• 👋 Hi, I’m Himanshu.
• 👀 I’m interested in SDN and IoT.
• 🌱 I’m currently learning networking, high-level and low-level design.
• 💞️ I’m looking to collaborate on SDN, Embedded Systems, DLMS/COSEM smart meters.
• 📫 Reach me via:
  • Gmail
  • IITR Webmail
  • IITD Webmail
  • LinkedIn.

CGPA - 8.143/10
Projects:

• DLMS/COSEM implementation on following platforms:
  • Linux (Ubuntu, RaspberryPi)
  • STM32F207ZG
• MERN website for displaying smart meters data.
• DLMS-Wireshark for displaying expanded (array and structure in) DLMS normal packets in Wireshark.
• SDN TCP SYN flood detection and mitigation (ongoing).

CGPA - 8.383/10
Projects:

• FPGA-based implementation of Numerical Overcurrent relay.
• Industry-Oriented Project: Apportionate Battery-Capacitor (ABC) for Electric Vehicles.
• B.Tech. Project: 3D real-time mapping using Kinect v2 vision sensor in ROS environment.

• Percentage - 93.8%

• CGPA - 9.6/10

Projects:

• Cross-lingual text classification

Projects:

• Vehicle Control Unit (VCU) for Formula Student race car for Formula Green 2020 competition.

• DLMS/COSEM implementation on Linux (Ubuntu, RaspberryPi)
  • Implemented Register class in fork from EPRI:DLMS/COSEM in collaboration with github:sudeshna
  • Implemented a Data Compression algorithm in Python and integrated with the inter-host DLMS/COSEM communication using TCP/IP and shared memory approaches for inter-process communication.
  • Implemented the Data Compression algorithm in C/C++ using TNT Matrix libraries. This allowed implementation of the complete application in C/C++ completely, and a seamless transition of the algorithm to STM32 implementation later on.
  • Deployed 10 RaspberryPi-based DLMS/COSEM-enabled smart meters around IIT Delhi campus.
  • Deployed a DLMS/COSEM-enabled Head-End System on Linux VMs hosted on IIT Delhi Cloud - Baadal - for collecting data from the meters at r intervals and store in MongoDB database.
• DLMS/COSEM implementation on STM32F207ZG
  • Modified Linux libmodbus library to integrate to STM32 application, exposing read, write and delay functions for custom handling of modbus requests/responses.
  • Implemented ESP8266 libraries in two different ways to integrate with multi-threaded STM32 DLMS/COSEM application in collaboration with github:venus696.
• MERN website for displaying smart meters data
  • Hosted on Linux VMs on IIT Delhi Cloud - Baadal - accessible over IITD intranet for displaying data collected from the deployed smart meters.
  • The website is written for two projects - Air Pollution Monitoring Device (APMD) and Smart Meters (SM).
  • The website allows selecting a date-range or a rolling-plot option, and desired metrics from a list of available metrics.
  • OpenLayers used to render a map of IIT Delhi with an additional layer with overlays to show checkboxes with location-popups marking locations of the installed smart meters around the campus.
• DLMS-Wireshark for displaying expanded (array and structure in) DLMS normal packets in Wireshark
  • Modified fork from matousp:dlms-analysis github repository in lua to expand and display the COSEM structure and array (ASN.1 format) DLMS normal packets in Wireshark.

• Developed Vehicle Control Unit (VCU) for Formula Student Electric race car as per Formula Green 2020 rule-book:
  • Shutdown circuit to control and cut power supply from battery to motor controller (and motor).
  • Several PCBs (schematic and layout design, and soldering) to control the shutdown circuit.
  • Programming STM32F103RB controller for algorithms to implement some controls in and out of rule-book.
  • SPI communication of the STM32 board with an ADC and an nRF24L01+ board.

• Sentiment classification for a target language for which only small amounts of labelled data is available using transfer learning on a source language for which large datasets for the same are available:
  • Sentiment classification training of base layers of the Adversarial Network model on source language datasets.
  • Adversarial training with language detector branch of the Adversarial Network model.
• The model was implemented on Tensorflow-Keras framework for Amazon reviews dataset with English as source language and French as target language.