Hi, I'm Ahmed

PhD ECE Student | Radar Systems & ML Researcher

Specializing in Radar Signal Processing, Vital Signs Monitoring, and AI-powered Healthcare Solutions

View My Work Get In Touch
signal_processor.py
# Radar Signal Model
def received_signal(t, f_d, τ):
  # x(t) = Σ αᵢ·s(t-τᵢ)·e^(j2πfᵈt)
  return A * np.cos(2*π*f_0*t + φ)

def range_doppler_fft(x):
  # 2D-FFT → Range-Doppler Map
  return np.fft.fft2(x)

researcher = {
  "focus": "Coherent-On-Receive",
  "method": "SVD + Phase Est."
}
Radar Range R = cτ/2
Doppler fᵈ = 2v/λ
FFT X(k) = Σxₙe^(-j2πkn/N)
Vital Signs φ(t) = 4πd(t)/λ
SNR SNR = P·G²λ²σ / (4π)³R⁴kTB

About Me

I'm a PhD student in Electrical and Computer Engineering at the University of Oklahoma, specializing in radar signal processing, machine learning, and advanced digital signal processing.

My research at the Advanced Radar Research Center focuses on developing novel signal processing algorithms for non-coherent radar systems, including coherent-on-receive processing frameworks and Vision Transformer architectures for micro-Doppler classification. I've achieved state-of-the-art results with 98% classification accuracy using 98% fewer parameters than conventional models.

I'm passionate about translating complex engineering concepts into practical solutions. From non-contact vital signs monitoring using mmWave radar to precision agriculture platforms combining IoT sensors with AI, I focus on building end-to-end systems that solve real-world problems in healthcare, agriculture, and industrial monitoring.

5+

Years Experience

15+

Projects Completed

10+

Technologies

Ahmed

Featured Projects

Here are some of my recent works

VitalFlow Radar Architecture
Featured

VitalFlow Radar

A comprehensive non-contact vital signs monitoring system using mmWave radar technology. Features real-time heart rate and breathing rate detection with AI-powered anomaly detection.

Python FastAPI Kafka Vertex AI Docker React TypeScript
  • Real-time radar signal processing
  • Cloud-based data streaming with Confluent Kafka
  • AI-powered anomaly detection with Google Vertex AI
  • Interactive React dashboard with live visualizations
  • Edge computing support for Raspberry Pi
View on GitHub
Startup

LoteAI - Precision Agriculture Platform

Hybrid AI architecture built for agriculture. LoteAI is an open, sensor-agnostic data fusion platform that works with existing equipment to deliver high-resolution, time-series driven crop insights and real-time anomaly detection.

Machine Learning Data Fusion Remote Sensing IoT Agriculture
  • Hybrid AI architecture built specifically for agriculture
  • Open, sensor-agnostic platform (LoRaWAN or cellular)
  • 22 specialized ML models (LSTM, GBDT, ensembles, CNN feature extractors)
  • 10-meter spatial resolution multispectral imagery for sub-field mapping
  • 5-day temporal revisit cycles for dense vegetation time-series
  • Real-time anomaly detection in under 200ms response time
  • Offline-first architecture with edge buffering and sync when connected
Visit Website

FMCW Heart & Breathing Rate Estimation

End-to-end MATLAB framework for non-contact vital signs monitoring in children using 60 GHz FMCW radar. Features dual signal processing backends and validated on 25,000 radar frames.

MATLAB Radar Signal Processing MODWT
View on GitHub

Maze Solving with Deep Q-Network

Reinforcement learning agent using DQN algorithm to autonomously navigate and solve complex mazes. Features experience replay and target networks for stable learning.

Python PyTorch Reinforcement Learning DQN
View on GitHub

Air Compressor Monitor

Real-time predictive maintenance system for industrial air compressors using IoT sensors and Raspberry Pi. Monitors pressure, temperature, and humidity with live Streamlit dashboard.

