WiROS: WiFi sensing toolbox for robotics

WiROS is a plug-and-play WiFi sensing toolbox allowing researchers to access coarse grained WiFi signal strength (RSSI), fine grained WiFi channel state information (CSI), and other MAC-layer information (device address, packet id’s or frequency-channel information). Additionally, WiROS open-sources state of-art algorithms to calibration and process WiFi measurements to furnish accurate bearing information for received WiFi signals.

This is an index repository to access the following components of WiROS

1. CSI Node - Extends the Nexmon CSI toolkit[1] to provide a ROS overlay.
2. Processing Node - Provide calibration and post-processing of WiFi CSI measurments. Open-sources mulitple state-of-art bearing extraction algorithms to measure both the angle of arrival (at the receiver) and angle of departure (from the transmitter) of the WiFi signal.
3. RF Messages format - Custom ROS messages to structure WiFi measurements information.

Overview of Features

1. Easily integrate WiFi channel state measurements, received signal strength and other WiFi MAC-layer information into your robot sensor stack.
2. Exposes all relevant measurements as accessible ROS topics. See rf_msgs for more details.
3. Builds a framework for hassle-free wireless calibration of wireless sensors.
4. Provides visualizations for WiFi signals which are helpful for algorithm paramter tuning and debugging
5. Open-sources implementation of various state-of-art WiFi processing algorithms[2, 3, 4].

Getting Started

To get started with WiROS, clone these repositories into the src folder of your catkin workspace, and follow the README in the CSI Node to configure your hardware.

Example usage of WiROS

WiROS can be easily leveraged to incorporate WiFi sensors to solve many applicable problems in robotics. We provide the following sample use-cases:

1. Kidnapped Robot Problem: A lost robot in an indoor envrionment can be conveniently localized using WiROS. Given a prior map of the existing Access points and additional details of their antenna geometry, the robot's location can be triangulated in a space.
2. Correct for Robot Location Drift: WiFi measurements can be additionally fused with Camera and odometry measurements to more accurately correct for sensor drifts and resolve ambiguities arising from perceptual aliasing in indoor environment[3].
3. IoT device localiztion: Often IoT devices are hard to localize visually. However, we can leverage WiFi-based bearing measurements to trinagulate their position in the envrionment. This can be useful for both IoT device management or to ensure security/privacy of users in a space.

Citations

1. Blanco, Alejandro, et al. "Accurate ubiquitous localization with off-the-shelf ieee 802.11 ac devices." The 19th Annual International Conference on Mobile Systems, Applications, and Services (MobiSys 2021). 2021.
2. Kotaru, Manikanta, et al. "Spotfi: Decimeter level localization using wifi." Proceedings of the 2015 ACM Conference on Special Interest Group on Data Communication. 2015.
3. Arun, Aditya, et al. "ViWiD: Leveraging WiFi for Robust and Resource-Efficient SLAM." arXiv preprint arXiv:2209.08091 (2022).
4. Schmidt, Ralph. "Multiple emitter location and signal parameter estimation." IEEE transactions on antennas and propagation 34.3 (1986): 276-280.