File size: 2,000 Bytes
d257a54
 
 
 
591f721
2a80a7a
 
 
 
 
d257a54
b5dfe52
d5e796d
f7e9d70
d5e796d
 
 
591f721
87351e9
d5e796d
591f721
d5e796d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
591f721
d5e796d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cc38892
 
 
 
 
 
591f721
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
---
license: apache-2.0
language:
- en
pretty_name: satcom_qa
tags:
- satellite
- communications
size_categories:
- n<1K
---


# 🛰️ SATCOM Question-Answer Dataset

## Dataset Summary

**SATCOM-QA** is a **human-generated evaluation dataset** designed for testing LLM reasoning in Satellite Communications (SatCom).
It contains around 1,000 open-ended questions, of which about 30% involve mathematical or quantitative SatCom reasoning (e.g., link budget calculations, path loss, beamforming, SNR, orbital geometry).

The dataset is designed for evaluation purpose.

---

## Key Features

* **Human-written** questions & answers
* **Open-ended** free-form responses
* **Covers core SatCom/NTN domains**, including:

  * Link Budget
  * Propagation & Path Loss
  * Frequency bands
  * Beamforming & antennas
  * Multi-layer NTNs (satellites, HAPS, UAVs)
  * 6G integration
  * Localization with NTNs
* 30% math-focused problems
* Includes **reference links** to the answers provided
* No synthetic content, no model-generated text

---

## Intended Use

* Benchmarking LLMs on **technical SatCom reasoning**
* Testing domain-specific expertise for

  * NTN/6G research
  * Satellite engineering
  * Telecommunications academia/industry
* Evaluating mathematical reasoning in real-world SatCom tasks


---

## Data Structure

### Fields

* **question** *(string)*
* **answer** *(string, human-written)*
* **topic** *(string: e.g., Link Budget, Beamforming, NTNs, etc.)*
* **sources** *(list of URLs — academic references)*
* **math_required** *(bool)*

---

## Source Material

Questions are based on **public academic literature**, including (examples):

* Multi-layer NTN architectures (ResearchGate)
* Integrated 6G TN/NTN (ResearchGate)
* NTN localization (arXiv)
* Link budget fundamentals (open-access journals)

## Example  
```python
from datasets import load_dataset
ds = load_dataset("esa-sceva/satcom-qa")
test_ds = ds["train"]
print(test_ds[0]["question"])
print(test_ds[0]["answer"])