, DAIICT. Dr Yogesh Trivedi, Professor, EC and alumnus of the EC department, Mr Dhaval Upadhyay, Scientist, SAC – ISRO, Ahmedabad judged the oral presentation session with Dr Vasavada. A total of 9 papers were presented in the session. Mr Ramji Makwana, Chairman, IETE, Ahmedabad was a Guest of Honour in the valedictory function of the conference on February 13, 2021. He encouraged students to be innovative in life and to 28

Events by EC Department participate in such activities in the future. He also briefed the students about the IETE and the activities carried out by their centre. Three prizes per session were awarded for the three best papers. The first prize of Rs 5000/-, second prize of Rs 3000/- and the third prize of Rs 750/- were awarded in each technical session. The Electronics and Communication Students’ Organization (ECO) of the Department was also an active part of the entire organisation. The student members of the ECO engaged the students during the breaks in the event and also awarded prizes for on-the-spot events. Total prizes worth Rs 37000/- were awarded to the winners in the conference. Student participants appreciated the organisation of the conference and requested to continue with the future versions of the event on a large scale. GUJCOST Sponsored One Day Online National Workshop on “Linux Shell Scripting” A One-Day National Workshop on “Linux Shell Scripting” was organised on March 20, 2021, through the WebEx platform from 9:00 am to 4:30 pm. This workshop was partially funded by the Gujarat Council on Science and Technology (GUJCOST). The registration of the webinar was open for UG, PG, and PhD program students, faculty members, and industry personnel across the nation. More than 75 participants from different regions have participated in this National level workshop. The workshop was coordinated by Dr Vaishali Dhare and Prof Dipesh Panchal from the Department of Electronics and Communication Engineering, Institute of Technology, Nirma University. The inaugural session was chaired by the HoD, Electronics and Communication Engineering Department, ITNU. The prime objective of this workshop was to provide an introduction and have hands-on experience of the Linux basics, terminologies, commands, and shell scripting through interaction with academicians and industry experts. Keeping these aims in mind, the sessions were conducted through Linux installed machine. 1. Introduction to Linux and its Installation 2. Linux Commands 3. Linux Shell Scripting 29

Events by EC Department 4. Application of Linux Shell Scripting 5. Demonstration of Real-Time Application Dr Vaishali Dhare conducted a session on Linux Commands, Installation of Linux with different methods, and execution of commands through an online platform. The pre and post-lunch sessions were conducted by Mr Rahul Mehta, Physical Design Engineer, eInfochips, Ahmedabad. The advanced Linux commands were demonstrated to the students, which are commonly used in the industry. The various possible shell scripts were demonstrated. Also, the industry applications of shell scripts were explained and demonstrated. The Sessions were conducted through an online platform using WebEx. Formal and informal feedback was taken at the end of the workshop and received good responses from the participants. A Certificate is issued to all participants who had registered and attended the workshop. 30

Events by EC Department GUJCOST sponsored Two-Day National Webinar on “Etiquettes of Placement, Higher Studies & Research (EoPHR)” A Two-Day National Webinar on “Etiquettes of Placement, Higher Studies & Research (EoPHR)” was organized during April 16-17, 2021 through the WebEx platform. This webinar was partially funded by the Gujarat Council on Science and Technology (GUJCOST). All the sessions in this webinar were conducted by renowned experts from Industry and Academic Institutes. The registration of the webinar was open for engineering students of the UG and PG programs. More than 104 participants from different regions and branches of engineering have participated in this National level seminar. The webinar was coordinated by Dr Vijay Savani, Dr Bhupendra Fataniya, and Prof Rutul Patel from Electronics and Communication Engineering Department, IT-NU. The Chief Guest of the inaugural function was Dr Rajesh N. Patel, Director, Institute of Technology, Nirma University and chaired by the HoD, Electronics and Communication Engineering Department, IT-NU. The webinar was planned for two from 09:00 am to 05:00 pm with expert sessions. 31

