dichasus-0c5x Dataset: Outdoor - Street on University Campus
Outdoor, mostly line-of-sight dataset with transmitter on a street between two tall buildings on the university campus
50.000 MHz
Signal Bandwidth
1024
OFDM Subcarriers
27922
Data Points
3389.1 s
Total Duration
7.3 GB
Total Download Size
32
Number of Antennas
Outdoor
Type of Environment
1.272000 GHz
Carrier Frequency
Co-Located
Antenna Setup
3D GNSS
Position-Tagged
3D GNSS
Position-Tagged
Experiment Setup
Outdoor environment in which measurements were carried out.
Locations of data points in dichasus-0c64 dataset visualized in a map.
32-antenna array mounted just outside a window.
Data Analysis
Mean received signal-to-noise ratio by position
SNR by receive antenna index
Warnings
The GNSS coordinates in this dataset may be inaccurate (several meters of drift), regardless of the reported standard deviations. They are suitable for visualization purposes, but not for localization experiments.
In particular, do not rely on the GNSS height coordinate, it is inaccurate. You can assume a constant transmitter height of approx. 518m above sea level.
GNSS Ground Truth Map
Antenna Configuration
Antenna 1: Main Array
This array has a vertical spacing of 0.118m and a
horizontal spacing of 0.118m.
In the dataset's longitude / latitude coordinate system, it is located at N 48.747036°, E 9.105075°, with the
center 10.1m above the ground, which corresponds to 527m above sea level.
The antenna points in a direction that is characterized by a true-north based
azimuth angle of 94° a downwards tilt of 26.5°.
Antenna Channel Assignments
| 28 | 5 | 10 | 14 | 6 | 2 | 16 | 18 |
| 19 | 4 | 23 | 17 | 20 | 11 | 9 | 27 |
| 31 | 29 | 0 | 13 | 1 | 12 | 3 | 7 |
| 30 | 26 | 21 | 25 | 22 | 15 | 24 | 8 |
Python: Import with TensorFlow
#!/usr/bin/env python3
import tensorflow as tf
raw_dataset = tf.data.TFRecordDataset(["tfrecords/dichasus-0c53.tfrecords", "tfrecords/dichasus-0c54.tfrecords", "tfrecords/dichasus-0c55.tfrecords", "tfrecords/dichasus-0c56.tfrecords", "tfrecords/dichasus-0c57.tfrecords", "tfrecords/dichasus-0c58.tfrecords", "tfrecords/dichasus-0c59.tfrecords", "tfrecords/dichasus-0c5a.tfrecords", "tfrecords/dichasus-0c5b.tfrecords", "tfrecords/dichasus-0c5c.tfrecords", "tfrecords/dichasus-0c5d.tfrecords", "tfrecords/dichasus-0c5e.tfrecords", "tfrecords/dichasus-0c5f.tfrecords", "tfrecords/dichasus-0c60.tfrecords", "tfrecords/dichasus-0c61.tfrecords", "tfrecords/dichasus-0c62.tfrecords", "tfrecords/dichasus-0c63.tfrecords", "tfrecords/dichasus-0c64.tfrecords"])
feature_description = {
"cfo": tf.io.FixedLenFeature([], tf.string, default_value = ''),
"csi": tf.io.FixedLenFeature([], tf.string, default_value = ''),
"pos-gnss": tf.io.FixedLenFeature([], tf.string, default_value = ''),
"snr": tf.io.FixedLenFeature([], tf.string, default_value = ''),
"stddev-gnss": tf.io.FixedLenFeature([], tf.string, default_value = ''),
"time": tf.io.FixedLenFeature([], tf.float32, default_value = 0),
}
def record_parse_function(proto):
record = tf.io.parse_single_example(proto, feature_description)
# Measured carrier frequency offset between MOBTX and each receive antenna.
cfo = tf.ensure_shape(tf.io.parse_tensor(record["cfo"], out_type = tf.float32), (32))
# Channel coefficients for all antennas, over all subcarriers, real and imaginary parts
csi = tf.ensure_shape(tf.io.parse_tensor(record["csi"], out_type = tf.float32), (32, 1024, 2))
# Position of transmitter determined by GNSS, in latitude / longitude / height above sea level
pos_gnss = tf.ensure_shape(tf.io.parse_tensor(record["pos-gnss"], out_type = tf.float64), (3))
# Signal-to-Noise ratio estimates for all antennas
snr = tf.ensure_shape(tf.io.parse_tensor(record["snr"], out_type = tf.float32), (32))
# Standard deviation of position label estimate, in meters
stddev_gnss = tf.ensure_shape(tf.io.parse_tensor(record["stddev-gnss"], out_type = tf.float32), (3))
# Timestamp since start of measurement campaign, in seconds
time = tf.ensure_shape(record["time"], ())
return cfo, csi, pos_gnss, snr, stddev_gnss, time
dataset = raw_dataset.map(record_parse_function, num_parallel_calls = tf.data.experimental.AUTOTUNE)
# Optional: Cache dataset in RAM for faster training
dataset = dataset.cache()Reference Channel Compensation
For this dataset, we are able to provide estimated antenna-specific carrier phase and sampling time offsets. These offsets occur due to the fact that the reference transmitter channel is not perfectly frequency-flat. To learn more about why these offsets occur and about their compensation, visit our offset calibration tutorial on this topic. Note that the estimates provided here are "best-effort" calculations. The phase and time offsets between antennas in the same array are usually very accurate, but for antennas that are spaced far apart, the results may be less precise. The offset are constant over the complete dataset.How to Cite
Please refer to the home page for information on how to cite any of our datasets in your research. For this dataset in particular, you may use the following BibTeX:
@data{dataset-dichasus-0c5x,
author = {Euchner, Florian and Gauger, Marc},
publisher = {DaRUS},
title = {{CSI Dataset dichasus-0c5x: Outdoor - Street on University Campus}},
doi = {doi:10.18419/darus-2186},
url = {https://doi.org/doi:10.18419/darus-2186},
year = {2021}
}Download
This dataset consists of 18 files. Descriptions of these files as well as download links are provided below.
dichasus-0c53
dichasus-0c54
dichasus-0c55
dichasus-0c56
dichasus-0c57
dichasus-0c58
dichasus-0c59
dichasus-0c5a
dichasus-0c5b
dichasus-0c5c
dichasus-0c5d
dichasus-0c5e
dichasus-0c5f
dichasus-0c60
dichasus-0c61
dichasus-0c62
dichasus-0c63
Derived Channel Statistics
Channel statistics such as delay spread, k-Factor and path loss exponent are a good way to characterize a wireless channel measurement and to parametrize a channel model. Using estimation algorithms contributed by Janina Sanzi, we automatically extract the following channel statistics from the measured datasets:
RMS Delay Spread
The delay spread of a wireless channel is inversely proportional to the channel's coherence bandwidth and indicates how "spread out" the lengths of the various multipath propagation paths are. For every datapoint, the delay spread can be characterized by its root mean square value and the resulting delay spreads can be plotted over the measurement area:
Rician K-Factor
The Rician K-factor is defined as the power ratio between dominant and diffuse component, usually expressed in decibels. We estimate the K-factor with a moment-method based on the distribution of of channel coefficient powers. The resulting K-Factors be plotted over the measurement area: