http://repository.iiitd.edu.in/xmlui/handle/123456789/1081 2026-04-10T22:12:00Z 2026-04-10T22:12:00Z Srinivasan, Shriya Girish, Vaibhav Sethi, Tavpritesh (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1211 2023-04-16T22:00:25Z 2022-05-01T00:00:00Z

Resource allocation using reinforcement learning Srinivasan, Shriya; Girish, Vaibhav; Sethi, Tavpritesh (Advisor) Resource allocation is a problem that requires complex decision making. Our focus is to solve this problem in the healthcare sector using Reinforcement Learning. We propose an RL pipeline that starts with a sequential decision deep RL model and combines it with a Contextual Bandit approach. The Reinforcement Learning models suggest actions and rewards whereas the Contextual Bandits allows for dynamic change in context so that allocation can take place in real-world scenarios too, where the environment is not static. We have also used mathematical randomised optimisation models to compare the results received by the RL models. Our goal is to make Reinforcement Learning models Plug and Playable through a platform so that the user can use these models to solve their Resource Allocation Problem without knowing Reinforcement Learning.

2022-05-01T00:00:00Z Goyal, Anmol Darak, Sumit Jagdish (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1210 2023-04-16T22:00:25Z 2022-07-01T00:00:00Z

Developing solutions for remote SoC labs Goyal, Anmol; Darak, Sumit Jagdish (Advisor) Embedded System labs involve working on Field Programmable Gate Arrays (FPGAs) and Zynq System-on-Chips (SoC). The key aspect is that these boards can be programmed and tested remotely. However, conducting remote labs for such courses with large class sizes, particularly during COVID-19, breeds a lot of challenges, due to limitations in time and resources. Different institutes have different systems in place for conducting remote SoC labs. We propose a system which provides an organized, efficient and convenient way of allocating lab resources and timed access to students, requiring minimal number of steps for the students and minimum intervention from the teaching staff. Not only that, it attempts to provide an experience as close to the physical labs as possible. The implementation involves a well-designed Graphical User Interface(GUI) using Python Django, along with Python scripts running on lab desktops and Linux SSH server, and Arduino UNO-based circuits. The GUI also enables the live monitoring of the board being worked upon. We aim to deploy this solution to technical institutes which include embedded system courses in their curriculum. It also has scope to help the conduction of these courses on online learning platforms. It aims to guarantee the smooth functioning of remote SoC labs, yet provide scope for future improvements.

2022-07-01T00:00:00Z Pal, Tathagat Bohara, Vivek Ashok (Advisor) Srivastava, Anand (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1122 2023-04-11T22:00:20Z 2022-12-01T00:00:00Z

RIS aided B5G V2V solution for road safety applications Pal, Tathagat; Bohara, Vivek Ashok (Advisor); Srivastava, Anand (Advisor) Terahertz (THz) communication is a viable technology for the 6G wireless networks. THz frequencies often have a limited coverage area because of their extremely high spread attenuation and molecular absorption. Consequently, new methods are necessary to get around these limitations. Additionally, reconfigurable intelligent surfaces (RIS) are used to modify the wireless propagation environment and enhance the coverage area. As a result, we highlight the benefit of deploying RIS for enhanced vehicular message dissemination. We utilize a novel momentgenerating functional based approach to derive the ergodic signal-to-noise ratio (SNR) for the proposed system. Numerical results and findings reveal that the proposed RIS aided THz-V2V system can significantly enhance BER and sumrate performance as compared to opical IRS aided vehicular-visible light communication (V-VLC). Moreover, the project demonstrates the impact of different weather conditions such as snow, rain, etc. on the proposed RIS-aided THz-V2V communication system in terms of SNR and BER.

2022-12-01T00:00:00Z Singhal, Shubhi Ram, Shobha Sundar (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1121 2023-04-11T22:00:20Z 2022-05-01T00:00:00Z

Probability of target detection in bistatic and multistatic radars Singhal, Shubhi; Ram, Shobha Sundar (Advisor) In the first part of the work, we evaluate the detection performance of a bistatic radar using a metric called bistatic radar detection coverage probability. The setting considered for the target detection consists of noise as well as clutter scatterers, which creates the need to distinguish the target responses from the noise and clutter responses. The number of clutters is modelled as a Poisson distribution and the location of clutters is modelled as a uniform distribution. The considered noise has been derived from the Johnson-Nyquist noise formula. This setting has been experimentally simulated using Monte Carlo method and the corresponding probability of target detection has been found out under various conditions. The experimental results are used to verify the theoretical results from the formula giving the detection coverage probability directly in terms of target, clutter and radar parameters. In the second part of the work, we go forward and work with multistatic radars under the scenario when there is a single receiver and multiple transmitters, where the no. of transmitters as well as the position of transmitters are not fixed. For a target to be detectable by the multistatic radar, it should be detectable by atleast one transmitter-receiver combination. We use stochastic geometry to find out a theoretical formula for the probability of a target being detected by the multistatic radar. Using Monte Carlo simulations, we experimentally verify the correctness of the formula obtained and it is found that it almost exactly matches the results from the experimental method. We have also analysed various trends of probability of detection on varying the various parameters. The derived formula can be used to gain useful insights on the limits on parameters like the transmitted power and the bandwidth.

2022-05-01T00:00:00Z