conferences | speakers | series

Cooperative Perception in Future Cars using GNU Radio

home

Cooperative Perception in Future Cars using GNU Radio
FOSDEM 2020

Title: Cooperative Perception in Future Cars using GNU Radio

Speaker: Augusto Vega, IBM Research (NY, USA)

Abstract: The phenomenon of self-driving (autonomous) vehicles is a symbol of the grand re-emergence of artificial intelligence and robotics as a promising technology. The most general model of future vehicular transportation is that of artificially intelligent, connected, autonomous vehicles (CAVs) [1].

In this talk, we present a representative open-source application for CAVs operating as a collaborative swarm and communicating via GNU Radio. The application, called ERA [2], incorporates local sensing, creation of occupancy grid maps, vehicle-to-vehicle (V2V) communication of grid maps between neighboring vehicles using GNU Radio-based dedicated short-range communication (DSRC), and map fusion to create a joint higher-accuracy grid map [3]. Specifically, each vehicle in ERA uses its onboard sensors to generate local occupancy grid maps, which it communicates to other nearby vehicles using DSRC. When a vehicle receives occupancy maps from nearby cars, it merges the received ones with the locally-generated occupancy maps, expanding the scope and increasing the accuracy of this vehicle's perception. The DSRC transceiver adopted in ERA is an open-source GNU Radio implementation of the IEEE 802.11p standard by Bastian Bloessl [4]; while perception and map creation is implemented using ROS (Robot Operating System) [5]. We created an appropriate software interface between GNU Radio and ROS which enables proper execution and interaction of both frameworks.

In addition to presenting a deep dive into ERA's code, we will also show performance analysis results of ERA (including its GNU Radio components) and discuss potential acceleration opportunities for performance and efficiency improvement -- including optimizations of Viterbi decoding and complex exponential through hardware acceleration. We believe that ERA can help to fill the gap between the fast-growing CAV R&D domain and GNU Radio, specifically when it comes to the wireless communication aspect of future vehicles.

[1] A. Vega, A. Buyuktosunoglu, P. Bose, “Towards "Smarter" Vehicles Through Cloud-Backed Swarm Cognition,” Intelligent Vehicles Symposium 2018: 1079-1086.

[2] ERA. URL: https://github.com/IBM/era

[3] E. Sisbot, A. Vega, A. Paidimarri, J. Wellman, A. Buyuktosunoglu, P. Bose, D. Trilla, “Multi-Vehicle Map Fusion using GNU Radio,” Proceedings of The GNU Radio Conference 2019, 4(1).

[4] B. Bloessl, “IEEE 802.11 a/g/p transceiver for GNU radio,” URL: https://github.com/bastibl/gr-ieee802-11

[5] ROS. URL: https://www.ros.org

Desired slot time: 30 mins

Speakers: Augusto Vega