Publications
1.
Delooz, Q.; Maksimovski, D.; Festag, A.; Facchi, C.
Design and Evaluation of V2X Communication Protocols for Cooperatively Interacting Automobiles Book Section
In: Stiller, Christoph; Althoff, Matthias; Burger, Christoph; Deml, Barbara; Eckstein, Lutz; Flemisch, Frank (Ed.): Cooperatively Interacting Vehicles: Methods and Effects of Automated Cooperation in Traffic, pp. 159–199, Springer International Publishing, Cham, 2024.
Abstract | Links | BibTeX | Tags: Cooperative ITS, Intelligent Transport Systems, maneuver coordination, V2X communication, vehicle automation
@incollection{Delooz:CoInCar:2024,
title = {Design and Evaluation of V2X Communication Protocols for Cooperatively Interacting Automobiles},
author = {Q. Delooz and D. Maksimovski and A. Festag and C. Facchi},
editor = {Christoph Stiller and Matthias Althoff and Christoph Burger and Barbara Deml and Lutz Eckstein and Frank Flemisch},
url = {https://link.springer.com/chapter/10.1007/978-3-031-60494-2_6},
doi = {10.1007/978-3-031-60494-2_6},
year = {2024},
date = {2024-08-12},
urldate = {2024-08-03},
booktitle = {Cooperatively Interacting Vehicles: Methods and Effects of Automated Cooperation in Traffic},
pages = {159–199},
publisher = {Springer International Publishing},
address = {Cham},
abstract = {This chapter studies two key communication services for the support of cooperative driving capabilities using Vehicle-to-Everything (V2X) communications: sensor data sharing and maneuver coordination. Based on the current state of the art in research and pre-standardization of V2X communications, we enhance the protocol design for both services and assess their performance by discrete-event simulations in highway and city scenarios. The first part of this chapter addresses the performance improvement of sensor data sharing by two complementary strategies. The shared sensor data are adapted to the available resources on the used channel. Furthermore, the redundancy of the transmitted information is reduced to lower the load on the wireless channel, whereas several approaches are proposed and assessed. The second part of the chapter analyzes cooperative maneuver coordination protocols. We propose a distributed approach based on the explicit exchange of V2X messages, which introduces priorities in maneuver coordination and studies several communication patterns for the negotiation and coordination of maneuvers among two and more vehicles. The results demonstrate the potential of V2X communications for automated driving, showcase several approaches for enhancements of sensor data sharing and maneuver coordination, and indicate the performance of these enhancements.},
keywords = {Cooperative ITS, Intelligent Transport Systems, maneuver coordination, V2X communication, vehicle automation},
pubstate = {published},
tppubtype = {incollection}
}
This chapter studies two key communication services for the support of cooperative driving capabilities using Vehicle-to-Everything (V2X) communications: sensor data sharing and maneuver coordination. Based on the current state of the art in research and pre-standardization of V2X communications, we enhance the protocol design for both services and assess their performance by discrete-event simulations in highway and city scenarios. The first part of this chapter addresses the performance improvement of sensor data sharing by two complementary strategies. The shared sensor data are adapted to the available resources on the used channel. Furthermore, the redundancy of the transmitted information is reduced to lower the load on the wireless channel, whereas several approaches are proposed and assessed. The second part of the chapter analyzes cooperative maneuver coordination protocols. We propose a distributed approach based on the explicit exchange of V2X messages, which introduces priorities in maneuver coordination and studies several communication patterns for the negotiation and coordination of maneuvers among two and more vehicles. The results demonstrate the potential of V2X communications for automated driving, showcase several approaches for enhancements of sensor data sharing and maneuver coordination, and indicate the performance of these enhancements.
2.
Maksimovski, D.; Facchi, C.; Festag, A.
Packet Rate Control for Maneuver Coordination in Congested V2X Communication Environments Proceedings Article
In: IEEE 99th Vehicular Technology Conference (VTC 2024 Fall), Washington, DC, USA, 2024.
Abstract | BibTeX | Tags: Cooperative ITS, Intelligent Transport Systems, maneuver coordination, V2X communication, vehicle automation
@inproceedings{Maksimovski:VTC-Spring-Fall:2024,
title = {Packet Rate Control for Maneuver Coordination in Congested V2X Communication Environments},
author = {D. Maksimovski and C. Facchi and A. Festag},
year = {2024},
date = {2024-07-31},
booktitle = {IEEE 99th Vehicular Technology Conference (VTC 2024 Fall)},
address = {Washington, DC, USA},
abstract = {Maneuver coordination is envisioned as a future Vehicle-to-Everything (V2X) communication-enabled application for cooperative driving to improve the safety, comfort, and efficiency of driving. A Maneuver Coordination Service (MCS) is currently being standardized to facilitate negotiation and coordination of cooperative maneuvers for the Connected and Automated Vehicles (CAVs) by exchanging Maneuver Coordination Messages (MCMs). In order to ensure fair use of available communication channel resources and limit the channel load to reduce message collisions, the European Telecommunications Standards Institute (ETSI) has defined a Decentralized Congestion Control (DCC) framework with packet rate control at the access and facilities layer. However, the DCC mechanisms are content agnostic and can have a negative impact on the MC application in congested communication environments. This paper proposes a new method by combining an adaptive DCC mechanism with a no packet rate control method for important MCMs during the negotiation and execution of safety-critical and emergency maneuvers. A simulation environment is used to extensively evaluate the new method and demonstrate its advantages over the standardized approach. A challenging V2X scenario with a high number of vehicles is simulated in congested V2X channels with only MCMs and with a mix of V2X message types. The results demonstrate the significant improvement of the proposed method for packet rate control for the MCS.},
keywords = {Cooperative ITS, Intelligent Transport Systems, maneuver coordination, V2X communication, vehicle automation},
pubstate = {accepted},
tppubtype = {inproceedings}
}
Maneuver coordination is envisioned as a future Vehicle-to-Everything (V2X) communication-enabled application for cooperative driving to improve the safety, comfort, and efficiency of driving. A Maneuver Coordination Service (MCS) is currently being standardized to facilitate negotiation and coordination of cooperative maneuvers for the Connected and Automated Vehicles (CAVs) by exchanging Maneuver Coordination Messages (MCMs). In order to ensure fair use of available communication channel resources and limit the channel load to reduce message collisions, the European Telecommunications Standards Institute (ETSI) has defined a Decentralized Congestion Control (DCC) framework with packet rate control at the access and facilities layer. However, the DCC mechanisms are content agnostic and can have a negative impact on the MC application in congested communication environments. This paper proposes a new method by combining an adaptive DCC mechanism with a no packet rate control method for important MCMs during the negotiation and execution of safety-critical and emergency maneuvers. A simulation environment is used to extensively evaluate the new method and demonstrate its advantages over the standardized approach. A challenging V2X scenario with a high number of vehicles is simulated in congested V2X channels with only MCMs and with a mix of V2X message types. The results demonstrate the significant improvement of the proposed method for packet rate control for the MCS.
3.
Lobo, S.; Facchi, C.; Festag, A.
Adaptive Message Prioritization: How to Prioritize VRU Awareness Messages in a Congested V2X Network Proceedings Article
In: IEEE International Conference on Intelligent Transportation Systems (ITSC), Edmonton, Canada, 2024.
