Quantum Measurement–Driven Security Lab (QKD + ODMR + QML)

QuantumSec-QKD-ODMR-QML is a research-grade, pure Python framework that unifies Information Security × Quantum Measurement × AI/QML into a reproducible pipeline. It demonstrates how quantum measurement statistics act as "security sensors" in two distinct domains: identifying eavesdroppers in Quantum Key Distribution (QKD) and detecting anomalies in Quantum Sensing (NV-ODMR).

English | 中文 (Chinese)

---

🇬🇧 English Description

🌟 Key Features

• Unified Security Pipeline (Task A & B):
  • Task A (QKD / BB84): Simulates the BB84 protocol to generate Quantum Bit Error Rate (QBER). We map QBER to Secret Key Rate (SKR) and use a Quantum Support Vector Machine (QSVM) to detect eavesdropping attacks (acting as an Intrusion Detection System).
  • Task B (NV-ODMR): Simulates NV Center ODMR spectra. We extract spectral features and utilize quantum kernels to detect magnetic field anomalies or sensor integrity issues.
• Multi-Backend Support: Implements identical quantum kernel feature maps across PennyLane, Qiskit Aer, Cirq, and CUDA-Q, with automated cross-validation to ensure numerical equivalence.
• Scientific Rigor: Built-in calculation of information-theoretic bounds (entropy, fidelity) and rigorous validation of physical monotonicity.

🚀 Quick Start

1. Installation

This repository follows a standard src/ layout. We recommend using a virtual environment (Conda or venv).

# Create and activate environment
conda create -n quantum-sec python=3.9
conda activate quantum-sec

# Install dependencies
pip install -r requirements.txt

Note: If you already have a quantum stack installed (PennyLane, Qiskit, etc.), you can skip full installation. CUDA-Q is optional if you lack a compatible environment.

2. Usage (Reproducible Scripts)

We provide a unified CLI to run experiments without modifying PYTHONPATH manually.

# Recommended: Run full experimental suite (BB84 + ODMR + Kernel Benchmarks)
./scripts/run_all.sh

# OR run individual modules via CLI
# 1. Run BB84 Simulation & Analysis
python -m qmsl.cli bb84 --backend numpy --seed 0 --out results

# 2. Run ODMR Simulation & Anomaly Detection
python -m qmsl.cli odmr --seed 0 --out results

# 3. Benchmark Quantum Kernels across Backends
python -m qmsl.cli kernel-bench --seed 0 --out results

Outputs:

• results/metrics.json: JSON report containing AUC, Accuracy, and SKR metrics.
• docs/figures/*.png: Generated plots (ROC curves, Confusion Matrices, Spectra).

📚 Repository Structure

• src/qmsl/datasets/: Data generators for BB84 (QBER stats) and ODMR (Lorentzian spectra).
• src/qmsl/kernels/: Quantum kernel implementations (PennyLane, Qiskit, Cirq, CUDA-Q).
• src/qmsl/models/: Classical baselines (SVM) and QSVM integration.
• src/qmsl/eval/: Evaluation metrics (AUC, Signal-to-Noise) and plotting tools.
• docs/: Supplementary documentation and figures.

📖 Citation

If you use this code in your research, please cite:

@misc{purohit2025qml,
  title = {Quantum Machine Learning for Quantum Key Distribution and Sensing},
  author = {Purohit, A. and Vyas, V.},
  year = {2025},
  note = {See refs/ for full bibliography}
}

---

🇨🇳 中文介绍 (Chinese Description)

🌟 核心特性

• 统一安全流水线 (任务 A & B):
  • 任务 A (QKD / BB84): 模拟 BB84 协议生成量子比特误码率 (QBER)。我们将 QBER 映射到信息论中的 安全密钥率 (SKR)，并使用量子支持向量机 (QSVM) 构建类似入侵检测系统 (IDS) 的防御机制。
  • 任务 B (NV-ODMR): 模拟 NV 色心 ODMR 光谱。提取光谱特征并利用量子核函数检测磁场异常或传感器完整性问题。
• 多后端支持: 在 PennyLane, Qiskit Aer, Cirq, 和 CUDA-Q 上实现了完全一致的量子核特征映射，并通过了严格的数值交叉验证。
• 科学严谨性: 内置香农熵、保真度等物理量的计算，确保模拟结果符合物理学理论边界。

🚀 快速开始

1. 安装环境

本项目采用标准的 python src/ 包结构。建议使用 Conda 或 venv 管理环境。

# 创建并激活环境
conda create -n quantum-sec python=3.9
conda activate quantum-sec

# 安装依赖
pip install -r requirements.txt

注意: 如果您已安装常用的量子计算库，可跳过完整安装。CUDA-Q 为可选依赖。

2. 运行实验

可以通过 CLI 轻松复现所有实验结果。

# 方案 A: 一键运行全套实验 (推荐)
./scripts/run_all.sh

# 方案 B: 单独运行模块
# 1. BB84 模拟与分析
python -m qmsl.cli bb84 --backend numpy --seed 0 --out results

# 2. ODMR 模拟与异常检测
python -m qmsl.cli odmr --seed 0 --out results

# 3. 跨后端量子核基准测试
python -m qmsl.cli kernel-bench --seed 0 --out results

输出产物:

• results/metrics.json: 包含 AUC、准确率 (Accuracy) 和安全密钥率等关键指标。
• docs/figures/*.png: 自动生成的 ROC 曲线、混淆矩阵和光谱分布图。

📂 项目结构

• src/qmsl/datasets/: BB84 (QBER 统计) 与 ODMR (洛伦兹光谱) 数据生成器。
• src/qmsl/kernels/: 适配多种后端的量子核函数实现 (PennyLane, Qiskit, Cirq, CUDA-Q)。
• src/qmsl/models/: 经典机器学习基线 (SVM) 与 QSVM 实现。
• src/qmsl/eval/: 评估指标 (AUC, SNR) 与绘图工具。
• docs/: 项目文档与图表。

📚 深度科普: 当信息安全遇见量子测量——基于 Purohit & Vyas (2025) 综述的 QML-IDS 复现探索

📜 许可证 (License)

本项目基于 MIT License 开源。