Events by EC Department The following experts were the resource persons who conducted sessions in the webinar: Name of Speaker Affiliation ● Dr Dhaval Pujara Professor and Head – ECE, Inst. of Technology, Nirma University ● Mr Sachin Sehgal CEO at Fincasys Heritage Private Limited & First Placement ● Ms Suhasini Assistant Manager (USA) at IDP Education Ranganathan ● Dr Khushi Vyas Research Associate, The Hamlyn Centre for Robotic Surgery, ● Mr Vivek Imperial College, London Randeria Founder and CHO - HUNCH CONSULTANTS ● Dr Nimrat Singh Founder Tangram ● Mr Rohan Garg Founder - ICE Gate Institute, & Founder - ● Mr Dhaval Amin Teacher & Dy. Section Officer - Gujarat Secretariat The objective of the webinar was to provide insight on the following topics like Resume Writing / Build your Bio-Data, how to Face an Interview, Higher Studies in Abroad & Its relative Competitive Examinations (GRE, IELTS, etc.), Academic Research as a career path, Corporate Etiquettes, Emotional Intelligence, Higher Studies in India & Its relative Competitive Examinations (GATE, CAT, etc.), and Civil Services Examinations. The key focus of this webinar was to familiarize the participants with aspects of Placement, Higher Studies, Research, and Personal Behaviour through interaction with renowned academicians, professional speakers, private consultants, and experts who are working on it. 32

Events by EC Department The content covered in the webinar was as follows: 1. Building resume: ● Difference between resume, bio-data, and CV, Email etiquettes, Do’s and don'ts while preparing a resume, or drafting an email 2. Facing Interviews: ● Desired skills from the candidate (like Intelligence, leadership, competence, courage, inner strength, connects, etc.), Tactics while facing interviews, Questions to be answered before an interview 3. Higher studies abroad and Competitive examination: ● 7-steps to study abroad, Opportunities at different universities, Q&A 4. Academic research as a career path): ● What is research, why should one pursue research, Research facts, myths and misconceptions, Research components: Academic writing, plagiarism 5. Corporate Etiquettes: ● Shaking hands, Exchange business cards, Expectations of corporate, College vs corporate, Workplace etiquettes 6. Emotional Intelligence: ● Emotions (Like sad, happiness, anger, etc.), Emotional Quotient Vs Intelligence Quotient, Emotional Reaction, The secret for a long and healthy life 7. Higher Studies in India & Its relative Competitive Examinations (GATE, CAT, etc.): ● Orthodox career paths, Higher education after engineering, Choosing an option of higher study MBA or GATE/Research, Eligibility and options for higher studies in India 8. Civil Services Examinations: ● Job opportunities after graduation (in central and state government), Competitive exams after graduation for a government job, IBPS for bank jobs, etc. 33

Events by EC Department The Sessions were conducted through an online platform using WebEx. Formal and informal feedback was taken at the end of the webinar and the participants gave a very good response. An E-Certificate is issued to all participants who had registered and attended all the sessions. 34

Interview Mantra Goldman Sachs | Summer Analyst Interview Experience Hi Everyone, Firstly, I would like to thank the department and ECO board for giving me an opportunity to share my interview experience for a Summer Analyst position at Goldman Sachs. I applied through the company’s career Purvansh Shah portal for the role of a Summer Analyst (18BEC099) under the Engineering Campus Hiring Program in February 2021. Around 1.5 lakh students from all over India (2021 and 2022 batch) had registered. Everyone who applied was eligible for the first round. Aptitude Test The test was conducted on HackerRank and was 1 hour 45 minutes long. It had a total of 7 sections, consisting of 68 questions distributed such: 1. Numerical Computation: 8 MCQs 2. Numerical Reasoning: 12 MCQs 3. Comprehension: 10 MCQs 4. Abstract Reasoning: 12 MCQs 5. Diagrammatic Reasoning: 12 MCQs 6. Logical Reasoning: 12 MCQs 7. Subjective: 2 Each question in the MCQs carried +5 for a correct answer and -2 for an incorrect answer. The subjective questions carried 10 marks each. Around 3000 students were selected for the next round. 35

Interview Mantra Technical Assessment 1. The coding section had 2 programming questions and the duration was 30 mins | 1 LeetCode Easy and 1 LeetCode Medium Level question. 2. CS MCQ section had 7 MCQs and its duration was 25 mins | OOPs, Operating Systems, Database Management Systems and DSA | +5/-2 Marking Scheme. 3. Problem Solving MCQ section had 8 MCQs and its duration was 20 mins | Moderately difficult Mathematics questions on Linear Algebra, Probability, Statistics, P&C, and Trigonometry| +5/-2 Marking Scheme. 4. The advanced section had 1 programming question and its duration was 45 mins | I had a question on the graphs data structure. 5. The subjective section had 2 questions and its duration was 15 mins | Common HR-based questions. Considering the current CGPA as a separate section, we had to clear the cut-off in at least three sections (out of 5 + 1 sections) to be shortlisted. The cut-off for each section was normalized and it depended on the college’s CGPA pattern and the specific test-scoring pattern. We could jump between sections till the time for that section was exhausted. Around 200 students were selected for the virtual interviews. Virtual Interviews I had 3 virtual interviews, all on the same day. We were informed that all of the interviews are elimination rounds. There can be any number of interviews between 1 and 3. 1ts Interview – Technical Round This was purely a coding-based round taken on Hackerrank Codepair. The interviewer was considerably amicable and happened to know Nirma University already, which boosted my confidence. 36

Interview Mantra The interview started with a brief “Tell me something about yourself”. Post this, the interviewer asked a question. The question was pretty short but rightly framed (LeetCode Hard). I asked him a lot of clarifying questions to make sure that I understood the question correctly. I began with the discussion of a brute force solution, although I thought I could improve upon it. After thinking a little more, I came up with the most optimized solution. We had some discussion over the possible corner cases and finally, we agreed that this could be the best possible solution. I later offered to quickly code the solution but he asked me not to. 2nd Interview – Technical Round I felt that this interview was a bar-raiser round. This was again conducted on the Hackerrank Codepair platform. After a quick introduction, the interviewer asked me a very interesting but open- ended question. In the initial 60% of the time, I just asked questions and understood the exact problem and discussed a lot of ‘what-ifs’ with the interviewer. I ensured that I kept speaking my mind out and wrote a pseudo-code alongside discussing the solution. Finally, I drafted an optimized solution and the interviewer seemed happy with it. I was also grilled on topics like Data Analysis, DSA, Mathematics, and Database Design because I had mentioned them in my resume. 3rd Interview – Behavioural Round This was a resume-based round. Initially, I was asked to give a walk-through of my resume. The interviewer was curious about my projects and went in-depth for some of my recent ones. Following this, he asked me 4 behavioural questions. They were pretty simple and straightforward. I was prepared for them and thus, I could drive through them easily. In the end, the interviewer asked me if I had any questions for him and he replied to my question very patiently. All of the 3 interviews were taken by Vice Presidents who made sure that I was comfortable before starting the actual interview. My takeaways from the interviews (These are based on personal opinion and might not apply to everyone): 37

Interview Mantra ● Asking your friend/a senior to take your mock interview a day before the actual interview can help you to find flaws in yourself and then you can work on the feedback to improve. ● Learning only a handful of the tech stack, but in-depth is better than learning a lot of tech but not being confident in any. ● Mention only those tech stack/projects in your resume that you are confident in. ● If there is something you don’t know about, you can admit it to the interviewer honestly. The interviewer mostly will take it positively. ● Understanding the question correctly before starting to think of its possible solution is very important. ● Give as many mock interviews as you can before the actual interview. This helps you overcome your anxiety during the interviews. ● Keep communicating your approach during the interviews. It might not be the correct one but in this way, the interviewer would help you think in the right direction. ● Be prepared for common behavioural questions before the interview. I hope you find this helpful. 38

Khoj - Student Articles Making Choices Neel Joshi (18BEC045) In life, we often have to decision is difficult. Two make a decision from available choices. At every stage and on researchers from Columbia and every day. It may be a small one like selecting a biscuit pack Stanford university carried out while you are shopping at the mall or a big one like deciding an experiment. Research your college. Sometimes the process of making a decision is assistances, dressed as easy but sometimes it is hard. shopkeepers, invited people to purchase Jam. The experiment was carried out twice, with a different number of Jam varieties: 6 and 24. Customers were allowed to taste as many as they want. Results were quite different from one’s intuition. People were more likely to test and get attracted towards Jam when a number of choices were more. But they were more likely to purchase one when there were only 6 choices. Research shows that while The concept is not new. overwhelmed by choices we feel It’s known as FOMO - Fear Of that the process of making a Missing Out, if you are choosing between two (or more) 39

