HELLO, I'M

Apala Pramanik

PhD Researcher • mmWave + Machine Learning • V2I / V2X

Lincoln, NE, USA

Open to Internship & Research Opportunities — ML for Wireless / V2I / V2X

I am a PhD student at the CPN Lab, University of Nebraska-Lincoln, advised by Dr. Mehmet Can Vuran. My research focuses on enhancing mmWave communication through advanced beamforming, beam tracking, and prediction techniques, integrating machine learning and sensory data for improved wireless connectivity.

5

Publications

2

First-Author

1

Best Paper Candidate

3+

Years Research

View CV Download Email Me

Experience & Education.

Aug 2024 — Present

PhD, Computer Science

University of Nebraska-Lincoln • CPN Lab

Advised by Dr. Mehmet Can Vuran, Dale M. Jensen Chair Professor.

Camera-Primed Double-Directional Beam Management (IEEE SECON 2026, co-first author): vision-assisted mmWave beamforming with real-time TX/RX beam selection from camera-derived spatial cues and adaptive correction under hardware constraints.
Reinforcement Learning for mmWave Power Control: hardware-in-the-loop RL framework to optimize transmit power and RF gain under SNR constraints on SDR-based testbeds with real-time feedback.
Modeled mmWave propagation with Wireless InSite and Sionna RT ray-tracing to validate experimental observations.
Characterized surface-guided propagation and bistatic radar scattering using liquid-metal (EGaIn) and metallic reflectors.

mmWave Beamforming Reinforcement Learning GNU Radio / USRP Ray Tracing

Aug 2021 — May 2025

MS, Computer Science

University of Nebraska-Lincoln

Advised by Dr. Dung Hoang Tran.

Perception-based Runtime Monitoring and Verification for Human-Robot Construction Systems — presented at ACM/IEEE MEMOCODE 2024 (Best Paper Candidate).
Used Neural Network Verification (NNV) for formal analysis of deep learning models in safety-critical settings.
Applied Signal Temporal Logic (STL) to specify and verify temporal properties in cyber-physical systems.
Built an edge-assisted SLAM pipeline with YOLOv3 detection and Kalman filtering on depth-camera point clouds, enforcing a 1.25 m human-safety radius in ROS/Gazebo.

Runtime Verification NNV Signal Temporal Logic ROS / Gazebo

Aug 2017 — May 2021

B.Tech, Electronics & Communication

Guru Gobind Singh Indraprastha University • Delhi, India

Thesis: Deep learning techniques for Magnetic Resonance Image (MRI) denoising and reconstruction from undersampled scans.

Published ASTITVA (ICCCIS 2021) — assistive technologies for visually challenged users.
Co-authored START: Smart Stick based on TLC Algorithm in IoT Networks (I-SMAC 2020).
Co-authored SERI: Secure Routing in IoT (ICIoTCT).

Deep Learning MRI Reconstruction IoT

Research Statement.

My research focuses on making mmWave wireless links survive the real world — where vehicles move, blockages happen, and hardware imposes hard constraints. I combine machine learning, camera and sensor cues, and ray-traced propagation models to enable fast, reliable double-directional beam management for V2I and V2X connectivity.

I'm especially interested in closing the loop between perception and the physical layer: using what a camera sees to prime where a beam points, and using reinforcement learning to adapt transmit power and gain in real time under SNR constraints. The goal is practical — algorithms that run on SDR testbeds today and scale to vehicular networks tomorrow.

mmWave Beam Management Vision-Aided Wireless Reinforcement Learning Ray Tracing V2I / V2X Software-Defined Radio

Publications.

Look Once, Beam Twice: Camera‑Primed Real‑Time Double‑Directional mmWave Beam Management for Vehicular Connectivity

2026 IEEE International Conference on Sensing, Communication, and Networking (SECON) (to appear)

Authors: Avhishek Biswas*, Apala Pramanik*, Eylem Ekici, Mehmet Can Vuran
*Equal contribution — co-first authors

Perception-based Runtime Monitoring and Verification for Human-Robot Construction Systems Best Paper Candidate

ACM/IEEE MEMOCODE 2024

Authors: Apala Pramanik, Sung Woo Choi, Yuntao Li, Luan Viet Nguyen, Kyungki Kim, Hoang-Dung Tran

ASTITVA: assistive special tools and technologies for inclusion of visually challenged

2021 international conference on computing, communication, and intelligent systems (ICCCIS)

Authors: Apala Pramanik, Rahul Johari, Nitesh Kumar Gaurav, Sapna Chaudhary, Rohan Tripathi

START: smart stick based on TLC algorithm in IoT network for visually challenged persons

2020 Fourth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud)(I-SMAC)

Authors: Rahul Johari, Nitesh Kumar Gaurav, Sapna Chaudhary, Apala Pramanik

SERI: SEcure Routing in IoT

International Conference on Internet of Things and Connected Technologies

Authors: Varnika Gaur, Rahul Johari, Parth Khandelwal, Apala Pramanik

Blogs.

Reinforcement Learning

Power Control for mmWave: From Optimization to RL

November 2025

Surveying the evolution of transmit power control methods for mmWave, from classical optimization to deep reinforcement learning.

Machine Learning

ML-Driven Beam Tracking for V2I Communication

July 2025

How machine learning can enhance beam prediction and tracking in vehicle-to-infrastructure communication scenarios.

mmWave

Designing Indoor Spaces for mmWave Coverage

April 2025

A modular design approach for building interiors that inherently support mmWave propagation through pre-characterized layout elements.

mmWave

Understanding Millimeter Wave Beamforming

October 2024

An introduction to mmWave beamforming techniques and their role in next-generation wireless communication systems.

