Available Data Sets

2016 Indoor/Basement RTI Data Set

This data set (in the basement directory) was collected in Fall 2016 for Neal Patwari's ECE 6960 Applications of Fading Channels course as Exam 2. It provides ground truth coordinates as Neal walked through the basement of his house in a piecewise linear path. The experiment for this exam was conducted in an approximately 8 meter by 8 meter section in his home. The sensors are placed inside the home. There were 10 sensors operating on 8 channels. The sensors are at the typical power outlet height. The accompanying code shows how to calculate the true coordinate of the person at any given time using the ground-truth files that are also included in the directory.

Mager Performance in Changing Environments Data Set

This data set is an extensive set of RSS Data in a house during device-free localization experiements conducted repeatedly, focusing on the changes in the RSS measurements and thus DFL performance after normal changes in the house. Five sets of experiments were performed. Each set includes results from several experiments, where data was gathered while the subject moved from one location to the next. Over the course of the experiments, objects were moved or altered in the house to determine the effect on the location accuracy of the fingerprint-based method. In the paper, the data is used to show the rate of performance degradation in fingerprint-based DFL.

Multi-target RTI Data Set

An RSS multichannel data set collected in an indoor area with multiple people moving in the area (at known positions).
  • Maurizio Bocca, Ossi Kaltiokallio, Neal Patwari, and Suresh Venkatasubramanian, Multiple target tracking with RF sensor networks, arXiv:1302.4720 [cs.NI], IEEE Transactions on Mobile Computing, appeared 25 July 2013, DOI: 10.1109/TMC.2013.92.

2013 Through-wall RTI Data Set

Data from two through-wall experiments under noisy conditions. TelosB sensors deployed outside of a part of a home, through exterior walls.

2010 Outdoor RTI Data Set

Data from outdoor RTI experiments consisting of 28 network nodes arranged in a 21 foot square. The nodes were TelosB devices. The data is in csv files.

Measured MIMO Channel Impulse Response Data

Measured MIMO channel impulse response (CIR) data which we used to experimentally evaluate MIMO location distinction.

Measured CIR in an indoor wireless sensor network

These measurements are of a set of complex impulse response measurements between each pair of nodes in a 44-node network in an indoor area.

Measured RSS and TOA in an indoor wireless network

Measurements of received signal strength (RSS) and time-of-arrival (TOA) in an indoor 44-node wireless sensor network, originally made publicly available in July 2005. This data set comprises pair-wise measurements in a 44-node network. The page also links to code for a MLE location estimation algorithm, and a function to calculate the Cramer-Rao Lower Bound for location estimation.

Available Tools

RF Respiration Monitoring Sleep-Wake Dataset

A joint study of the SPAN Lab and the Sleep-Wake Center at the University of Utah, led by Peter Hillyard, collected RF sensing data during patients' sleep studies at the Sleep-Wake Center. Data from 20 patients were collected. For each patient, we simultaneously collected data from ultra-wideband impulse response (UWB-IR) transceivers, from WiFi tranceivers with CSI-Tool, from RSSI from Zigbee transceivers, and from the fine-grained RSS from the Sub-dB system tranceivers. Each sleep study is about 8 hours long, and the Sleep-Wake Center data includes polysomnograph data indicating ground-truth for the sleep state, breathing rate, as well as other sensor data.

PRISMS Sensor Network Architecture

The PRISMS project provides software for the deployment of sensor networks in homes, particularly for researchers who want to run human subjects studies in which the activities and state of a home is recorded over a long duration of time. The Architecture builds upon Home Assistant, which provides a sensor data bus for a wide variety of commercial IoT sensors and devices, by adding support for a remote InfuxDB database instance, for custom WiFi sensors, and for highly robust protocols for data transfer to ensure near-zero data loss. The repositories also include code and instructions on how to build the Utah Modified Dylos Sensor (UMDS) which measures particulate matter, temerature, and humidity, and uses a Beaglebone to upload the data in real time via WiFi.

Neal Patwari Multichannel RTI Algorithms

This code implements multichannel RTI using both attenuation and variance measurements. It follows the multi-channel attenuation-based (traditional) RTI. This is the method published in MASS 2012: O. Kaltiokallio, M. Bocca, and N. Patwari, Enhancing the accuracy of radio tomographic imaging using channel diversity, 9th IEEE International Conference on Mobile Ad hoc and Sensor Systems, October 8-11, 2012. The code also implements multi-channel variance-based RTI. This is the straightforward extension of VRTI to multiple channels. Single-channel VRTI is described in: J. Wilson and N. Patwari, See Through Walls: Motion Tracking Using Variance-Based Radio Tomography Networks, IEEE Transactions on Mobile Computing, vol. 10, no. 5, pp. 612-621, May 2011.

SPAN Lab Github Repositories

The SPAN Lab public repositories for software and hardware for our recent projects.

Multi-Spin Communication Protocol

Multichannel switching and TX/RX protocol for a network of n RF sensors to be able to measure RSS on multiple frequency channels without self-interference. Developed by Maurizio Bocca for the TI CC253x SoC.

SPAN Lab Beamer Template

This is a Beamer template with the SPAN and U of Utah logo and a navigation bar on top. Please replace the .pdf file extension with .zip and read the sample.pdf file before using the template.