Abstract | Links | BibTeX | Tags: Intelligent Transport Systems, safety, V2X communication, vulnerable road users
@inproceedings{lobo:ITSC:2024,
title = {Adaptive Message Prioritization: How to Prioritize VRU Awareness Messages in a Congested V2X Network},
author = {S. Lobo and C. Facchi and A. Festag},
url = {https://ieee-itsc.org/2024/},
year = {2024},
date = {2024-07-18},
urldate = {2024-07-18},
booktitle = {IEEE International Conference on Intelligent Transportation Systems (ITSC)},
address = {Edmonton, Canada},
abstract = {The safety of Vulnerable Road Users (VRUs) can be improved through the utilization of Vehicle-to-Everything (V2X) communication, enabling the exchange of the VRUs' position and kinematic data. In the European V2X system, the dissemination relies on a dedicated message type, the VRU Awareness Message (VAM), which integrates VRUs into the V2X network as active transceivers. However, in the long term, the high number of V2X nodes transmitting messages poses challenges in the wireless resource allocation, which may result in safety-critical situations. To mitigate this issue, the current standardization of VRU awareness services in Europe, considers clustering VRUs with similar kinematic pattern. Nevertheless, the VRU process for joining a cluster may encounter errors, resulting in high channel load due to multiple requests. In this paper, an adaptive message prioritization is proposed to reduce this impact. This method is triggered only under heavy network load and takes into account the number of potential V2X senders. Via simulations based on a crossing scenario, improvements in network performance are observed, as well as an increase in VRU awareness measured by vehicles, i.e. more VRUs being detected.},
keywords = {Intelligent Transport Systems, safety, V2X communication, vulnerable road users},
pubstate = {accepted},
tppubtype = {inproceedings}
}
The safety of Vulnerable Road Users (VRUs) can be improved through the utilization of Vehicle-to-Everything (V2X) communication, enabling the exchange of the VRUs' position and kinematic data. In the European V2X system, the dissemination relies on a dedicated message type, the VRU Awareness Message (VAM), which integrates VRUs into the V2X network as active transceivers. However, in the long term, the high number of V2X nodes transmitting messages poses challenges in the wireless resource allocation, which may result in safety-critical situations. To mitigate this issue, the current standardization of VRU awareness services in Europe, considers clustering VRUs with similar kinematic pattern. Nevertheless, the VRU process for joining a cluster may encounter errors, resulting in high channel load due to multiple requests. In this paper, an adaptive message prioritization is proposed to reduce this impact. This method is triggered only under heavy network load and takes into account the number of potential V2X senders. Via simulations based on a crossing scenario, improvements in network performance are observed, as well as an increase in VRU awareness measured by vehicles, i.e. more VRUs being detected.
4.
Maksimowski, D.; Facchi, C.; Festag, A.
A Framework of Use Cases, Scenarios, and Metrics for Evaluation of V2X Maneuver Coordination Proceedings Article
In: IEEE Vehicular Networking Conference (VNC), Kobe, Japan, 2024.
Abstract | Links | BibTeX | Tags: Cooperative ITS, Intelligent Transport Systems, maneuver coordination, V2X communication, vehicle automation
@inproceedings{maksimovski:VNC:2024,
title = {A Framework of Use Cases, Scenarios, and Metrics for Evaluation of V2X Maneuver Coordination},
author = {D. Maksimowski and C. Facchi and A. Festag},
url = {https://ieee-vnc.org/2024},
doi = {10.1109/VNC61989.2024.10575992},
year = {2024},
date = {2024-07-05},
urldate = {2024-07-05},
booktitle = {IEEE Vehicular Networking Conference (VNC)},
address = {Kobe, Japan},
abstract = {Vehicle-to-everything (V2X) communication can enhance the capabilities of connected and automated vehicles (CAVs) to perform cooperative maneuver coordination (CMC), offering improvements in traffic safety, awareness, comfort and efficiency. Recent years have seen progress in the development of cooperative driving applications that leverage V2X maneuver coordination service and message (MCS and MCM), driven by both research and standardization efforts. However, the evaluation of such applications, along with defining relevant use cases and scenarios, lacks a common framework. This paper aims to fill this gap by proposing a framework for classifying use cases, scenarios and metrics for simulation-based performance evaluation of CMC applications. A simulation environment is employed to assess the effectiveness of the proposed metrics and evaluate a highway merging use case across various scenarios without coordination, with coordination, and in different communication conditions. Thus, this paper serves as a contribution to the development, testing and evaluation of CMC applications.},
keywords = {Cooperative ITS, Intelligent Transport Systems, maneuver coordination, V2X communication, vehicle automation},
pubstate = {published},
tppubtype = {inproceedings}
}
Vehicle-to-everything (V2X) communication can enhance the capabilities of connected and automated vehicles (CAVs) to perform cooperative maneuver coordination (CMC), offering improvements in traffic safety, awareness, comfort and efficiency. Recent years have seen progress in the development of cooperative driving applications that leverage V2X maneuver coordination service and message (MCS and MCM), driven by both research and standardization efforts. However, the evaluation of such applications, along with defining relevant use cases and scenarios, lacks a common framework. This paper aims to fill this gap by proposing a framework for classifying use cases, scenarios and metrics for simulation-based performance evaluation of CMC applications. A simulation environment is employed to assess the effectiveness of the proposed metrics and evaluate a highway merging use case across various scenarios without coordination, with coordination, and in different communication conditions. Thus, this paper serves as a contribution to the development, testing and evaluation of CMC applications.
5.
Song, R.; Festag, A.; Jagtap, A. Dinkar; Bialdyga, M.; Yan, Z.; Otte, M.; A, Knoll
``First Mile'' – An Open Innovation Lab for Infrastructure-Assisted Cooperative Intelligent Transportation Systems: Architecture, Validation and Outlook Proceedings Article
In: 35th IEEE Intelligent Vehicles Symposium (IV), Jeju Island, Korea, 2024.
Abstract | Links | BibTeX | Tags: Cooperative ITS, Intelligent Transport Systems, testfield, V2X communication
@inproceedings{Song:IV:2024,
title = {``First Mile'' – An Open Innovation Lab for Infrastructure-Assisted Cooperative Intelligent Transportation Systems: Architecture, Validation and Outlook},
author = {R. Song and A. Festag and A. Dinkar Jagtap and M. Bialdyga and Z. Yan and M. Otte and Knoll A},
url = {https://ieeexplore.ieee.org/document/10588500},
doi = {10.1109/IV55156.2024.10588500},
year = {2024},
date = {2024-04-02},
urldate = {2024-07-15},
booktitle = {35th IEEE Intelligent Vehicles Symposium (IV)},
address = {Jeju Island, Korea},
abstract = {Infrastructure-assisted cooperative intelligent transportation systems (C-ITS) leverage roadside intelligent infrastructure and vehicular network technology to facilitate information exchange among traffic participants, enhancing road safety, efficiency, and sustainability. However, this requires not only the massive deployment of infrastructure, including advanced sensors, communication devices, and computing units at various levels, but also the involvement of various stakeholders in the development of functions and real-road testing. In this paper, we present our test field – myTestField, as an emphopen innovation lab for C-ITS in Ingolstadt, Germany, offering an open environment for world-wide stakeholders to conduct research and testing in C-ITS. In particular, we equip a 3.5 km area with 22 roadside intelligent masts and 89 sensors, achieving dense deployment on public roads. Our design includes a protocol stack tailored for different C-ITS stations and services, conforming to European communication standards. Furthermore, we conduct quantitative analyzes of key performance metrics, such as radio signal quality and End-to-End delay, to assess the efficacy of myTestField in different data processing pipelines. Finally, we delve into the future prospects of large-scale C-ITS deployment, guided by extensive and prolonged measurement studies.},
keywords = {Cooperative ITS, Intelligent Transport Systems, testfield, V2X communication},
pubstate = {published},
tppubtype = {inproceedings}
}
Infrastructure-assisted cooperative intelligent transportation systems (C-ITS) leverage roadside intelligent infrastructure and vehicular network technology to facilitate information exchange among traffic participants, enhancing road safety, efficiency, and sustainability. However, this requires not only the massive deployment of infrastructure, including advanced sensors, communication devices, and computing units at various levels, but also the involvement of various stakeholders in the development of functions and real-road testing. In this paper, we present our test field – myTestField, as an emphopen innovation lab for C-ITS in Ingolstadt, Germany, offering an open environment for world-wide stakeholders to conduct research and testing in C-ITS. In particular, we equip a 3.5 km area with 22 roadside intelligent masts and 89 sensors, achieving dense deployment on public roads. Our design includes a protocol stack tailored for different C-ITS stations and services, conforming to European communication standards. Furthermore, we conduct quantitative analyzes of key performance metrics, such as radio signal quality and End-to-End delay, to assess the efficacy of myTestField in different data processing pipelines. Finally, we delve into the future prospects of large-scale C-ITS deployment, guided by extensive and prolonged measurement studies.