Khoj - Student Articles experiences or events. For because it feels like I am example, when you decide not to supposed to do it? Is it because go to a friend’s birthday party) my friend is doing it, or is it for something more important to because it gives me money you, you feel left out. The first (/certificate or any kind of two principles of Economics state reward)? If the answer is yes to is as ‘making a choice comes these questions, I avoid doing with its trade-offs and an such things and you should as opportunity cost’. When you well. It may keep you happy for decide to go for a certain thing, some time but friends or money you miss other ones as you have can't help you to be engaged in a limited amount of time. any activity, which you don't like, for a longer period. So, what should be our approach in such cases? Well, I found 2. How much time it will take? decision making quite difficult. So, I've come with an algorithm Secondly, I try to estimate the that (I think) works for me. It time for the choice I am may help you as well! making. But don't waste too much time in determining time. There are few questions I Just remember any similar event ask myself while deciding in past and find out how much whether to go for something or time you spent on that. not. 1. Why I want to do it? 3. Do I have that much time? The very first question is why I want to do something. Is it just The last question I ask myself is do I have that much amount of time? While answering this question just don't consider your present self but take care of your future self, too. For example, let's say the vacations are ongoing and you decide to learn a new instrument or to take an interesting MOOC course. Now, this type of thing requires a large investment of time. You may enjoy it now but 40

Khoj - Student Articles not when you will be busy on a The last thing I want to regular schedule. So, what you mention is, once you got the can do is you can think of answer for all the questions and something small. Maybe instead if it’s ‘yes, I want to do it then of being committed to writing a give your 100%. Make a priority blog, you can write an article list if you don’t have one and for the department newsletter! fit it wherever it fits in it. An Employment Scam “Money!!! I want lots of Money”, “I want Rishita Chaliawala a job that pays so much money”, “When I (19BEC113) earn money, I will spend it on…..”, etc. are arbitrary thoughts of a college-going student in our country. I bet everyone at least once in their lifetime has Googled, “How to earn money by doing some simple tasks”. Why? Why such kind of thoughts? Simple, because college students are seeing their teenage days fly and opening the gates of adulthood, where they think to become independent, free-spirit and strong enough to not to ask their parents for their expenses. Sometimes they feel like they have become a burden to their family which is not their fault, it has been in our history for parents to make their children aware of them being a burden, to push them into the industrial world. With fluctuating hormones and pressure, sometimes they tend to make bad decisions and then pay for its consequences. Being a teenager, one wants to take their own decision and responsibilities but all they lack is experience. For an engineering student, summer breaks are usually for internships because they are an advantage in one’s resume. But the scenario in the current COVID situation is different. Those who are getting and allowed for internships are doing well but those who 41

Khoj - Student Articles aren’t able to, are feeling inferior and depressed because of peer pressure. These students are now focusing their ray of hope on google work from home companies or freelancing sites. RED ALERT! Seriously be very careful in this option. Data states that 70% of all cybercriminals target students by luring them big pay amounts. This awareness is a must for all the college students of this era. These sites will be showcasing some thousands of INR just for a simple task such as form filling, text to handwriting, etc. They lure students like honey to bees. After one fill in all the details mentioned, they WhatsApp message the brief details, terms and conditions and deadline for work submission. There is still a wall between the task and student and that is of registration fee or security fee. This is some guaranteed money from the student side which they promise to refund with payment. It is the same as our methodical reference of a promise by King Dashratha to his Queen Kekei in Ramayana. The payment displayed would be 10 times the security fee, so obviously one would invest. So many students fell into this rat trap and pay security fees and get the task, but when the time comes to pay for the submitted work, these people falsely accuse the students of cheating or states their work insufficient. Thus, gulping down their security fee along with the legit payment that the student deserved. Sometimes, if a student is unable to complete the work within the time, then these fraudsters threaten to sue the former in court and as a penalty have to pay around 50k INR. These are all bullshit ways to thug money from young students. So, do not believe all those online people and still, someone unknowingly fell into this, then there is always an option to file a complaint in the cyber-crime department. So, this summer, I plead all my fellow peers to please don’t chase this work from home jobs because if earning money was as simple as these cozeners depict then unemployment or poverty would not have become a major concern in our country. 42

