Strategies for determining the trustworthiness of received data
The successful implementation of a trust model for external data in the V2X ecosystem is vital for future use cases like sensor data sharing. This can improve safety, efficiency and the overall experience. It enables advanced functions such as “green wave”, intersection management and priority switching for emergency vehicles. However, challenges in connection with the functional safety of vehicle functions in the automotive area need to be addressed to ensure the safety and reliability of V2X systems.
Local assessment of data
An assessment of the incoming data compared to the local environment model could lead to increased trust in the ego vehicle's model if the external data validates the model. The external data could then be used to add more details to the environment model. If external data deviates in some details from the ego vehicle's environment model, the ego vehicle could use the model with less confidence.
Application-related trust model
Depending on the data source, the ego vehicle could utilize various trust levels. Data from the manufacturer’s backend is usually highly trustworthy. In the V2X network, data from specific applications such as traffic lights could also be given a higher trust level. The prerequisite for this increased trust level are high security and integrity standards for data in the operation of these applications. Otherwise, attackers could manipulate these devices and could send fake data with a high trust level.
Log-based trust model
Functional safety requires a specific ASIL to use data for vehicle functions. The standards and regulations for functional safety ensure a common understanding of the requirements among vehicle manufacturers. When a vehicle sends data into the V2X network, it could add some information via the basic ASIL, under which the data has been created, and it could also provide a trust level for the correctness of the data. This would enable the ego vehicle to classify the received data and to assess its usability in a specific function or in a specific use case.
Entity-centric and data-centric trust models
The focus of entity-centric trust models is on assessing the reliability of individual devices, whereas data-centric trust models prioritize the reliability of the received data. To ensure an exact validation of both devices and data, these trust model rely on collaborative information from multiple sources, such as neighboring vehicles or traffic infrastructure. Combined or hybrid trust models are used to calculate the trustworthiness of data and assess the trustworthiness of data-transferring devices. The reliability of data is assessed using additional data collected from numerous vehicles. Functional trust and trust based on recommendations are used to assess the trustworthiness of an end device.
The underlying properties of the V2X network contradict the effort to establish trust between vehicles as the timespan in which the vehicles are in the same area is short and the vehicles regularly change their identity with pseudonymization. Therefore, traffic infrastructure could play an important role in such a trust model.
Advanced trust models
There are also studies and discussions on advanced confidence models:
- Bayesian inference-based trust models use probability theory to determine the trust level.
- Trust models based on deep and machine learning use algorithms that can recognize patterns in a huge volume data and can predict the future.
- By assigning different membership levels to different trust levels, trust models based on fuzzy logic manage ambiguity and inaccurate information.
- Blockchain-based trust models use the unchangeable and decentralized properties of the blockchain to create trust in a distributed environment.
External sensor data as a potential and challenge for future V2X use cases
In conclusion, the incorporation of external sensor data in future V2X (vehicle-to-everything) use cases has the potential to significantly enhance safety, efficiency, and the overall user experience. V2X systems can gather a more thorough and almost instantaneous understanding of the surroundings by incorporating information from external sensors like traffic cameras, weather monitors, and infrastructure sensors. This enables vehicles to access data beyond the reach of their sensors, empowering them to make well-informed decisions and take necessary actions to ensure the safety of fellow road users.
The availability of external sensor data can enable precise and advanced functions such as “green wave” automation, intersection management and priority switching for emergency vehicles. For example, by accessing traffic camera images, vehicles can receive accurate and up-to-date information on traffic congestion, road conditions and accidents to choose optimal routes and avoid potential hazards.
However, a successful implementation of external sensor data in V2X use cases requires overcoming challenges such as data reliability, security and data protection. Ensuring exactness, consistency and integrity of external sensor data sources is critical to maintaining the trustworthiness and reliability of V2X systems. Furthermore, stringent data security and data protection protocols should be implemented to safeguard sensitive information and to thwart unauthorized data access or manipulation.
In general, as sensor technology, connectivity, and artificial intelligence continue to advance, the incorporation of external sensor data in future V2X use cases has the potential to transform the interaction between vehicles and their surroundings. This transformation can
Your experts at msg
Are you engaged in the ever-evolving realm of V2X communication, connectivity, and autonomous driving? You undoubtedly comprehend the intricacy and difficulties associated with these technologies. msg's experts are available to offer guidance throughout your journey and cater to your unique requirements.
We possess vast knowledge and experience in systems engineering, encompassing architecture, safety, and security. Additionally, we specialize in testing advanced, widely distributed vehicle architectures. This expertise allows us to offer you comprehensive assistance in developing and implementing cloud-based remote ADAS solutions. We not only assist you in designing, but also in ensuring the protection of these functionalities to transform the concept of autonomous driving into a tangible reality.
Furthermore, we leverage our proficiency in machine learning and analytics to efficiently examine and harness the vast amounts of data produced throughout the development of these cutting-edge technologies. It is essential to have these capabilities in place to guarantee the reliability and safety of driving functions.
If you are encountering difficulties with connectivity, V2X, and ADAS, or if you simply wish to learn more about our services, please feel free to contact us. We are well-prepared to enhance your projects with our expertise and collaborate with you to shape the future of mobility.