6.
Bauder, M.; Festag, A.; Kubjatko, T.; Schweiger, H. -G.
Data Accuracy in Vehicle-to-X Cooperative Awareness Messages: An Experimental Study for the First Commercial Deployment of C-ITS in Europe Journal Article
In: Elsevier Vehicular Communications, vol. 47, pp. 31, 2024.
Abstract | Links | BibTeX | Tags: Cooperative ITS, Intelligent Transport Systems, measurements, V2X communication
@article{Bauder:ElsevierVehComm:2024_DataAccuracy_V2X,
title = {Data Accuracy in Vehicle-to-X Cooperative Awareness Messages: An Experimental Study for the First Commercial Deployment of C-ITS in Europe},
author = {M. Bauder and A. Festag and T. Kubjatko and H. -G. Schweiger},
url = {https://www.sciencedirect.com/science/article/pii/S2214209624000196},
doi = {10.1016/j.vehcom.2024.100744},
year = {2024},
date = {2024-03-14},
urldate = {2024-03-14},
journal = {Elsevier Vehicular Communications},
volume = {47},
pages = {31},
abstract = {Cooperative Intelligent Transportation Systems have achieved a mature technology stage and are in an early phase of mass deployment in Europe. Relying on Vehicle-to-X communication, these systems were primarily developed to improve traffic safety, efficiency and driving comfort. However, they also offer great opportunities for other use cases. One of them is forensic accident analysis, where the received data provide details about the status of other traffic participants, give insights into the accident scenario and therefore help in understanding accident causes. A high accuracy of the sent information is essential: For safety use cases, such as traffic jam warning, a poor accuracy of the data may result in wrong driver information, undermine the usability of the system and even create new safety risks. For accident analysis, a low accuracy may prevent the correct reconstruction of an accident. This paper presents an experimental study of the first generation of Cooperative Intelligent Transportation Systems in Europe. The results indicate a high accuracy for most of the data fields in the Vehicle-to-X messages, namely speed, acceleration, heading and yaw rate information, which meet the accuracy requirements for safety use cases and accident analysis. In contrast, the position data, which are also carried in the messages, have larger errors. Specifically, we observed that the lateral position has still an acceptable accuracy. The error of the longitudinal position is larger and may compromise safety use cases with high accuracy requirements. Even with limited accuracy, the data provide a high value for the accident analysis. Since we also found that the accuracy of the data increases for newer vehicle models, we presume that Vehicle-to-X data have the potential for exact accident reconstruction.},
keywords = {Cooperative ITS, Intelligent Transport Systems, measurements, V2X communication},
pubstate = {published},
tppubtype = {article}
}
Cooperative Intelligent Transportation Systems have achieved a mature technology stage and are in an early phase of mass deployment in Europe. Relying on Vehicle-to-X communication, these systems were primarily developed to improve traffic safety, efficiency and driving comfort. However, they also offer great opportunities for other use cases. One of them is forensic accident analysis, where the received data provide details about the status of other traffic participants, give insights into the accident scenario and therefore help in understanding accident causes. A high accuracy of the sent information is essential: For safety use cases, such as traffic jam warning, a poor accuracy of the data may result in wrong driver information, undermine the usability of the system and even create new safety risks. For accident analysis, a low accuracy may prevent the correct reconstruction of an accident. This paper presents an experimental study of the first generation of Cooperative Intelligent Transportation Systems in Europe. The results indicate a high accuracy for most of the data fields in the Vehicle-to-X messages, namely speed, acceleration, heading and yaw rate information, which meet the accuracy requirements for safety use cases and accident analysis. In contrast, the position data, which are also carried in the messages, have larger errors. Specifically, we observed that the lateral position has still an acceptable accuracy. The error of the longitudinal position is larger and may compromise safety use cases with high accuracy requirements. Even with limited accuracy, the data provide a high value for the accident analysis. Since we also found that the accuracy of the data increases for newer vehicle models, we presume that Vehicle-to-X data have the potential for exact accident reconstruction.
7.
Maksimovski, D.; Facchi, C.; Festag, A.
Prioritization of Maneuver Coordination Messages and the Impact of Decentralized Congestion Control Proceedings Article
In: IEEE 99th Vehicular Technology Conference (VTC 2024 Spring), Singapore, 2024.
Abstract | BibTeX | Tags: Cooperative ITS, Intelligent Transport Systems, maneuver coordination, V2X communication, vehicle automation
@inproceedings{Maksimovski:VTC-Spring:2024,
title = {Prioritization of Maneuver Coordination Messages and the Impact of Decentralized Congestion Control},
author = {D. Maksimovski and C. Facchi and A. Festag},
year = {2024},
date = {2024-03-13},
booktitle = {IEEE 99th Vehicular Technology Conference (VTC 2024 Spring)},
address = {Singapore},
abstract = {Vehicle-to-everything (V2X) communication can enable connected and automated vehicles (CAVs) to drive cooperatively by coordinating their maneuvers. Envisioned as a future Day 3+ V2X application, CAVs can broadcast Maneuver Coordination Messages (MCMs) as part of a Maneuver Coordination Service (MCS) to exchange the necessary data. In order to prevent congestion and ensure fair use of the communication channel by the increasing number of V2X services, a Decentralized Congestion Control (DCC) mechanism with packet rate control has been developed and standardized by the European Telecommunications Standards Institute (ETSI). In this paper, a prioritization of MCMs is proposed based on the necessity of the maneuver and the operation mode of the MCS. Several application and network related metrics are used to analyze the impact of the DCC on the MCS with and without prioritization. A challenging scenario with a high number of transmitting vehicles is simulated, congesting the communication channel with only MCMs and with the standardized Cooperative Awareness Messages (CAMs). Evaluations are performed with DCC at the access and facilities layers. The results demonstrate the benefits of MCM prioritization and discuss the impact and issues of DCC mechanisms in the context of MCS.},
keywords = {Cooperative ITS, Intelligent Transport Systems, maneuver coordination, V2X communication, vehicle automation},
pubstate = {accepted},
tppubtype = {inproceedings}
}
Vehicle-to-everything (V2X) communication can enable connected and automated vehicles (CAVs) to drive cooperatively by coordinating their maneuvers. Envisioned as a future Day 3+ V2X application, CAVs can broadcast Maneuver Coordination Messages (MCMs) as part of a Maneuver Coordination Service (MCS) to exchange the necessary data. In order to prevent congestion and ensure fair use of the communication channel by the increasing number of V2X services, a Decentralized Congestion Control (DCC) mechanism with packet rate control has been developed and standardized by the European Telecommunications Standards Institute (ETSI). In this paper, a prioritization of MCMs is proposed based on the necessity of the maneuver and the operation mode of the MCS. Several application and network related metrics are used to analyze the impact of the DCC on the MCS with and without prioritization. A challenging scenario with a high number of transmitting vehicles is simulated, congesting the communication channel with only MCMs and with the standardized Cooperative Awareness Messages (CAMs). Evaluations are performed with DCC at the access and facilities layers. The results demonstrate the benefits of MCM prioritization and discuss the impact and issues of DCC mechanisms in the context of MCS.
8.
Song, R.; Liang, C.; Cao, H.; Yan, Z.; Zimmer, W.; Gross, M.; Festag, A.; Knoll, A.