Khoj - Student Articles CUDA – Parallel Programming for GPUs CUDA is an acronym for Compute Pallav Rathod Unified Device Architecture. CUDA is a (19BEC106) parallel computing platform and Application Programming Interface (APIs) model created by NVIDIA. It enables the use of Graphics Processing Unit (GPUs) for general purpose processing – an approach termed as GPGPU (i.e. General – Purpose computing on Graphics Processing Units). The CUDA platform is a software layer that gives the direct access to GPU’s instruction set and parallel computational elements, for the execution of compute kernels. CUDA Cores are parallel processors like CPUs processors that are quad – core or hexa – core device. NVIDIA GPUs have several hundred or thousands of cores which are responsible for all types of processing of data that are fed into and out of the GPUs. In GPU-accelerated applications, the sequential part of the workload runs on the CPU – which is optimized for single-threaded performance – while the compute intensive portion of the application runs on thousands of GPU cores in parallel. With CUDA, computing applications can be dramatically speed up by using the power of GPUs. CUDA accelerates applications across wide range of domains from Image Processing, to Deep- Learning, Numerical Analytics and Computational Science like performing game graphics calculations that are resolved visually to the end – user like rendering scenery in – game, drawing character models, shading and lighting within the game environment, etc. 43

Khoj - Student Articles CUDA is a specialized programming language that uses the GPUs in a specific way to perform tasks with greater performance. Each GPU contains hundreds to thousands of CUDA cores. The CUDA architecture changes that makes cross – generation comparisons non – linear but generally more CUDA cores means more raw computing power from the GPUs. The jump from the Kepler to Maxwell architecture saw around 40% efficiency gain in CUDA core processing ability, which shows the difficulty in linearity for drawing comparisons between both architectures. The CUDA performance is designed to work with popular programming languages like C, C++, Fortran, Python and MATLAB and can express the parallelism by using extensions in the form of a basic keywords. This accessibility makes it easier to use GPU resources, in contrast to prior APIs like Direct3D and OpenGL, which required advanced skills in graphics processing. CUDA also supports programming frameworks like OpenACC, OpenCL and OpenMP, and HIP by compiling codes to CUDA. Autonomous RC vehicle and Electric vehicle Future An autonomous remote-controlled car is Harsh Panara a vehicle that can be given directions and (19BEC083) the vehicle will automatically reach there without the need of any driver or anyone. The car is a very basic car consisting of the chassis, wheels, motors and body. The other components are RC receiver and transmitter, GPS, sensors for various purposes, lipo batteries and charger. U need to choose a ground control station for the project among various available like Arm Planner, Mission Planner, MAVProxy etc. 44

Khoj - Student Articles It is a small project based idea for the autonomous car. Similarly, we can use the same on a larger scale with the proper functioning of everything. But why autonomous vehicles and why the electric vehicles are the future? The graph and the image shows the market cap of Tesla and the number of units sold in 2020 creating much more revenue for the company. Now raises the question, why people are going for tesla over the other companies which are existing for many years? The reasons behind opting for electric vehicles are listed below: ▪ Minimum fuel consumption ▪ Environment friendly ▪ Convenient and easy to maintain ▪ No need for a driver in the case of autonomous car ▪ Less physical parts as compared to petrol and diesel cars ▪ Less maintenance Electric vehicles are the future of the world and sooner or later they will replace the vehicles running on petrol and diesel. 45

Khoj - Student Articles Oxygen Concentrator You've probably heard of an oxygen Kanak Tekwani concentrator, but are you sure what it is? (19BEC134) Do you have any questions about how an oxygen concentrator works? People with low blood oxygen levels can benefit from oxygen devices. They used to be big and cumbersome, making it difficult for individuals who required oxygen treatment to receive it outside of their homes. Modern oxygen generators, on the other hand, are smaller, lighter, and designed particularly for portability. A medical device that concentrates oxygen from ambient air is known as an oxygen concentrator. Atmospheric air contains around 78 % Nitrogen and 21% oxygen, with the remaining 1% made up of various gases. The oxygen concentrator takes in this air, filters it via a sieve, releases the nitrogen back into the atmosphere, and concentrates on the oxygen that remains. This oxygen is 90-95 % pure when compressed and delivered through a cannula. Concentrators include a pressure valve that helps control supply, which can range from 1 to 10 litres per minute. Concentrators are intended for continuous operation, according to 2015 WHO study, and may generate oxygen 24 hours a day, 7 days a week, for up to 5 years or more. Is 90 – 95 % purity enough? Experts believe it's excellent enough for mild and moderate Covid-19 patients with oxygen saturation levels of 85 % or above, even if it's not as pure as LMO (99 %). It is not, however, recommended for ICU patients. Although several tubes may be connected to a concentrator to serve two patients at once, doctors do not suggest it because of the danger of cross-infection. 46

Like this book? You can publish your book online for free in a few minutes!
Create your own flipbook