May 26, 2024

C-ITS allows vehicles to communicate with each other. Its development focuses on improving safety, comfort, traffic, and energy efficiency. Its main communication strategy is vehicle-to-everything (V2X), in which data sharing is performed through vehicle-to-vehicle communication (V2V) and vehicle-to-infrastructure communication (V2I). In a CITS environment, cooperative-awareness messages (CAMs). The decentralized environmental notifications (DENMs), on the other hand, provide real-time data about individual vehicles. [] provide information about events that could have an impact on traffic or road safety (e.g., roadworks, accidents, vehicle failures, etc.), allowing proactive incident prevention. Despite the novelty of the concept, it is still not possible to fully record and analyze the impact that C-ITS services have on road networks. Due to the C-ITS’s cooperative nature, its participants share a large amount of data. These data can reveal valuable information about the performance of the C-ITS system. Road and telecommunications operators, as well as the end users, could detect critical information, such as system failures or anomalies. These units are installed along the roadside and provide access to the V2I infrastructure for vehicles. These units are essential and most useful when there is a low density of C-ITS-equipped vehicles ]. RSUs also provide services to vehicles that are implemented in the infrastructure. RSUs provide services that are implemented on the infrastructure. ]:

  • Traffic light maneuver is a service that administers the generation and transmission of Signal Phase and Timing Messages. Its goal is to control access by vehicles to intersections and conflict areas. It provides safety information for cars at a meeting, including the current status of a traffic light as well as its future status and the time between them.
  • Road and lane Topology (RLT) is a service that manages the transmission and reception of Map Extended Messages. A MAPEM message contains a digital topology map of a particular area. This topology defines lanes, crossings, conflict zones, and allowed maneuvers.
  • Infrastructure to Vehicle Information (IVI) provides information on road signs, such as road warnings or contextual speed limits.

In low-density scenarios, a failure of an RSU may result in the unavailability of infrastructure services (or partial availability) and may even degrade the entire C-ITS system. In order to ensure the proper functioning of a system, RSUs need to be monitored and their failures identified and reported immediately. This study has a specific objective: to detect transmission failures in an RSU. The end user (vehicles) will be responsible for detecting the failure and informing road operators.

  • New method to detect RSU failures in vehicles.
  • Evaluation of the method using a real dataset generated by CAMs in a CITS naturalistic driving environment in three countries: France, Germany, and Italy.
  • Proposal of a C-ITS message that will be used to alert users to detected problems.

This paper is organized as follows: Section 2 The C-ITS is briefly introduced to help understand the role played by the communication stack over the CITS. Section 3 Presents our solution as well as the mechanisms that we use to detect RSU failures. Section 4 This section describes the message that must be sent by a vehicle to inform it of the RSU’s status. Section 5 presents the conclusions and future works.

State of the art

This section discusses various aspects of our work, including the C-ITS system, which we analyzed for this paper, the anomaly detection with relevant datasets, and the clustering algorithm used to classify data.

2.1. C-ITS Systems

In Europe, C-ITS is based on a communication stack that has been standardized and defined by ETSI. The Facilities layer was designed to interface efficiently with the Application layer (closer to the driver and vehicle sensors) over the Network layer. This layer provides a variety of messages to address a range of use cases, including traffic jam detection, road works warnings, traffic light controls, and logistics management. We have focused our study on a single message type: the CAM. The goal of the CAM is to promote a cooperative consciousness among vehicles. Its use is to give dynamic information about the car, such as its location, speed, and heading. A CAM is an intermittent message sent between 1 Hz and 10 Hz, depending on the vehicle’s speed. The message can be sent via V2V or V2I communication. Figure 1. Each car must have a pseudonym certification. The vehicle’s letters could be forwarded to distant vehicles using multi-hop forwarding. Figure 1. The general architecture for V2V communications is shown in Figure 2. Figure 2. A General Scheme for Vehicle-to-Infrastructure Communication. An RSU plays the role of the infrastructure as it handles all the received messages from the vehicles and runs the road operator’s computations, such as traffic management and event recording. It can also be used as a vehicle in the forwarding aspect, and it may even disseminate events to other RSUs in the network of the operator if those RSUs are within its range. The ETSI has defined the ITS G5 protocol stack. The protocol is the networking layer. In addition to the CAMs in our work, we use the geo-networking layers of the packets in order to access additional data, such as an accurate localization of stations. Ref. [ [ proposes a method to detect anomalies in the energy consumption of wireless equipment.

2.2. Anomaly detection

In ] Anomalies are further classified into three categories:

  1. Point anomalies which is when a data point is considered aberrant from the rest of the data.
  2. Contextual anomalies occur when a data instance is strange in a specific context.
  3. Collective anomalies, where a collection of related data instances may be strange but not each data instance.

Anomalies are further classified into three types: (a) point anomalies (when a particular data point is seen as being out of place with the rest of the set of data); (b) contextual anomalies (when a certain data instance is deemed to be abnormal in a given context); and (c), collective anomalies (when a group of data instances are deemed to be abnormal, even though the individual instances themselves may not have been uncommon). RSUs form a key component of a CITS. They are primarily responsible for V2I communications. The omnidirectional RSU antenna is supposed to provide equal radio transmission in all directions. However, obstacles like buildings, tunnels, or rivers can reduce the signal in some tips. This RSU is technically sound. A high attenuation of the movement in certain areas could be an indication of failure. A probabilistic model was used to detect antenna failure. The model was built using data collected from real field tests. Health assessment was done by comparing the behaviors of RSUs in terms of radio transmission. In 10 In [ ], an error detector for VANET systems that has some notions about signal attenuation is proposed. In 11 In order to determine the factors that affect a signal’s attenuation, numerical simulations have been performed. In line-of-sight (LOS) situations, a call will tend to weaken over long distances. However, in non-line of sight (NLOS), the presence or obstruction can have a significant impact on signal intensity (in decibels). The behavior of radio propagation depends on the geographic context. This was also noted in 12 [], where the communication range between a vehicle and an RSU is greater when the car moves toward the RSU. The authors also observed a relationship between the vehicle’s velocity and the range and explained it by the Doppler Effect. This is also followed in 13 ]. However, the opposite case can be observed. 14 DeepADV 15 [ is a VANET anomaly detection framework based on deep neural networks. The difference between an authentic and unconventional message is used to calculate a threshold and classify the messages. The algorithm will be used on RSUs to detect faulty messages. In ] 16, A VANET anomaly detection system using edge computing has been proposed. The faults were related to transmission omissions and detected by RSU-based edge networks and vehicular edge computing. These RSUs store information on a variety of vehicles and the number of packets collected. This data is then used to determine if an anomaly has occurred (a change in numbers). The strategy proved to be highly effective when tested on a simulation with high fault rates (25%). EVAD was a method that was proposed in 17 [To detect anomalies using edge computing. A correlation is made between sensor variables and is used to detect anomalies.


Leave a Reply

Your email address will not be published. Required fields are marked *