Collaborative Semantic Occupancy Prediction with Hybrid Feature Fusion in Connected Automated Vehicles Proceedings Article
In: 2024 Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 2024.
Abstract | Links | BibTeX | Tags: Intelligent Transport Systems, perception, sensor data sharing
@inproceedings{Song:CVPR:2024,
title = {Collaborative Semantic Occupancy Prediction with Hybrid Feature Fusion in Connected Automated Vehicles},
author = {R. Song and C. Liang and H. Cao and Z. Yan and W. Zimmer and M. Gross and A. Festag and A. Knoll},
url = {https://cvpr.thecvf.com/Conferences/2024},
year = {2024},
date = {2024-02-27},
urldate = {2024-02-27},
booktitle = {2024 Conference on Computer Vision and Pattern Recognition (CVPR)},
address = {Seattle, WA, USA},
abstract = {Collaborative perception in automated vehicles leverages the exchange of information between agents, aiming to elevate perception results. Previous camera-based collaborative 3D perception methods typically employ 3D bounding boxes or bird’s eye views as representations of the environment. However, these approaches fall short in offering a comprehensive 3D environmental prediction. To bridge this gap, we introduce the first method for collaborative 3D semantic occupancy prediction. Particularly, it improves local 3D semantic occupancy predictions by hybrid fusion of (i) semantic and occupancy task features, and (ii) compressed orthogonal attention features shared between vehicles. Additionally, due to the lack of a collaborative perception dataset designed for semantic occupancy prediction, we augment a current collaborative perception dataset to include 3D collaborative semantic occupancy labels for amore robust evaluation. The experimental findings highlight that: (i) our collaborative semantic occupancy predictions excel above the results from single vehicles by over 30%, and (ii) models anchored on semantic occupancy outpace state-of-the-art collaborative 3D detection techniques in subsequent perception applications, showcasing enhanced accuracy and enriched semantic-awareness in road environments.},
keywords = {Intelligent Transport Systems, perception, sensor data sharing},
pubstate = {accepted},
tppubtype = {inproceedings}
}
Collaborative perception in automated vehicles leverages the exchange of information between agents, aiming to elevate perception results. Previous camera-based collaborative 3D perception methods typically employ 3D bounding boxes or bird’s eye views as representations of the environment. However, these approaches fall short in offering a comprehensive 3D environmental prediction. To bridge this gap, we introduce the first method for collaborative 3D semantic occupancy prediction. Particularly, it improves local 3D semantic occupancy predictions by hybrid fusion of (i) semantic and occupancy task features, and (ii) compressed orthogonal attention features shared between vehicles. Additionally, due to the lack of a collaborative perception dataset designed for semantic occupancy prediction, we augment a current collaborative perception dataset to include 3D collaborative semantic occupancy labels for amore robust evaluation. The experimental findings highlight that: (i) our collaborative semantic occupancy predictions excel above the results from single vehicles by over 30%, and (ii) models anchored on semantic occupancy outpace state-of-the-art collaborative 3D detection techniques in subsequent perception applications, showcasing enhanced accuracy and enriched semantic-awareness in road environments.
9.
Bazzi, A.; Sepulcre, M.; Delooz, Q.; Festag, A.; Vogt, J.; Wieker, H.; Berens, F.; Spaanderman, P.
Multi-Channel Operation for the Release 2 of ETSI Cooperative Intelligent Transport Systems Journal Article
In: IEEE Communications Standards Magazine, vol. 8, no. 1, pp. 28-35, 2023.
Abstract | Links | BibTeX | Tags: Cooperative ITS, Intelligent Transport Systems, Multi-channel operation, V2X communication
@article{Bazzi:IEEE-CommStdMag:2023_ETSIMCO,
title = {Multi-Channel Operation for the Release 2 of ETSI Cooperative Intelligent Transport Systems},
author = {A. Bazzi and M. Sepulcre and Q. Delooz and A. Festag and J. Vogt and H. Wieker and F. Berens and P. Spaanderman},
url = {https://ieeexplore.ieee.org/document/10467184},
doi = {10.1109/MCOMSTD.0001.2200080},
year = {2023},
date = {2023-10-12},
urldate = {2024-03-18},
journal = {IEEE Communications Standards Magazine},
volume = {8},
number = {1},
pages = {28-35},
abstract = {Vehicles and road infrastructure are starting to be equipped with vehicle-to-everything (V2X) communication solutions to increase road safety and provide new services to drivers and passengers. In Europe, the deployment is based on a set of Release 1 standards developed by ETSI to support basic use cases for cooperative intelligent transport systems (C-ITS). For them, the capacity of a single 10 MHz channel in the ITS band at 5.9 GHz is considered sufficient. At the same time, the ITS stakeholders are working towards several advanced use cases, which imply a significant increment of data traffic and the need for multiple channels. To address this issue, ETSI has recently standardized a new multi-channel operation (MCO) concept for flexible, efficient, and future-proof use of multiple channels. This new concept is defined in a set of new specifications that represent the foundation for the future releases of C-ITS standards. The present paper provides a comprehensive review of the new set of specifications, describing the main entities extending the C-ITS architecture at the different layers of the protocol stack, In addition, the paper provides representative examples that describe how these MCO standards will be used in the future and discusses some of the main open issues arising. The review and analysis of this paper facilitate the understanding and motivation of the new set of Release 2 ETSI specifications for MCO and the identification of new research opportunities.},
keywords = {Cooperative ITS, Intelligent Transport Systems, Multi-channel operation, V2X communication},
pubstate = {published},
tppubtype = {article}
}
Vehicles and road infrastructure are starting to be equipped with vehicle-to-everything (V2X) communication solutions to increase road safety and provide new services to drivers and passengers. In Europe, the deployment is based on a set of Release 1 standards developed by ETSI to support basic use cases for cooperative intelligent transport systems (C-ITS). For them, the capacity of a single 10 MHz channel in the ITS band at 5.9 GHz is considered sufficient. At the same time, the ITS stakeholders are working towards several advanced use cases, which imply a significant increment of data traffic and the need for multiple channels. To address this issue, ETSI has recently standardized a new multi-channel operation (MCO) concept for flexible, efficient, and future-proof use of multiple channels. This new concept is defined in a set of new specifications that represent the foundation for the future releases of C-ITS standards. The present paper provides a comprehensive review of the new set of specifications, describing the main entities extending the C-ITS architecture at the different layers of the protocol stack, In addition, the paper provides representative examples that describe how these MCO standards will be used in the future and discusses some of the main open issues arising. The review and analysis of this paper facilitate the understanding and motivation of the new set of Release 2 ETSI specifications for MCO and the identification of new research opportunities.
10.
Song, R.; Zhou, L.; Lyu, L.; Festag, A.; Knoll, A.
ResFed: Communication Efficient Federated Learning by Transmitting Deep Compressed Residuals Journal Article
In: IEEE Internet of Things Journal, vol. 11, no. 6, pp. 9458-9472, 2023.
