M.Sc. Researcher in Electrical & Electronics Engineering
Koç University · Wireless Networks Laboratory · Istanbul, Türkiye
Supervised by Prof. Sinem Coleri
I am a Master's researcher at Koç University, working in the Wireless Networks Laboratory under the supervision of Prof. Sinem Coleri. My research focuses on developing uncertainty-aware machine learning, safe reinforcement learning, and resource-efficient neural architectures for next-generation (5G-Advanced / 6G) wireless communications and Open RAN (O-RAN) systems.
Between my B.Sc. and M.Sc., I worked as a Medical Imaging and AI intern at Radiologics, building CT segmentation and detection pipelines. Previously, I earned my B.Sc. in Electrical and Electronics Engineering from Bilkent University on an 80% Merit Scholarship (Top 15% class ranking).
🎯 I am actively seeking PhD positions starting Fall 2027 in machine learning for wireless communications, safe & risk-sensitive reinforcement learning, and statistical signal processing.
- Safe & Uncertainty-Aware Reinforcement Learning: Multi-agent DRL, exact-cardinality sampling, execution-aware safety shields, and constrained policy optimization.
- Statistical Learning & Reliability Guarantees: Online conformal prediction, risk-sensitive measures (CVaR), lower-tail latency guarantees for URLLC.
- Hierarchical O-RAN Control & Network Slicing: Multi-timescale resource budgeting (Non-RT / Near-RT RIC xApps / O-DU) for heterogeneous eMBB–URLLC coexistence.
- Resource-Efficient Deep Learning & PHY/MAC Layer: Low-complexity neural architectures, Channel State Information (CSI) prediction, MIMO/Massive MIMO spatio-temporal modeling.
|
Mohammad Hussain, Maedeh Adibag, Dilara Gurer, Gokhan Kalem, Kerim Serin, Sinem Coleri Proposed a lightweight GRU-Attention predictor with bottleneck gated fusion and a Dimension-wise Separable Linear Head (DSLH). Achieves −13.84 dB NMSE with 26% fewer parameters and 2.3× higher inference throughput than LinFormer baselines on 3GPP TR 38.901-compliant channels. |
- Graduate Research Assistant, Koç University (Oct 2025 – Present) — AI-driven O-RAN resource allocation with Turkcell and Opticoms; GRU-attention encoder achieving 26% fewer parameters and 2.3× throughput over LinFormer at −13.84 dB NMSE; hierarchical scheduling pipeline combining learned surrogates, conformal demand bounds, and convex optimization for URLLC/eMBB workloads.
- Medical Imaging and AI Intern, Radiologics (Jul 2025 – Oct 2025) — CT segmentation (nnU-Net v2, 44 structures), YOLOX detection on 50K pediatric radiographs, and a QCT bone-density pipeline with HU-to-BMD calibration.
- 🌟 resource-efficient-csi-prediction: Official implementation of the IEEE Communications Letters 2026 paper, featuring GRU-Attn-DSLH, baseline models (LinFormer, LSTM, GRU), and MATLAB QuaDRiGa dataset generation.
- 📘 MHDoh.github.io: Academic portfolio & research website hosted via GitHub Pages (mhdoh.github.io).
| Languages |
|
| Deep Learning |
|
| Wireless & Sim |
|
| Tools & Systems |
|
| Degree | University | Field | Period | Details |
|---|---|---|---|---|
| 🎓 M.Sc. | Koç University | Electrical & Electronics Engineering | Oct 2025 – Present | GPA: 3.57/4.00 · Wireless Networks Lab (Advisor: Prof. Sinem Coleri) |
| 🎓 B.Sc. | Bilkent University | Electrical & Electronics Engineering | Graduated Jun 2025 | GPA: 3.27/4.00 · Top 15% · 80% Merit Scholarship |
Open to research discussions, PhD opportunities, and collaborations.