UbiNet: Ubiquitous Networked Systems Lab
@ubinetiitkgp
The research group at CSE, IIT Kharagpur led by Dr. Sandip Chakraborty (@sandip2201)
🚀 Excited to share that @prasenjit52282 has joined @smusg as a Visiting Postgraduate Research Student! He’ll be working with Prof. Archan Misra on energy-efficient ML sensing & smart homes powered by Foundation Models. Wishing him a fruitful journey ahead!🔬 #ResearchInternship

🚀 Excited to welcome Shrest, Pranjal, Tamanna & Nishant to #UbiNet @IITKgp! They bring fresh energy to research in sensing, HCI, edge computing, AI & mmWave tech. Looking forward to impactful work ahead! 🙌 #AI #HCI #Networking #IITKharagpur #Research

🤖 Our work "Generating Cost Optimized Smart Contracts from DL Code using LLMs" is accepted at #BCCA2025! PSTL uses LLMs to convert PyTorch models to optimized Solidity code, enabling CNNs on #Ethereum testnet! 🔥 #AI #SmartContracts #LLM #Blockchain #DeepLearning
🚗 Our paper "DeRide: An Incentivized P2P Ride Sharing Service using Permissioned Blockchain Networks" is accepted at #BCCA2025! Built on #HyperledgerFabric, DeRide uses non-monetary tokens for fair, decentralized ride sharing. Congrats to all! 🎉 #Blockchain #Web3 #Mobility
Thrilled to share that @argha_sen10 has joined @BellLabs Cambridge, UK as a Summer Research Intern! 🌟 He’ll be working on sensory wearables for medical technology, advancing cutting-edge healthcare solutions. 🚀 #NokiaBellLabs #WearableTech #HealthcareInnovation

From EE undergrads to key contributors in our lab’s mmWave journey — Soham Chakraborty (@ UIUC PhD) & Soham Tripathy (@ Eightfold.AI) helped shape our research, resulting in a MobiSys '24 poster, SmartComp '25 paper, and a patent! Wishing them the very best ahead!


Excited to serve as TPC Vice Chair for IEEE PerCom 2026—a top-tier (Core A*) conf on pervasive systems & networks! 📡✨ Join us in historic Pisa, Italy 🇮🇹 — home of the iconic Leaning Tower 🏛️ Submit your best work! #PerCom2026 #PervasiveComputing #IEEE @ubinetiitkgp
🌐 The #PerCom2026 website is live! 📍 Join us in Pisa 🇮🇹 at the CNR Area, March 16–20, 2026! Check the new website at: 🔗 percom.org Updates coming soon: stay tuned for workshops, CfP, and more! #IEEE #IEEEPerCom #PervasiveComputing #PerCom
After 10 impactful months at #UbiNet, we bid farewell to Manjeet Yadav, JRF and HCI enthusiast, who joined us post-MS from IISER Bhopal. He now begins his #PhD at IIT Jodhpur! 🎓 Wishing you the best, Manjeet—keep exploring and stay in touch! 🌟 #Farewell #IITJodhpur #HCI




An article "Pervasive Sensing to Correlate Vehicle Driving Behavior with City-Scale Traffic Dynamics" by @das_jee, Shameek Bhattacharjee, @sandip2201, Bivas Mitra, and Sajal Das has been accepted to IEEE Pervasive Computing Magazine. Congratulations to all the authors! #Pervasive


It has been an amazing experience chairing @IEEEPerCom 2025 workshops with Franca Delmastro from IIT-CNR, Italy. #PerCom this year had 14 different workshops on various topics, ranging from context & activity modeling, security, privacy & trust, pervasive applications, and so on.
Congratulations to @AritraHota, @Soumyaj44848812, and @sandip2201 for the Best Work in Progress Paper at @IEEEPerCom 2025, for the paper "Exploring Large Language Models in Active Learning for Annotating Physical Sensing Data."


And ... This work received the best WiP Award. Congratulations to @AritraHota, @Soumyaj44848812 , and @sandip2201.
The third WiP paper that @Soumyaj44848812 will be presenting tomorrow at @IEEEPerCom is on using LLMs for active learning over physical sensing data. We show that LLMs can be useful to support annotation of human activity data through active learning. youtu.be/djXGIx3DqLg?si…
Happy to share that I have been recognized as a distinguished member of IEEE INFOCOM 2025 TPC. This distinction is awarded based upon ratings by peer TPC members, fairness in review scores, and promptness in meeting various deadlines during the review process.
The third WiP paper that @Soumyaj44848812 will be presenting tomorrow at @IEEEPerCom is on using LLMs for active learning over physical sensing data. We show that LLMs can be useful to support annotation of human activity data through active learning. youtu.be/djXGIx3DqLg?si…