Abstract | Links | BibTeX | Tags: federated learning, Intelligent Transport Systems
@article{Song:IEEE-IOT:2023_ResFed,
title = {ResFed: Communication Efficient Federated Learning by Transmitting Deep Compressed Residuals},
author = {R. Song and L. Zhou and L. Lyu and A. Festag and A. Knoll},
url = {https://ieeexplore.ieee.org/document/10283999},
doi = {10.1109/JIOT.2023.3324079},
year = {2023},
date = {2023-10-12},
urldate = {2023-10-12},
journal = {IEEE Internet of Things Journal},
volume = {11},
number = {6},
pages = {9458-9472},
abstract = {Federated learning allows for cooperative training among distributed clients by sharing their locally learned model parameters, such as weights or gradients. However, as model size increases, the communication bandwidth required for deployment in wireless networks becomes a bottleneck. To address this, we propose a residual-based federated learning framework (ResFed) that transmits residuals instead of gradients or weights in networks. By predicting model updates at both clients and the server, residuals are calculated as the difference between updated and predicted models and contain more dense information than weights or gradients. We find that the residuals are less sensitive to an increasing compression ratio than other parameters, and hence use lossy compression techniques on residuals to improve communication efficiency for training in federated settings. With the same compression ratio, ResFed outperforms current methods (weight- or gradient-based federated learning) by over 1.4x on federated datasets, including MNIST, FashionMNIST, SVHN, CIFAR-10, CIFAR-100, FEMNIST, in client-to-server communication, and can also be applied to reduce communication costs for server-to-client communication.},
keywords = {federated learning, Intelligent Transport Systems},
pubstate = {published},
tppubtype = {article}
}
Federated learning allows for cooperative training among distributed clients by sharing their locally learned model parameters, such as weights or gradients. However, as model size increases, the communication bandwidth required for deployment in wireless networks becomes a bottleneck. To address this, we propose a residual-based federated learning framework (ResFed) that transmits residuals instead of gradients or weights in networks. By predicting model updates at both clients and the server, residuals are calculated as the difference between updated and predicted models and contain more dense information than weights or gradients. We find that the residuals are less sensitive to an increasing compression ratio than other parameters, and hence use lossy compression techniques on residuals to improve communication efficiency for training in federated settings. With the same compression ratio, ResFed outperforms current methods (weight- or gradient-based federated learning) by over 1.4x on federated datasets, including MNIST, FashionMNIST, SVHN, CIFAR-10, CIFAR-100, FEMNIST, in client-to-server communication, and can also be applied to reduce communication costs for server-to-client communication.
11.
Zhou, L.; Song, R.; Chen, G.; Festag, A.; Knoll, A.
Residual Encoding Framework to Compress DNN Parameters for Fast Transfer Journal Article
In: Knowledge-Based Systems, vol. 277, pp. 110815, 2023, ISSN: 0950-7051.
Abstract | Links | BibTeX | Tags: federated learning, Intelligent Transport Systems
@article{Zhou:KnowledgeBasedSystems:2023_ResidualEncoding,
title = {Residual Encoding Framework to Compress DNN Parameters for Fast Transfer},
author = {L. Zhou and R. Song and G. Chen and A. Festag and A. Knoll},
url = {https://www.sciencedirect.com/science/article/pii/S0950705123005658},
doi = {10.1016/j.knosys.2023.110815},
issn = {0950-7051},
year = {2023},
date = {2023-10-09},
urldate = {2023-10-09},
journal = {Knowledge-Based Systems},
volume = {277},
pages = {110815},
abstract = {Efficient communication is significant for federated learning and DNN model deployment. However, transferring hundreds of millions of DNN parameters over networks with limited bandwidth results in long communication delays or even data losses. To alleviate or even remove the communication bottleneck, efficient methods for parameter compression can be applied. Inspired by video encoding, which exploits inter-frame similarity for compression, we investigate the strong temporal correlations of parameter updates in two near epochs of the DNN model and introduce a model parameter residual encoding framework. By transmitting encoded residual between model parameters in two near epochs, the receiver can reconstruct new model parameters and finish the updates with less communication cost. Furthermore, with respect to our framework, we develop lossless and lossy model parameter compression methods and demonstrate them on popular classification and detection networks. The results show that the lossless method can compress the data size of the parameters to less than 90%, and the lossy method can shrink the parameter size to less than 50% with a fair low loss. Our source code is released at https://github.com/zhouliguo/DNN_param_encode.},
keywords = {federated learning, Intelligent Transport Systems},
pubstate = {published},
tppubtype = {article}
}
Efficient communication is significant for federated learning and DNN model deployment. However, transferring hundreds of millions of DNN parameters over networks with limited bandwidth results in long communication delays or even data losses. To alleviate or even remove the communication bottleneck, efficient methods for parameter compression can be applied. Inspired by video encoding, which exploits inter-frame similarity for compression, we investigate the strong temporal correlations of parameter updates in two near epochs of the DNN model and introduce a model parameter residual encoding framework. By transmitting encoded residual between model parameters in two near epochs, the receiver can reconstruct new model parameters and finish the updates with less communication cost. Furthermore, with respect to our framework, we develop lossless and lossy model parameter compression methods and demonstrate them on popular classification and detection networks. The results show that the lossless method can compress the data size of the parameters to less than 90%, and the lossy method can shrink the parameter size to less than 50% with a fair low loss. Our source code is released at https://github.com/zhouliguo/DNN_param_encode.
12.
Hegde, A.; Song, R.; Festag, A.
Radio Resource Allocation in 5G-NR V2X: A Multi-Agent Actor-Critic Based Approach Journal Article
In: IEEE Access, vol. 11, pp. 87225-87244, 2023.
Abstract | Links | BibTeX | Tags: Cellular V2X, Cooperative ITS, Intelligent Transport Systems, machine learning, V2X communication
@article{Hegde:Access:2023_ResAlloc_ActorCritic,
title = {Radio Resource Allocation in 5G-NR V2X: A Multi-Agent Actor-Critic Based Approach},
author = {A. Hegde and R. Song and A. Festag},
url = {https://ieeexplore.ieee.org/document/10216959},
doi = {10.1109/ACCESS.2023.3305267},
year = {2023},
date = {2023-08-04},
urldate = {2023-08-15},
journal = {IEEE Access},
volume = {11},
pages = {87225-87244},
abstract = {The efficiency of radio resource allocation and scheduling procedures in Cellular Vehicle-to-X (Cellular V2X) communication networks directly affects link quality in terms of latency and reliability. However, owing to the continuous movement of vehicles, it is impossible to have a centralized coordinating unit at all times to manage the allocation of radio resources. In the unmanaged mode of the fifth generation new radio (5G-NR) V2X, the sensing-based semi-persistent scheduling (SB-SPS) loses its effectiveness when V2X data messages become aperiodic with varying data sizes. This leads to misinformed resource allocation decisions among vehicles and frequent resource collisions. To improve resource selection, this study formulates the Cellular V2X communication network as a decentralized multi-agent networked markov decision process (MDP) where each vehicle agent executes an actor-critic-based radio resource scheduler. Developing further the actor-critic methodology for the radio resource allocation problem in Cellular V2X, two variants are derived: independent actor-critic (IAC) and shared experience actor-critic (SEAC). Results from simulation studies indicate that the actor-critic schedulers improve reliability, achieving a 15-20% higher probability of reception under high vehicular density scenarios with aperiodic traffic patterns.},
keywords = {Cellular V2X, Cooperative ITS, Intelligent Transport Systems, machine learning, V2X communication},
pubstate = {published},
tppubtype = {article}
}
The efficiency of radio resource allocation and scheduling procedures in Cellular Vehicle-to-X (Cellular V2X) communication networks directly affects link quality in terms of latency and reliability. However, owing to the continuous movement of vehicles, it is impossible to have a centralized coordinating unit at all times to manage the allocation of radio resources. In the unmanaged mode of the fifth generation new radio (5G-NR) V2X, the sensing-based semi-persistent scheduling (SB-SPS) loses its effectiveness when V2X data messages become aperiodic with varying data sizes. This leads to misinformed resource allocation decisions among vehicles and frequent resource collisions. To improve resource selection, this study formulates the Cellular V2X communication network as a decentralized multi-agent networked markov decision process (MDP) where each vehicle agent executes an actor-critic-based radio resource scheduler. Developing further the actor-critic methodology for the radio resource allocation problem in Cellular V2X, two variants are derived: independent actor-critic (IAC) and shared experience actor-critic (SEAC). Results from simulation studies indicate that the actor-critic schedulers improve reliability, achieving a 15-20% higher probability of reception under high vehicular density scenarios with aperiodic traffic patterns.
13.
Song, R.; Lyu, L.; Jiang, W.; Festag, A.; Knoll, A.
