»IIIT Delhi UAV Team - Aurora led by undergraduate students Akhil Jarodia, Aseem Saini and Mukul Chhabra secured 6th rank in the 17th edition of the coveted Association for Unmanned Vehicle Systems International Student Unmanned Aerial Systems (AUVSI SUAS) Competition held at the US Navy Air Station in Maryland, USA from 12th-15th June 2019. A total of 75 teams participated in the competition. In addition to this, the team has been awarded prize money of $2250. The team used a hexacopter UAV platform which cleared all milestones and deliverables including fully autonomous mission, stationary & moving obstacles avoidance, autonomous air-delivery of a water bottle and object detection & classification.
»SPARC project [Ministry of HRD, Govt of India] titled "Post-Silicon Debug Platform for Secured Systems" with Dr. Sujay Deb (Indian-PI), Prof. Prabhat Mishra (International PI), Dr. Saket Anand (Indian Co-PI), and Dr. Kanad Basu (International Co-PI) got approved with a budget sanction for 48 lakhs. This project is a collaboration work between IIIT Delhi, University of Florida and New York University.
»Viraj Parimi (B.Tech CSE, 2015), Sharmistha Gupta (B.Tech ECE, 2016) and Raghavv Goel (B.Tech ECE, 2016) have been selected for CMU-RISS 2019, Carnegie Mellon University’s eleven-week summer research internship program.The RISS (Robotics Institute Summer Scholars) program is one of the best undergraduate robotics research opportunities in the world. It is a fully funded program that selects around 35 undergraduate students every year worldwide to carry out research in the domain of robotics and related fields. The selected students work on state-of-the-art projects under the mentorship of faculty at CMU.
» Dr. Abhijit Mitra has been awarded the Professional Achievement Award by the British Council Alumni Awards India-2019 on March 15.
» Our Ph.D. candidate, Deepayan Banerjee, have received Student Travel Grant for Outstanding Work for his paper at the IEEE 69th ECTC to be held at Las Vegas, Nevada, USA. 10 students have been selected from all over the world and Deepayan is the only one to be selected from India.
» Best graduate forum award at ICDCN 2019 for Mr. Rohit Kumar (PhD student NIT Delhi, co-supervised by Dr. Sumit)
» Second-best poster award (among 34 posters from various IITs and international universities: https://www.comsnets.org/accepted_posters.html) for the paper titled "LSTM Guided Modulation Classification and Experimental Validation for Sub-Nyquist Rate Wideband Spectrum Sensing" authored by S. Chandhok (JMI Delhi), Himani Joshi (PhD Student, IIIT-Delhi), Sumit J. Darak (Faculty, IIIT-Delhi) and A. Subramanyam (Faculty, IIIT-Delhi)
» Joint best graduate forum awards at COMSNETs 2019 to Himani Joshi (PhD Student, IIIT-Delhi) and Rohit Kumar (PhD student NIT Delhi, co-supervised by Dr. Sumit).
» Book by our faculty member Dr. Agnshul Majumdar, titled "Compressed sensing for Engineers" is out on stands. The book introduces the topic of compressed sending in a level accessible to engineering
students. Reading it requires only a basic understanding of linear and optimization. The complex mathematical concepts are explained intuitively. The book mainly concentrates on algorithms and applications (in signal processing and machine
learning). It contains Matlab implementation of almost all the algorithms discussed.
Compressed sensing (CS) studies the solution of a system of under-determined linear equations when the solution is known to be sparse. The subject revolutionised signal processing and machine learning since it's onset in late 2000s. CS paved way for faster MRIs and safer CT scans. It paved way for modern representation learning; CS concepts are being used today in sparse techniques for deep learning. This book introduces CS at the level accessible by professional engineers and does not expect an advanced mathematical background required for understanding other treatises on this topic. The emphasis is on algorithms and applications of compressed sensing.
» Paper titled "New Spectrum Efficient Reconfigurable Filtered-OFDM Based L-Band Digital Aeronautical Communication System" authored by Niharika Agrawal (Ph.D. Student, IIIT Delhi), Sumit Darak (Faculty,
IIIT Delhi) and Faouzi Bader (Faculty, CentraleSupelec, France) has been accepted for publication in the journal IEEE Transactions on Aerospace and Electronic Systems.
To meet an ever-increasing demand of the spectrum for aeronautical communication, we propose new L-band digital aeronautical communication system (LDACS) using reconfigurable filtered OFDM (Ref-OFDM). The proposed protocol enables transceivers to dynamically adapt the transmission bandwidth over a wide range to meet the desired quality of service, and high out-of-band attenuation leads to significant improvement in the vacant spectrum utilization. We support our claims via theoretical analysis as well as simulation and complexity comparison results for various realistic channel conditions.
» Infosys Center for Artificial Intelligence at IIITD is organizing winter school on Artificial Intelligence. This Winter School will focus on the fundamentals and advances in Artificial Intelligence and
Machine Learning, along with application domains of computer vision, natural language processing, autonomous driving, and robotics. The lectures will be accompanied with lab sessions in the evening and the last day will be a hackathon on "AI
for social good".
Students and professionals interested in learning algorithms in machine learning and artificial intelligence can apply. http://cai.iiitd.ac.in/AIwinterschool2019/
» Paper titled, "Distributed Learning and Optimal Assignment in Multiplayer Heterogeneous Networks" authored by H. Tibrewal (BTech student, IIT Bombay), S. Patchala (BTech student, IIT Bombay), M.
Hanawal (Faculty IIT Bombay) and Sumit Darak (Faculty IIIT Delhi) has been accepted in IEEE INFOCOM 2019 (Core A* conference) scheduled for 29th April - 2nd May at Paris, France.
We consider an ad hoc network where multiple users access the same set of channels. The channel characteristics are unknown and could be different for each user (heterogeneous). No controller is available to coordinate channel selections by the users, and if multiple users select the same channel, they collide and none of them receives any rate (or reward). For such a completely decentralized network we develop learning algorithms that aim to achieve optimal network throughput. For the wideband sensing and narrowband sensing scenarios, we first develop explore-and-commit algorithms that converge to near-optimal allocation with high probability in a small number of rounds. Building on this, we develop an algorithm that gives logarithmic regret.
» Congrats ! Dr. Vivek Ashok Bohara for got a reserach grant from SERB-TARE for Project titled "Machine Learning for Efficient Millimeter-Wave Communication"
» Congrats ! Naveen Gupta for got a travel grant to attend IEEE GLOBAL COMMUNICATIONS CONFERENCE (GLOBECOM) to be held in Singapore from 04-12-2017 to 08-12-2017approved by DST SERB for the paper titled "OFDMA Based Angle-Constrained Underlay Device-to-Device Communication" co-authored by Dr. Vivek Bohara (Faculty, IIIT-D).