File size: 4,871 Bytes
84111ef
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ecad033
 
 
 
 
 
84111ef
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b1b5f48
84111ef
 
 
 
 
 
 
 
 
 
b1b5f48
84111ef
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
# SC2026: Screaming Channel 2026 Dataset

SC2026 is a large-scale public dataset for **Far-Field Electromagnetic Side-Channel Attacks (FEM-SCAs)**, also known as **Screaming Channel Attacks**.  
The dataset is designed to support **systematic, realistic, and reproducible evaluation** of long-range EM side-channel attacks under diverse conditions.

It contains far-field EM traces captured from **Bluetooth-enabled IoT devices** executing multiple cryptographic algorithms, across **different distances and physical barriers**.

📌 This dataset accompanies the paper:

> **A Systematic Far Field EM Side-Channel Evaluation and Introduction of SC2026 Dataset**  
> *Not yet published*

---

## 🔍 Key Features

- 📡 **Far-field EM leakage** (wireless, over-the-air)
- 🔐 **Multiple cryptographic algorithms**
  - AES
  - SM4
  - CRYSTALS-Kyber
- 📏 **Multiple attack distances**
  - 0 m (coaxial cable baseline)
  - 3 m / 6 m / 9 m / 12 m / 15 m
  - 30 m (open-field scenario)
- 🧱 **Physical barriers**
  - Plastic
  - Wood
  - Metal
- 🧠 **Supports both classical and deep-learning SCAs**
  - CPA, Template Attack
  - CNN / MLP / Transformer / Domain Adaptation models
- 📂 **NumPy format (.npy)** for efficient loading and training

---

## 📦 Dataset Structure

To balance accessibility and research utility, this dataset is provided in two formats:

👁️ Preview (dataset_preview.parquet): A small sample (first 100 traces) containing plaintexts, keys, and trace snippets. Use the "Viewer" tab above to explore the data schema interactively.

⬇️ Full Dataset (.npy): The complete raw traces and metadata are stored in the data/ and metadata/ directories as NumPy files. Researchers should download these files for training and analysis.

The SC2026 dataset is divided into **two main subsets**:

### 1️⃣ Profiling Set (High-Quality Reference)

Used for building leakage models.

- Capture method: **Coaxial cable**
- Purpose: Profiling / training
- Traces per algorithm: **40,000**
- Each trace:
  - Averaged over **100 repeated measurements**
  - Plus a **corresponding single-trace (non-averaged) version**
- Algorithms & labels:
  - **AES / SM4**: plaintext & key
  - **Kyber**: message bit

📌 This design enables:
- Profiling attacks
- Transfer learning
- Domain adaptation
- Denoising and robustness studies

---

### 2️⃣ Testing Set (Realistic Attack Scenarios)

Used for evaluating attack performance under realistic conditions.

- Capture method: **Over-the-air (far-field EM)**
- Traces per scenario: **5,000**
- No averaging, no repetition
- Covers:
  - Different distances
  - Different environments
  - Different physical barriers

Each scenario is stored as an independent `.npy` file for clarity and reproducibility.

---

## 📁 Example Directory Layout

```text
SC2026/
├── profiling/
│   ├── AES/
│   │   ├── traces_avg.npy
│   │   ├── traces_single.npy
│   │   ├── plaintext.npy
│   │   └── key.npy
│   ├── SM4/
│   └── Kyber/

├── testing/
│   ├── distance_3m/
│   │   ├── AES.npy
│   │   ├── SM4.npy
│   │   └── Kyber.npy
│   ├── distance_15m/
│   ├── distance_30m/
│   ├── barrier_plastic/
│   ├── barrier_wood/
│   └── barrier_metal/

└── README.md
```
---

## How to Download
Full dataset:  
```python 
from huggingface_hub import snapshot_download
snapshot_download(repo_id="SCA-HNUST/SC2026", repo_type="dataset", local_dir="<download_path>")
```

One sub-dataset of choice:
```python 
from huggingface_hub import snapshot_download
snapshot_download(repo_id="SCA-HNUST/SC2026",repo_type="dataset",local_dir="<download_path>",allow_patterns="<sub_dataset>/*")
```
Replace `<sub_dataset>` with 'AES','SM4','CRYSTALS-Kyber'.

---


## 🧪 Experimental Setup Overview

- **Target device**: Nordic nRF52 DK (nRF52832)
- **Radio**: Bluetooth Low Energy (2.4 GHz)

**Receiver**:
- Ettus N210 USRP
- SBX RF daughterboard
- 24 dBi directional antenna

**Signal acquisition**:
- Sampling rate: 5 MHz
- Center frequency: 2.272 GHz

**Environments**:
- Indoor: office corridor
- Outdoor: open field

Both indoor (office corridor) and outdoor (open field) environments are included.

---
## 📜 Citation

If you use SC2026 in your research, please cite:

```bibtex
@article{wang2025sc2026,
  title={A Systematic Far Field EM Side-Channel Evaluation and Introduction of SC2026 Dataset},
  author={Wang, Huanyu and Wang, Xiaoxia and Ge, Kaiqiang and Yao, Jinjie and Tan, Xinyan and Wang, Junnian},
  journal={-----------------------------------},
  year={2025}
}
```
---

## 🤝 Contact

For questions, feedback, or collaboration, please contact:

Huanyu Wang 

School of Computer Science and Engineering  
Hunan University of Science and Technology