V2X-Boosted Federated Learning for Cooperative Intelligent Transportation Systems with Contextual Client Selection Journal Article
In: arXiv, 2023, (arXiv preprint arXiv:2305.11654).
Abstract | Links | BibTeX | Tags: client selection, federated learning, Intelligent Transport Systems
@article{Song:ARXIV:2023_V2XBoosted,
title = {V2X-Boosted Federated Learning for Cooperative Intelligent Transportation Systems with Contextual Client Selection},
author = {R. Song and L. Lyu and W. Jiang and A. Festag and A. Knoll},
url = {https://arxiv.org/pdf/2305.11654},
year = {2023},
date = {2023-06-21},
urldate = {2023-06-21},
journal = {arXiv},
abstract = {Machine learning (ML) has revolutionized transportation systems, enabling autonomous driving and smart traffic services. Federated learning (FL) overcomes privacy constraints by training ML models in distributed systems, exchanging model parameters instead of raw data. However, the dynamic states of connected vehicles affect the network connection quality and influence the FL performance. To tackle this challenge, we propose a contextual client selection pipeline that uses Vehicle-to-Everything (V2X) messages to select clients based on the predicted communication latency. The pipeline includes: (i) fusing V2X messages, (ii) predicting future traffic topology, (iii) pre-clustering clients based on local data distribution similarity, and (iv) selecting clients with minimal latency for future model aggregation. Experiments show that our pipelineoutperforms baselines on various datasets, particularly in non-iid settings.},
note = {arXiv preprint arXiv:2305.11654},
keywords = {client selection, federated learning, Intelligent Transport Systems},
pubstate = {published},
tppubtype = {article}
}
Machine learning (ML) has revolutionized transportation systems, enabling autonomous driving and smart traffic services. Federated learning (FL) overcomes privacy constraints by training ML models in distributed systems, exchanging model parameters instead of raw data. However, the dynamic states of connected vehicles affect the network connection quality and influence the FL performance. To tackle this challenge, we propose a contextual client selection pipeline that uses Vehicle-to-Everything (V2X) messages to select clients based on the predicted communication latency. The pipeline includes: (i) fusing V2X messages, (ii) predicting future traffic topology, (iii) pre-clustering clients based on local data distribution similarity, and (iv) selecting clients with minimal latency for future model aggregation. Experiments show that our pipelineoutperforms baselines on various datasets, particularly in non-iid settings.
14.
Song, R.; Liu, D.; Chen, D. Z.; Festag, A.; Trinitis, C.; Schulz, M.; Knoll, A.
Federated Learning via Decentralized Dataset Distillation in Resource-Constrained Edge Environments Proceedings Article
In: International Joint Conference on Neural Networks (IJCNN 2023), Queensland, Australia, 2023.
Abstract | Links | BibTeX | Tags: federated learning, Intelligent Transport Systems
@inproceedings{Song:IJCNN:2023,
title = {Federated Learning via Decentralized Dataset Distillation in Resource-Constrained Edge Environments},
author = {R. Song and D. Liu and D. Z. Chen and A. Festag and C. Trinitis and M. Schulz and A. Knoll},
url = {https://2023.ijcnn.org},
doi = {10.1109/IJCNN54540.2023.10191879},
year = {2023},
date = {2023-06-20},
urldate = {2023-06-20},
booktitle = {International Joint Conference on Neural Networks (IJCNN 2023)},
address = {Queensland, Australia},
abstract = {In federated learning, all networked clients contribute to the model training cooperatively. However, with model sizes increasing, even sharing the trained partial models often leads to severe communication bottlenecks in underlying networks, especially when communicated iteratively. In this paper, we introduce a federated learning framework FedD3 requiring only one-shot communication by integrating dataset distillation instances. Instead of sharing model updates in other federated learning approaches, FedD3 allows the connected clients to distill the local datasets independently, and then aggregates those decentralized distilled datasets (e.g. a few unrecognizable images) from networks for model training. Our experimental results show that FedD3 significantly outperforms other federated learning frameworks in terms of needed communication volumes, while it provides the additional benefit to be able to balance the trade-off between accuracy and communication cost, depending on usage scenario or target dataset. For instance, for training an AlexNet model on CIFAR-10 with 10 clients under non-independent and identically distributed (Non-IID) setting, FedD3 can either increase the accuracy by over 71% with a similar communication volume, or save 98% of communication volume, while reaching the same accuracy, compared to other one-shot federated learning approaches.},
keywords = {federated learning, Intelligent Transport Systems},
pubstate = {published},
tppubtype = {inproceedings}
}
In federated learning, all networked clients contribute to the model training cooperatively. However, with model sizes increasing, even sharing the trained partial models often leads to severe communication bottlenecks in underlying networks, especially when communicated iteratively. In this paper, we introduce a federated learning framework FedD3 requiring only one-shot communication by integrating dataset distillation instances. Instead of sharing model updates in other federated learning approaches, FedD3 allows the connected clients to distill the local datasets independently, and then aggregates those decentralized distilled datasets (e.g. a few unrecognizable images) from networks for model training. Our experimental results show that FedD3 significantly outperforms other federated learning frameworks in terms of needed communication volumes, while it provides the additional benefit to be able to balance the trade-off between accuracy and communication cost, depending on usage scenario or target dataset. For instance, for training an AlexNet model on CIFAR-10 with 10 clients under non-independent and identically distributed (Non-IID) setting, FedD3 can either increase the accuracy by over 71% with a similar communication volume, or save 98% of communication volume, while reaching the same accuracy, compared to other one-shot federated learning approaches.
15.
Song, R.; Lyu, L.; Jiang, W.; Festag, A.; Knoll, A.
V2X-Boosted Federated Learning for Cooperative Intelligent Transportation Systems with Contextual Client Selection Proceedings Article
In: International Joint Conference on Neural Networks (IJCNN 2023) Workshop on Collaborative Perception and Learning, London, UK, 2023.
Abstract | Links | BibTeX | Tags: federated learning, Intelligent Transport Systems
@inproceedings{Song:ICRA:2023,
title = {V2X-Boosted Federated Learning for Cooperative Intelligent Transportation Systems with Contextual Client Selection},
author = {R. Song and L. Lyu and W. Jiang and A. Festag and A. Knoll},
url = {https://www.icra2023.org/programme/workshops-tutorials},
doi = {10.48550/arXiv.2305.11654},
year = {2023},
date = {2023-05-30},
urldate = {2023-05-30},
booktitle = {International Joint Conference on Neural Networks (IJCNN 2023) Workshop on Collaborative Perception and Learning},
address = {London, UK},
abstract = {In federated learning, all networked clients contribute to the model training cooperatively. However, with model sizes increasing, even sharing the trained partial models often leads to severe communication bottlenecks in underlying networks, especially when communicated iteratively. In this paper, we introduce a federated learning framework FedD3 requiring only one-shot communication by integrating dataset distillation instances. Instead of sharing model updates in other federated learning approaches, FedD3 allows the connected clients to distill the local datasets independently, and then aggregates those decentralized distilled datasets (e.g. a few unrecognizable images) from networks for model training. Our experimental results show that FedD3 significantly outperforms other federated learning frameworks in terms of needed communication volumes, while it provides the additional benefit to be able to balance the trade-off between accuracy and communication cost, depending on usage scenario or target dataset. For instance, for training an AlexNet model on CIFAR-10 with 10 clients under non-independent and identically distributed (Non-IID) setting, FedD3 can either increase the accuracy by over 71% with a similar communication volume, or save 98% of communication volume, while reaching the same accuracy, compared to other one-shot federated learning approaches.},
keywords = {federated learning, Intelligent Transport Systems},
pubstate = {published},
tppubtype = {inproceedings}
}
In federated learning, all networked clients contribute to the model training cooperatively. However, with model sizes increasing, even sharing the trained partial models often leads to severe communication bottlenecks in underlying networks, especially when communicated iteratively. In this paper, we introduce a federated learning framework FedD3 requiring only one-shot communication by integrating dataset distillation instances. Instead of sharing model updates in other federated learning approaches, FedD3 allows the connected clients to distill the local datasets independently, and then aggregates those decentralized distilled datasets (e.g. a few unrecognizable images) from networks for model training. Our experimental results show that FedD3 significantly outperforms other federated learning frameworks in terms of needed communication volumes, while it provides the additional benefit to be able to balance the trade-off between accuracy and communication cost, depending on usage scenario or target dataset. For instance, for training an AlexNet model on CIFAR-10 with 10 clients under non-independent and identically distributed (Non-IID) setting, FedD3 can either increase the accuracy by over 71% with a similar communication volume, or save 98% of communication volume, while reaching the same accuracy, compared to other one-shot federated learning approaches.
