Update README.md
Browse files
README.md
CHANGED
@@ -49,31 +49,27 @@ language:
|
|
49 |
tags:
|
50 |
- wireless
|
51 |
---
|
|
|
|
|
|
|
52 |
# πΆ Beam-Level (5G) Time-Series Dataset
|
53 |
|
54 |
-
This dataset introduces a **novel multivariate time series** specifically curated to support research
|
55 |
|
56 |
<p align="center">
|
57 |
-
|
58 |
</p>
|
59 |
|
60 |
-
Precise forecasting of network traffic is critical for
|
61 |
-
- Optimizing **network management**
|
62 |
-
- Enhancing **resource allocation efficiency**
|
63 |
-
|
64 |
-
This task is of both **practical and theoretical importance** to researchers in networking and machine learning, offering a strong benchmark for state-of-the-art (SOTA) time series models.
|
65 |
|
66 |
---
|
67 |
|
68 |
## π Dataset Overview
|
69 |
|
70 |
-
|
71 |
-
|
72 |
-
|
73 |
-
|
74 |
-
- **Frequency:** Hourly
|
75 |
-
- **Duration:** 5 weeks + 2 target weeks
|
76 |
-
- **Total Hours (per beam):** Up to 840 (train) and 1176 (total)
|
77 |
|
78 |
---
|
79 |
|
@@ -81,61 +77,48 @@ This task is of both **practical and theoretical importance** to researchers in
|
|
81 |
|
82 |
### ποΈββοΈ Training Set (Weeks 0β5)
|
83 |
|
84 |
-
| File Name
|
85 |
-
|
86 |
-
| `DLThpVol_train_0w-5w.csv`
|
87 |
-
| `DLThpTime_train_0w-5w.csv`
|
88 |
-
| `DLPRB_train_0w-5w.csv`
|
89 |
-
| `MR_number_train_0w-5w.csv`
|
90 |
|
91 |
### π― Forecast Targets
|
92 |
|
93 |
#### π 6th Week (Week 5β6)
|
94 |
|
95 |
-
| File Name
|
96 |
-
|
97 |
-
| `DLThpVol_test_5w-6w.csv`
|
98 |
-
| `DLThpTime_test_5w-6w.csv`
|
99 |
-
| `DLPRB_test_5w-6w.csv`
|
100 |
-
| `MR_number_test_5w-6w.csv`
|
101 |
|
102 |
#### π 11th Week (Week 10β11)
|
103 |
|
104 |
-
| File Name
|
105 |
-
|
106 |
-
| `DLThpVol_test_10w-11w.csv`
|
107 |
-
| `DLThpTime_test_10w-11w.csv`
|
108 |
-
| `DLPRB_test_10w-11w.csv`
|
109 |
-
| `MR_number_test_10w-11w.csv`
|
110 |
-
|
111 |
|
112 |
---
|
113 |
|
114 |
## π§ͺ Dataset Splits
|
115 |
|
116 |
<p align="center">
|
117 |
-
|
118 |
</p>
|
119 |
|
120 |
-
|
121 |
-
- **Forecast Targets:**
|
122 |
-
- **Week 6** (immediate future)
|
123 |
-
- **Week 11** (long-term future)
|
124 |
|
125 |
---
|
126 |
|
127 |
## π Data Format
|
128 |
|
129 |
-
Each CSV file
|
130 |
-
|
131 |
-
- **`Time` column**:
|
132 |
-
- Ranges:
|
133 |
-
- `0β839` for training (weeks 1β6)
|
134 |
-
- `0β167` for week 6
|
135 |
-
- `168β335` for week 11
|
136 |
-
|
137 |
-
- **Beam columns (`0_0_0`, ..., `29_2_31`)**:
|
138 |
-
- Each uniquely identifies one of the **2,880 beams** across 30 base stations.
|
139 |
|
140 |
---
|
141 |
|
@@ -143,13 +126,10 @@ Each CSV file follows this structure:
|
|
143 |
|
144 |
If you use this dataset in your research, please cite:
|
145 |
|
146 |
-
> **L. Fechete et al.**,
|
147 |
-
> *Goal-Oriented Time-Series Forecasting: Foundation Framework Design*,
|
148 |
-
> arXiv:2504.17493 (2025)
|
149 |
|
150 |
---
|
151 |
|
152 |
## π Code Repository
|
153 |
|
154 |
-
The official codebase for working with this dataset is available here:
|
155 |
-
π [https://github.com/netop-team/gotsf](https://github.com/netop-team/gotsf)
|
|
|
49 |
tags:
|
50 |
- wireless
|
51 |
---
|
52 |
+
|
53 |
+
|
54 |
+
|
55 |
# πΆ Beam-Level (5G) Time-Series Dataset
|
56 |
|
57 |
+
This dataset introduces a **novel multivariate time series** specifically curated to support research in enabling **accurate prediction of KPIs** across communication networks, as illustrated below:
|
58 |
|
59 |
<p align="center">
|
60 |
+
Β <img src="images/network.png" alt="Base station, cells, and beams" />
|
61 |
</p>
|
62 |
|
63 |
+
Precise forecasting of network traffic is critical for optimizing **network management** and enhancing **resource allocation efficiency**. This task is of both **practical and theoretical importance** to researchers in networking and machine learning, offering a strong benchmark for state-of-the-art (SOTA) time series models.