Robotics

Runtime Monitoring for Human-Robot Systems

March 2024

Exploring perception-based approaches to ensure safety in human-robot collaborative construction environments.

Skills.

Core skills are what I use daily in research. Familiar tools are ones I've shipped with but use less often.

Programming Core

Python (PyTorch, TensorFlow, NumPy, Pandas, Scikit-learn)
MATLAB
C / C++
Shell (Bash / Zsh)
ROS (rospy, roscpp)

Wireless & RF Core

GNU Radio
USRP / Software Defined Radio
Wireless InSite (Ray Tracing)
Sionna RT
mmWave Propagation Modeling
Bistatic Radar Measurement

Machine Learning Core

PyTorch / TensorFlow
CNN, RNN, GRU, LSTM
Transformers / Vision Transformers
Complex-Valued Neural Networks
Reinforcement Learning (hardware-in-the-loop)

Tools & OS Core

Ubuntu / Linux
Git
Docker
LaTeX (Overleaf)

Robotics & Verification Familiar

Gazebo / RViz
Point Cloud Library (PCL)
Neural Network Verification (NNV)
Signal Temporal Logic (STL)
Depth Cameras / Point Cloud Sensors

Embedded & Undergrad Toolchain Familiar

Arduino
AVR Studio
KEIL
Proteus Professional

Projects.

2024

3D Ray Tracing for mmWave Propagation

24 GHz 3-bounce reflections MATLAB

A MATLAB-based ray tracing simulator for modeling millimeter-wave signal propagation in indoor environments. Traces rays through a configurable 3D layout, computing received power with free-space path loss, multiple reflections, and log-normal shadowing. Modular environment modeling with power-gradient visualization.

View on GitHub

2024

Multi-Modal Traffic Pattern Detection

>80% accuracy <100 ms latency USRP + CVNN

Real-time multi-modal traffic classification system fusing synchronized USRP RF signals and camera images (±10 ms sync). A Complex-Valued Neural Network classifies traffic into Very Busy, Moderate, and Not Busy states with a watchdog mechanism for reliable continuous operation.

View on GitHub

2023

Edge-Assisted SLAM for Human-Robot Construction

1.25 m safety radius YOLOv3 + Kalman ROS

Tested the EdgeSLAM algorithm at a construction site where a robot navigates to goals while maintaining a 1.25 m safe distance from human workers. Uses YOLOv3 for detection, plus clustering and Kalman filtering on depth-camera point clouds to estimate relative human motion in the robot's local frame.

2023

Vision Transformer Re-implementation

Self-Attention PyTorch

Re-implemented the Vision Transformer (ViT) architecture from scratch, exploring self-attention mechanisms for image recognition. Studied scaling challenges, pretraining strategies in data-limited regimes, and fine-tuning techniques compared to conventional CNNs.

Sequence Modeling with GRU, LSTM, and Transformer

2023

Sequence Modeling: GRU, LSTM & Transformer

GRU / LSTM Transformer TensorFlow

Comparative study of recurrent and attention-based architectures for sequence modeling. Implemented and evaluated GRU, LSTM, and Transformer models on time-series prediction and NLP tasks, analyzing trade-offs between compute efficiency, long-range dependency handling, and accuracy.

Hobbies.

Beyond research, these are the things that keep me balanced and curious.

Bharatnatyam

Senior diploma in Indian classical dance with 10+ years of training.

Gym & Fitness

Passionate about weight training and maintaining a healthy, active lifestyle.

Travelling

Exploring new places, cultures, and cuisines whenever the chance comes up.

News.

Mar 2026 Paper “Look Once, Beam Twice: Camera‑Primed Real‑Time Double‑Directional mmWave Beam Management for Vehicular Connectivity” accepted at IEEE SECON 2026 (co-first author)
2026 Reviewer for IEEE Transactions on Wireless Communications
2025 Reviewer for IEEE Transactions on Wireless Communications
May 2025 Graduated with MS in Computer Science from University of Nebraska-Lincoln
Oct 2024 Presented “Perception-based Runtime Monitoring and Verification for Human-Robot Construction Systems” at ACM/IEEE MEMOCODE 2024 — Best Paper Candidate (Raleigh, NC)
Aug 2024 Started PhD in Wireless Communications with focus on millimeter wave for V2I at UNL

Outreach.

Sharing wireless research with students across every level — from curious kids to future PhDs.

Raikes School Fair

Presented live demonstrations of mmWave beamforming to Raikes School of Computer Science & Management undergraduates, walking them through how directional beams are formed and comparing the performance of different beamforming strategies. The goal was to pull back the curtain on what graduate research actually looks like and motivate them to pursue research careers of their own.

mmWave Demo Beamforming Research Motivation

Hour of Code

Designed interactive games inspired by mmWave beamforming and beam alignment to explain wireless research to K–12 students during the Hour of Code event. Kids got to physically “steer beams,” align transmitters and receivers, and see for themselves that research can be playful, creative, and genuinely cool.

Interactive Games Beam Alignment K–12 STEM

REU Summer Internship — UNL

Mentored undergraduate researchers through the NSF REU summer internship program at UNL, guiding them through hands-on contributions to our ongoing projects. Helped them ramp up on experimental design, data collection, SDR-based testbeds, and technical communication — so they could leave the summer with real research skills and a tangible project they built themselves.

NSF REU Experimental Design SDR Testbeds

Let's Connect.

I'm always open to conversations about wireless communication, ML research, internships, and collaborations.

I usually reply within 24 hours.

apramanik2@huskers.unl.edu