16.
Song, R.; Xu, R.; Festag, A.; Ma, J.; Knoll, A.
FedBEVT: Federated Learning Bird's Eye View Perception Transformer in Road Traffic Systems Journal Article
In: IEEE Transactions on Intelligent Vehicles, vol. 9, no. 1, pp. 958-969, 2023.
Abstract | Links | BibTeX | Tags: federated learning, Intelligent Transport Systems
@article{Song:TIV:2024-FedBEVT,
title = {FedBEVT: Federated Learning Bird's Eye View Perception Transformer in Road Traffic Systems},
author = {R. Song and R. Xu and A. Festag and J. Ma and A. Knoll},
url = {https://ieeexplore.ieee.org/document/10236488},
doi = {10.1109/TIV.2023.3310674},
year = {2023},
date = {2023-04-08},
urldate = {2023-04-04},
journal = {IEEE Transactions on Intelligent Vehicles},
volume = {9},
number = {1},
pages = {958-969},
abstract = {Bird's eye view (BEV) perception is becoming increasingly important in the field of autonomous driving. It uses multi-view camera data to learn a transformer model that directly projects the perception of the road environment onto the BEV perspective. However, training a transformer model often requires a large amount of data, and as camera data for road traffic are often private, they are typically not shared. Federated learning offers a solution that enables clients to collaborate and train models without exchanging data but model parameters. In this paper, we introduce FedBEVT, a federated transformer learning approach for BEV perception. In order to address two common data heterogeneity issues in FedBEVT: (i) diverse sensor poses, and (ii) varying sensor numbers in perception systems, we propose two approaches - Federated Learning with Camera-Attentive Personalization (FedCaP) and Adaptive Multi-Camera Masking (AMCM), respectively. To evaluate our method in real-world settings, we create a dataset consisting of four typical federated use cases. Our findings suggest that FedBEVT outperforms the baseline approaches in all four use cases, demonstrating the potential of our approach for improving BEV perception in autonomous driving.},
keywords = {federated learning, Intelligent Transport Systems},
pubstate = {published},
tppubtype = {article}
}
Bird's eye view (BEV) perception is becoming increasingly important in the field of autonomous driving. It uses multi-view camera data to learn a transformer model that directly projects the perception of the road environment onto the BEV perspective. However, training a transformer model often requires a large amount of data, and as camera data for road traffic are often private, they are typically not shared. Federated learning offers a solution that enables clients to collaborate and train models without exchanging data but model parameters. In this paper, we introduce FedBEVT, a federated transformer learning approach for BEV perception. In order to address two common data heterogeneity issues in FedBEVT: (i) diverse sensor poses, and (ii) varying sensor numbers in perception systems, we propose two approaches - Federated Learning with Camera-Attentive Personalization (FedCaP) and Adaptive Multi-Camera Masking (AMCM), respectively. To evaluate our method in real-world settings, we create a dataset consisting of four typical federated use cases. Our findings suggest that FedBEVT outperforms the baseline approaches in all four use cases, demonstrating the potential of our approach for improving BEV perception in autonomous driving.
17.
Song, R.; Xu, R.; Festag, A.; Ma, J.; Knoll, A
FedBEVT: Federated Learning Bird's Eye View Perception Transformer in Road Traffic Systems Journal Article
In: arXiv, 2023, (arXiv preprint arXiv:2304.01534).
Abstract | Links | BibTeX | Tags: federated learning, Intelligent Transport Systems
@article{Song:ARXIV:2023-FedBEVT,
title = {FedBEVT: Federated Learning Bird's Eye View Perception Transformer in Road Traffic Systems},
author = {R. Song and R. Xu and A. Festag and J. Ma and A Knoll},
url = {https://arxiv.org/pdf/2304.01534},
year = {2023},
date = {2023-04-08},
urldate = {2023-04-04},
journal = {arXiv},
abstract = {Bird's eye view (BEV) perception is becoming increasingly important in the field of autonomous driving. It uses multi-view camera data to learn a transformer model that directly projects the perception of the road environment onto the BEV perspective. However, training a transformer model often requires a large amount of data, and as camera data for road traffic are often private, they are typically not shared. Federated learning offers a solution that enables clients to collaborate and train models without exchanging data but model parameters. In this paper, we introduce FedBEVT, a federated transformer learning approach for BEV perception. In order to address two common data heterogeneity issues in FedBEVT: (i) diverse sensor poses, and (ii) varying sensor numbers in perception systems, we propose two approaches - Federated Learning with Camera-Attentive Personalization (FedCaP) and Adaptive Multi-Camera Masking (AMCM), respectively. To evaluate our method in real-world settings, we create a dataset consisting of four typical federated use cases. Our findings suggest that FedBEVT outperforms the baseline approaches in all four use cases, demonstrating the potential of our approach for improving BEV perception in autonomous driving.},
note = {arXiv preprint arXiv:2304.01534},
keywords = {federated learning, Intelligent Transport Systems},
pubstate = {published},
tppubtype = {article}
}
Bird's eye view (BEV) perception is becoming increasingly important in the field of autonomous driving. It uses multi-view camera data to learn a transformer model that directly projects the perception of the road environment onto the BEV perspective. However, training a transformer model often requires a large amount of data, and as camera data for road traffic are often private, they are typically not shared. Federated learning offers a solution that enables clients to collaborate and train models without exchanging data but model parameters. In this paper, we introduce FedBEVT, a federated transformer learning approach for BEV perception. In order to address two common data heterogeneity issues in FedBEVT: (i) diverse sensor poses, and (ii) varying sensor numbers in perception systems, we propose two approaches - Federated Learning with Camera-Attentive Personalization (FedCaP) and Adaptive Multi-Camera Masking (AMCM), respectively. To evaluate our method in real-world settings, we create a dataset consisting of four typical federated use cases. Our findings suggest that FedBEVT outperforms the baseline approaches in all four use cases, demonstrating the potential of our approach for improving BEV perception in autonomous driving.
18.
Delooz, Q.; Festag, A.; Vinel, A.; Lobo, S.
Simulation-Based Performance Optimization of V2X Collective Perception by Adaptive Object Filtering Proceedings Article
In: IEEE Intelligent Vehicles Symposium (IV), Anchorage, Alaska, US, 2023.