|
|
|
|
|
|
|
|
|
64 |
|
65 |
---
|
66 |
|
67 |
## π Dataset Overview
|
68 |
|
69 |
+
The dataset comprises:
|
70 |
+
* **2,880 Beams** across 30 Base Stations (3 Cells per Station, 32 Beams per Cell).
|
71 |
+
* **Hourly frequency**.
|
72 |
+
* **Duration:** 5 weeks + 2 target weeks, totaling up to 840 training hours and 1176 total hours per beam.
|
|
|
|
|
|
|
73 |
|
74 |
---
|
75 |
|
|
|
77 |
|
78 |
### ποΈββοΈ Training Set (Weeks 0β5)
|
79 |
|
80 |
+
| File Name | Metric |
|
81 |
+
|---|---|
|
82 |
+
| `DLThpVol_train_0w-5w.csv` | Downlink throughput volume |
|
83 |
+
| `DLThpTime_train_0w-5w.csv` | Throughput transmission time |
|
84 |
+
| `DLPRB_train_0w-5w.csv` | PRB (Physical Resource Block) usage |
|
85 |
+
| `MR_number_train_0w-5w.csv` | User count (Measurement Reports) |
|
86 |
|
87 |
### π― Forecast Targets
|
88 |
|
89 |
#### π 6th Week (Week 5β6)
|
90 |
|
91 |
+
| File Name | Metric |
|
92 |
+
|---|---|
|
93 |
+
| `DLThpVol_test_5w-6w.csv` | Downlink throughput volume |
|
94 |
+
| `DLThpTime_test_5w-6w.csv` | Throughput transmission time |
|
95 |
+
| `DLPRB_test_5w-6w.csv` | PRB usage |
|
96 |
+
| `MR_number_test_5w-6w.csv` | User count |
|
97 |
|
98 |
#### π 11th Week (Week 10β11)
|
99 |
|
100 |
+
| File Name | Metric |
|
101 |
+
|---|---|
|
102 |
+
| `DLThpVol_test_10w-11w.csv` | Downlink throughput volume |
|
103 |
+
| `DLThpTime_test_10w-11w.csv` | Throughput transmission time |
|
104 |
+
| `DLPRB_test_10w-11w.csv` | PRB usage |
|
105 |
+
| `MR_number_test_10w-11w.csv` | User count |
|
|
|
106 |
|
107 |
---
|
108 |
|
109 |
## π§ͺ Dataset Splits
|
110 |
|
111 |
<p align="center">
|
112 |
+
Β <img src="images/dataset_split.png" alt="Dataset train/forecast split" />
|
113 |
</p>
|
114 |
|
115 |
+
The dataset is split into a **Training Set** (first 5 weeks) and **Forecast Targets** for Week 6 (immediate future) and Week 11 (long-term future).
|
|
|
|
|
|
|
116 |
|
117 |
---
|
118 |
|
119 |
## π Data Format
|
120 |
|
121 |
+
Each CSV file contains a `Time` column and multiple beam columns (e.g., `0_0_0` to `29_2_31`). The `Time` column ranges from `0β839` for training (weeks 1β6), `0β167` for week 6, and `168β335` for week 11. Each beam column uniquely identifies one of the **2,880 beams** across 30 base stations.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
122 |
|
123 |
---
|
124 |
|
|
|
126 |
|
127 |
If you use this dataset in your research, please cite:
|
128 |
|
129 |
+
> **L. Fechete et al.**, *Goal-Oriented Time-Series Forecasting: Foundation Framework Design*, arXiv:2504.17493 (2025)
|
|
|
|
|
130 |
|
131 |
---
|
132 |
|
133 |
## π Code Repository
|
134 |
|
135 |
+
The official codebase for working with this dataset is available here: π [https://github.com/netop-team/gotsf](https://github.com/netop-team/gotsf)
|
|