Ozan Aygun

Ozan Aygun

PhD Candidate · NYU WIRELESS · CommIT

Brooklyn, NY · ozan at nyu.edu

about

I am an Electrical and Computer Engineering PhD candidate at NYU Tandon, advised by Prof. Elza Erkip in the CommIT Research Group. Previously, I spent two summers at Apple and one summer at SHARP Labs of America as a research intern. I received my B.S. and M.S. in Electrical and Electronics Engineering from Bilkent University, Türkiye, with high honors.

research

My research sits at the intersection of machine learning and wireless communications, spanning multiple layers of the network stack. I am broadly interested in building learning-based systems that are communication-aware, and in bringing communications perspectives to machine learning problems.

Federated learning over wireless channels: How can distributed devices collaboratively train models over noisy wireless channels using over-the-air aggregation, even under data heterogeneity and energy constraints?J1C1C2

Network control with constrained deep RL: How can constrained reinforcement learning enable cost-efficient, latency-aware resource management and application delivery in communication networks?C3

Learning-based mobility prediction: How can machine learning reduce signaling and measurement overhead in mobile networks through accurate prediction of radio resource measurements?C5

Learning-based compress-and-forward: How can neural networks learn efficient compression and forwarding strategies in relay channels, bridging information-theoretic coding and deep learning?C4

Always happy to chat about research—reach me via email.

selected work

Over-the-Air Federated Edge Learning with Hierarchical Clustering

O. Aygün, M. Kazemi, D. Gündüz, T.M. Duman

IEEE Transactions on Wireless Communications, 2024

Reference Signal Received Power Prediction for Measurement Gap Reduction

O. Aygün, O. Orhan, O. Sahin, N. Goris, A. Naguib

IEEE WCNC, 2026 — Work done during internship at Apple.

A Constrained RL Approach for Cost-Efficient Delivery of Latency-Sensitive Applications

O. Aygün, V. N. Vitale, A. Tulino, H. Feng, E. Erkip, J. Llorca

Asilomar Conference on Signals, Systems, and Computers, 2025

background

Education

2022 – present
Ph.D., Electrical and Computer Engineering — New York University Tandon
Advisor: Prof. Elza Erkip · Focus: ML, wireless communications, network control
2020 – 2022
M.S., Electrical and Electronics Engineering — Bilkent University
Advisor: Prof. Tolga M. Duman · Thesis: Federated Learning Over Wireless Channels with Over-the-Air Aggregation
2015 – 2020
B.S., Electrical and Electronics Engineering — Bilkent University
High Honour Student

Experience

May – Aug 2025
Software Intern — Apple Inc., Cupertino, CA
AI/ML-based mobility prediction; 3GPP RAN2 contributions; authored manuscript [C5]
May – Sep 2024
Software Intern — Apple Inc., San Diego, CA
Time-series forecasting pipelines for AI/ML-based mobility prediction (PyTorch, XGBoost); 3GPP RAN2 August 2024 contribution
May – Aug 2023
Communications Systems Research Intern — SHARP Labs of America, Vancouver, WA
3GPP-compliant AI/ML-based positioning algorithm; integrated with NYUSIM channel simulator

news