Abstract | Links | BibTeX | Tags: Cooperative ITS, Intelligent Transport Systems, sensor data sharing, V2X communication
@inproceedings{Delooz:IV:2023,
title = {Simulation-Based Performance Optimization of V2X Collective Perception by Adaptive Object Filtering},
author = {Q. Delooz and A. Festag and A. Vinel and S. Lobo},
url = {https://ieeexplore.ieee.org/document/10186788},
doi = {10.1109/IV55152.2023.10186788},
year = {2023},
date = {2023-04-04},
urldate = {2023-04-04},
booktitle = {IEEE Intelligent Vehicles Symposium (IV)},
address = {Anchorage, Alaska, US},
abstract = {V2X Collective Perception is the principle of exchanging sensor data among V2X-capable stations, such as vehicles or roadside units, by exchanging lists of perceived objects in the 5.9 GHz frequency band for road safety and traffic efficiency. An object can be anything relevant to traffic safety, e.g., vehicles or pedestrians. The current standardization of Collective Perception in Europe considers filtering objects for transmission based on their locally perceived dynamics and freshness to preserve channel resources. However, two remaining problems of object filtering are: information redundancy and adapting object filtering to the available channel resources. In this paper, we combine redundancy mitigation and congestion control-aware filtering. We evaluate the performance of the resulting object filtering techniques by realizing realistic, large-scale simulations of a mid-size city in Germany. We assess the performance using a scoring metric. The results show better information redundancy control and adjustable channel usage for object filtering.},
keywords = {Cooperative ITS, Intelligent Transport Systems, sensor data sharing, V2X communication},
pubstate = {published},
tppubtype = {inproceedings}
}
V2X Collective Perception is the principle of exchanging sensor data among V2X-capable stations, such as vehicles or roadside units, by exchanging lists of perceived objects in the 5.9 GHz frequency band for road safety and traffic efficiency. An object can be anything relevant to traffic safety, e.g., vehicles or pedestrians. The current standardization of Collective Perception in Europe considers filtering objects for transmission based on their locally perceived dynamics and freshness to preserve channel resources. However, two remaining problems of object filtering are: information redundancy and adapting object filtering to the available channel resources. In this paper, we combine redundancy mitigation and congestion control-aware filtering. We evaluate the performance of the resulting object filtering techniques by realizing realistic, large-scale simulations of a mid-size city in Germany. We assess the performance using a scoring metric. The results show better information redundancy control and adjustable channel usage for object filtering.
19.
Delooz, Q.; Vinel, A.; Festag, A.
Optimizing the Channel Resource Usage for Sensor Data Sharing with V2X Communications Journal Article
In: at – Automatisierungstechnik, vol. 71, no. 4, pp. 311-317, 2023.
Abstract | Links | BibTeX | Tags: Cooperative ITS, Intelligent Transport Systems, sensor data sharing, V2X communication
@article{Delooz:ATJournal:2023,
title = {Optimizing the Channel Resource Usage for Sensor Data Sharing with V2X Communications},
author = {Q. Delooz and A. Vinel and A. Festag},
url = {2023-02-21},
doi = {10.1515/auto-2022-0162},
year = {2023},
date = {2023-04-01},
urldate = {2023-02-21},
journal = {at – Automatisierungstechnik},
volume = {71},
number = {4},
pages = {311-317},
abstract = {Sensor data sharing in V2X communication enables vehicles to exchange locally perceived sensor data with each other to increase their environmental awareness. It relies on the periodic exchange of selected, safety-relevant objects. Object selection is used to reduce channel resource usage. Additionally, vehicles use congestion control mechanisms to avoid overloading the channel. Currently, both object selection and congestion control mechanisms operate independently. We study a congestion-aware object filtering approach combining both and improving the performance of sensor data sharing. Die Übertragung von Sensordaten mit V2X-Kommunikation ermöglicht Fahrzeugen, lokale Umgebungsinformationen auszutauschen, um die Wahrnehmungsreichweite zu erhöhen. Der Sensordatenaustausch basiert auf der periodischen Übertragung sicherheitsrelevanter Objekte. Dabei wird die Anzahl der Objekte reduziert, um die Datenlast zu verringern. Zusätzlich steuern Mechanismen die Datenlast um eine Überlast zu vermeiden. Bisher arbeiten die Objektauswahl und die Datenüberlaststeuerung unabhängig voneinander. Wir untersuchen einen kombinierten Ansatz zur Objektfilterung, der die Performanz des Sensordatenaustauschs verbessert.},
keywords = {Cooperative ITS, Intelligent Transport Systems, sensor data sharing, V2X communication},
pubstate = {published},
tppubtype = {article}
}
Sensor data sharing in V2X communication enables vehicles to exchange locally perceived sensor data with each other to increase their environmental awareness. It relies on the periodic exchange of selected, safety-relevant objects. Object selection is used to reduce channel resource usage. Additionally, vehicles use congestion control mechanisms to avoid overloading the channel. Currently, both object selection and congestion control mechanisms operate independently. We study a congestion-aware object filtering approach combining both and improving the performance of sensor data sharing. Die Übertragung von Sensordaten mit V2X-Kommunikation ermöglicht Fahrzeugen, lokale Umgebungsinformationen auszutauschen, um die Wahrnehmungsreichweite zu erhöhen. Der Sensordatenaustausch basiert auf der periodischen Übertragung sicherheitsrelevanter Objekte. Dabei wird die Anzahl der Objekte reduziert, um die Datenlast zu verringern. Zusätzlich steuern Mechanismen die Datenlast um eine Überlast zu vermeiden. Bisher arbeiten die Objektauswahl und die Datenüberlaststeuerung unabhängig voneinander. Wir untersuchen einen kombinierten Ansatz zur Objektfilterung, der die Performanz des Sensordatenaustauschs verbessert.
20.
Hegde, A.; Lobo, S.; Festag, A.
Cellular-V2X for Vulnerable Road User Protection in Cooperative ITS Proceedings Article
In: 2022 18th International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob), Thessaloniki, Greece, 2023.
Abstract | Links | BibTeX | Tags: Intelligent Transport Systems, safety, V2X communication, vulnerable road users
@inproceedings{Hegde:WiMob:2022,
title = {Cellular-V2X for Vulnerable Road User Protection in Cooperative ITS},
author = {A. Hegde and S. Lobo and A. Festag},
url = {http://www.wimob.org/wimob2022},
doi = {10.1109/WiMob55322.2022.9941707},
year = {2023},
date = {2023-02-21},
urldate = {2022-10-12},
booktitle = {2022 18th International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob)},
address = {Thessaloniki, Greece},
abstract = {Cooperative Intelligent Transport Systems (C-ITS) play a significant role in improving road traffic safety and efficiency. Primary use cases rely on vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication. In C-ITS, the safety of vulnerable road users (VRUs) such as pedestrians, motorists and other users with reduced mobility is being increasingly considered. A warning system, where VRUs actively send and receive messages instead of being passively monitored as road traffic objects, can play an important role in detecting risky situations and allowing the warning of vehicle drivers. However, in a dense urban traffic scenario with closely moving vehicles and pedestrians, the communication network can face severe resource constraints. Considering Cellular-V2X communication systems, the user equipments (UEs) may have to use the unmanaged mode of the sidelink interface. This paper analyzes the limitations of the existing radio resource allocation mechanisms and proposes adaptive strategies to ensure fairness in the distribution of radio resources between vehicles and VRUs. In addition, this work addresses the question about the situations in which VRU safety, being subject to a resource-constrained Cellular-V2X network, can be ensured.},
keywords = {Intelligent Transport Systems, safety, V2X communication, vulnerable road users},
pubstate = {published},
tppubtype = {inproceedings}
}
Cooperative Intelligent Transport Systems (C-ITS) play a significant role in improving road traffic safety and efficiency. Primary use cases rely on vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication. In C-ITS, the safety of vulnerable road users (VRUs) such as pedestrians, motorists and other users with reduced mobility is being increasingly considered. A warning system, where VRUs actively send and receive messages instead of being passively monitored as road traffic objects, can play an important role in detecting risky situations and allowing the warning of vehicle drivers. However, in a dense urban traffic scenario with closely moving vehicles and pedestrians, the communication network can face severe resource constraints. Considering Cellular-V2X communication systems, the user equipments (UEs) may have to use the unmanaged mode of the sidelink interface. This paper analyzes the limitations of the existing radio resource allocation mechanisms and proposes adaptive strategies to ensure fairness in the distribution of radio resources between vehicles and VRUs. In addition, this work addresses the question about the situations in which VRU safety, being subject to a resource-constrained Cellular-V2X network, can be ensured.