Saksham Dewan

PhD researcher University at Albany — SUNY

I study applications of machine learning and artificial intelligence for physical layer waveform and receiver design in wireless and optical communications.

Advisor : Dr. Hany Elgala

Portrait of Saksham Dewan

Research

My research focuses on waveform and receiver design for wireless and optical communications, with particular attention to robustness and receiver complexity.

Ongoing research

Learned wireless waveforms

Designing waveform bases that reorganize how symbols interact in fractional delay–Doppler channels, so a receiver can focus on a limited number of dominant interactions.

Comparing learned designs with OTFS and AFDM under matched channels and receiver budgets to understand where adaptation helps.

Waveform learning · Receiver complexityCollaboration between University at Albany — SUNY and IBM
IEEE MILCOM 2026 · Accepted

Residual CFO robustness and PAPR reduction

Jointly learning transmit and receive bases to reduce interference from residual carrier-frequency offset, alongside a model-guided receiver that refines symbol estimates.

The work explores the tradeoff between communication reliability and peak-to-average power ratio.

Synchronization · PAPR · Learned receiversPublication details
IEEE ICMLCN 2026 · Presented

Adaptive VLC constellations

Learning constellation geometry and receiver decisions together under LED clipping constraints, including modulation orders beyond the usual powers of two.

This allows finer control of transmission rate in visible light communication systems.

Optical OFDM · Constellation learningPublication details
Master’s research

Multi-task learning for VLC demodulation

Using cross-stitch neural networks to recover multiple information streams within a composite VLC waveform, sharing useful features between related demodulation tasks.

An interpretable measure of task dependence helps explain when that shared processing is useful.

Multi-task learning · Visible light communicationCode
  1. 2026

    Jointly Learned Waveforms for Residual CFO Robustness and PAPR Reduction

    S. Dewan, H. Elgala, A. Tzadok, and M.-C. Chang

    IEEE MILCOM Accepted for presentation

  2. 2026

    End-to-End Learned Non-Power-of-Two Constellations for Adaptive VLC Transmission

    S. Dewan and H. Elgala

    IEEE ICMLCN Presented

  3. 2025

    Leveraging Multi-Task Learning to Model Task Relationships in Composite VLC Waveform Demodulation

    S. Dewan, H. Elgala, S. Muhaidat, and P. Sofotasios

    Applied Optics

  4. 2024

    Towards Low Complexity VLC Systems: A Multi-Task Learning Approach

    S. Dewan and H. Elgala

    Communications & Networks Connect

  5. AI-Enabled Physical Layer Design for Visible Light Communication

    S. Dewan, M. Elamassie, M. Uysal, and H. Elgala

    IEEE Communications Magazine Submitted

Collaborative work

  1. 2026

    Optical OFDM Modulation Identification with Deep Learning

    H. Chen, H. Elgala, S. Dewan, and S. Cheng

    IEEE ICMLCN Presented

  2. 2025

Education

  1. 2023 — Present

    PhD, Electrical & Computer Engineering

    University at Albany — SUNY · Albany, NY, USA

  2. 2022 — 2024

    MS, Electrical & Computer Engineering

    University at Albany — SUNY · Albany, NY, USA

  3. 2017 — 2021

    BTech, Electrical Engineering

    Punjab Engineering College · Chandigarh, India

Experience & service

  1. Jan 2026 — Present

    Research Assistant

    University at Albany — SUNY · Albany, NY, USA

    Through the university’s collaboration with IBM, developing learned waveform bases for fractional delay–Doppler channels and evaluating them against OTFS and AFDM under matched receiver budgets.

  2. Jun 2026

    Peer reviewer

    IEEE Transactions on Communications

  3. Jan 2023 — Dec 2025

    Teaching Assistant

    University at Albany — SUNY · Albany, NY, USA

    • ECE 111: Introduction to ECE
    • ECE 231: Digital Systems
    • ECE 401: Advanced Electronic Circuits
    • ECE 422: Integrated Circuits
    • ECE 431: Reconfigurable Computing
    • ECE 481: Linear Control Theory
  4. Jun 2020 — Jan 2022

    Technical Product Manager

    Kolligate · Seattle, WA · Remote

    • Led delivery of a dental staffing platform, including feature planning, payment integration, and onboarding workflows; supported 300+ users in its first quarter.
    • Designed 120+ screens across web, iOS, and Android; coordinated engineering priorities, QA, and vendors.
  5. May — Dec 2021

    Full-Stack Developer

    Busywizzy · Gurugram, India

    • Led a seven-person team delivering Kolligate and designed Laravel/MySQL REST APIs for web and mobile applications.
    • Authored 200+ test cases covering functionality and performance requirements.
  6. Jan — Jun 2020

    Full-Stack Developer Intern

    Quantela · Bengaluru, India

    • Developed a real-time COVID-19 dashboard covering India and worldwide statistics.
    • Built Angular/TypeScript components and integrated Node.js APIs and municipal data feeds.