Python Raspberry Pi IoT Streamlit
View on GitHub
Research

Vision Transformer for Micro-Doppler

Novel hybrid ViT architecture combining ResNet-SE blocks with transformer attention for radar target classification. Achieved 98% accuracy with 98% fewer parameters than VGG19.

Python PyTorch Vision Transformer Radar
Research

Coherent-On-Receive Marine Radar

Software-only DSP framework transforming non-coherent magnetron radar into coherent systems. Novel SVD-based algorithm achieving phase RMSE < 0.01 radians.

MATLAB DSP SVD Marine Radar
Research

Hybrid Statistic Subspace Decomposition

Advanced clutter suppression framework for non-coherent radar using fourth-order cumulants and SVD. Separates target returns from non-stationary clutter and leakage.

MATLAB Signal Processing Higher-Order Statistics
Research

Brain-Computer Interface ErrP Classification

Interpretable ML framework for Error-Related Potential classification from EEG. Achieved F1-Score > 0.88, outperforming complex CNN models with data-efficient approach.

Python Scikit-learn EEG PCA

Research Experience

Academic & Industry Research Positions

Graduate Research Assistant

May 2024 - Present

Advanced Radar Research Center | Norman, OK

  • Coherent-on-Receive Radar Processing: Architected a multi-stage DSP pipeline that transforms a non-coherent marine radar into a coherent-on-receive system. Developed novel SVD-based algorithm achieving phase RMSE < 0.01 radians.
  • Vision Transformer Architecture: Engineered hybrid deep learning model combining CNN-based ResNet-SE blocks with transformer attention. Enhanced with Fourier-based FAN layer, achieving 98% accuracy with 98% fewer parameters.
  • High-Fidelity Simulation: Built comprehensive MATLAB framework generating radar data for 13,000 avian species with complex flight dynamics and micro-Doppler signatures.
  • National Lab Collaboration: Partnering with PNNL and USGS to design buoy-based radar systems for biological target detection in maritime environments.
MATLAB Python PyTorch DSP Vision Transformers

Undergraduate Research Assistant

Jul 2022 - May 2024

Advanced Radar Research Center | Norman, OK

  • Automotive Sensor Fusion: Spearheaded integration of automotive-grade radar and video sensors using ROS, creating low-latency data fusion pipeline.
  • Radar Interference Mitigation: Developed algorithms to detect and mitigate inter-system radar interference, eliminating "ghost target" artifacts in real-time.
  • High-Power Radar Deployment: Deployed X-band Coherent-On-Receive radar achieving 94-113 dB channel isolation for precision geofencing and object tracking.
ROS C++ Sensor Fusion Real-Time Systems

Undergraduate Researcher

Jan 2022 - May 2022

3SigmaLab | Norman, OK

  • BCI Signal Processing: Engineered EEG signal processing pipeline for Error-Related Potential classification using band-pass filtering and feature extraction.
  • Interpretable Machine Learning: Developed data-efficient ML framework achieving F1-Score > 0.88, outperforming published CNN (86.1% accuracy).
  • Advanced Feature Engineering: Demonstrated 79.5% accuracy using only signal peak characteristics, establishing computationally efficient baseline.
Python Scikit-learn EEG Processing PCA

Skills & Technologies

Technologies I work with

Languages

Python TypeScript JavaScript MATLAB C/C++

Frameworks

React FastAPI TailwindCSS NumPy SciPy

Cloud & DevOps

Google Cloud Vertex AI Docker Confluent Kafka Vercel

Hardware & IoT

TI AWR1642 Raspberry Pi mmWave Radar Edge Computing

AI & ML

Machine Learning Deep Learning TensorFlow Scikit-learn

Biomedical

Signal Processing Vital Signs Radar Systems Healthcare Tech

Get In Touch

Let's discuss healthcare technology, radar signal processing, or AI in medicine

LinkedIn

Connect with me

GitHub

@ahmed100553

Location

Norman, OK