Software
This page highlights research software and experimental platforms I have developed for quantum computing, quantum machine learning, quantum sensing, and distributed learning.
CMAB-ZNE
Adaptive Zero-Noise Extrapolation on Real Quantum Hardware — 2026
A contextual multi-armed-bandit framework for adaptively selecting zero-noise extrapolation strategies using accuracy, uncertainty, and execution-cost signals.
Hardware: Rigetti Cepheus-1-108Q through Amazon Braket
Scale: 240 real-QPU tasks
Capabilities: circuit-folding strategies, Pauli-Z expectation estimation, uncertainty analysis, runtime tracking, and hardware-cost evaluation.
SimQFL
Quantum Federated Learning Simulator — 2025–2026
SimQFL is a client-server simulation framework for heterogeneous and personalized quantum federated learning with QPU-level noise models, client participation controls, aggregation, and real-time experiment visualization.
Key capabilities
- Configurable quantum clients
- Non-IID and client-level heterogeneity
- QPU-specific noise models
- Personalized and noise-aware QFL
- Configurable client participation
- Real-time training and convergence visualization
QuNoise
Quantum Noise Simulator & Visual Analytics Platform — 2026
QuNoise is a research platform for analyzing how realistic quantum noise reshapes circuit outputs, model behavior, and latent representations.
Noise models
- Depolarizing noise
- Amplitude damping
- Phase damping
- Readout errors
Mitigation methods
- Zero-noise extrapolation
- Probabilistic error cancellation
- Measurement error mitigation
- Clifford data regression
The platform also supports fidelity and feature-map visualization for interpreting noise-induced degradation.
QuSenseSim
Variational Quantum Sensing Simulator — 2025–2026
QuSenseSim is an interactive multi-sensor quantum sensing framework supporting heterogeneous sensing conditions, quantum noise, mitigation workflows, Fisher Information analysis, and Cramér–Rao Bound evaluation.
Scalable Quantum-Kernel Malware Classification
KAUST Quantum Security Research — 2026
Developed an end-to-end hybrid quantum-classical pipeline for large-scale malware-family classification using supervised feature projection, fidelity-based quantum kernels, Nyström approximation, and multiclass ridge classification.
Evaluation scale: 18,836 PE samples across 23 labels
Validation: repeated stratified splits, strong classical baselines, ablation studies, statistical testing, and real-device validation on a Rigetti superconducting QPU through Amazon Braket.
The associated first-author manuscript is currently under revision at IEEE Transactions on Dependable and Secure Computing (TDSC).