An automatic driving anti-hacking attack monitoring method

CN115366824BActive Publication Date: 2026-09-15SHENZHEN JILIAN TECH CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202210791610.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-07-07
Publication Date
2026-09-15
Estimated Expiration
2042-07-07

AI Technical Summary

Technical Problem

另外一种是通过CAN协议网络,向自动驾驶汽车传送大量杂乱数据,使得车辆通信信道拥堵,无法和车联网系统正常交互,同时降低车辆计算性能,从而影响车辆快速应对道路异常状况的能力

Benefits of technology

[0027] This invention can identify anomalies in autonomous vehicle data and assess whether there are potential safety hazards. It uses a hacker hijacking detection system to determine if the autonomous driving system has been hijacked and assesses the level of hijacking. Different response strategies are determined based on the level of hijacking, allowing for timely intervention and ensuring the safety of passengers and other road vehicles. A vehicle-to-everything (V2X) system is used to provide safety warnings to surrounding connected vehicles, adjust driving plans, and control information interaction. This prevents more serious road consequences caused by untimely assessments or inaccurate notifications.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115366824B_ABST
    Figure CN115366824B_ABST
Patent Text Reader

Abstract

The application provides an automatic driving anti-hacking attack monitoring method, comprising: detecting automatic driving vehicle data anomaly, wherein the detection of automatic driving vehicle data anomaly specifically comprises: automatic driving vehicle sensor data anomaly, automatic driving vehicle control system anomaly, and automatic driving vehicle network data interaction anomaly; starting a driving vehicle safety mode; issuing a hacking hijacking detection for a driving data abnormal vehicle; analyzing a hacking hijacking level through a hijacking detection; prewarning and scheduling surrounding vehicles according to the hijacking level; requesting to intercept a vehicle for interception for a vehicle judged to be dangerous; and notifying relevant personnel to assist in intercepting the vehicle.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a monitoring method for preventing hacker attacks on autonomous vehicles. Background Technology

[0002] Hacking attacks on autonomous vehicles could lead to vehicle hijacking or reduced data processing performance, decreasing the ability to handle abnormal road conditions and endangering the lives and property of passengers and other road users. Currently, the communication between vehicle components uses the CAN protocol, which makes the vehicle network and sensor system vulnerable to hacking. Two common types of hacking attacks targeting autonomous vehicles are: one involves remotely accessing the vehicle control system via wireless transmission networks, infiltrating internal sensors and positioning systems to obtain erroneous data, thereby altering the vehicle's route or control functions; the other involves transmitting large amounts of noisy data to the autonomous vehicle via the CAN protocol network, congesting the vehicle's communication channels, preventing normal interaction with the vehicle-to-everything (V2X) system, and reducing the vehicle's computing performance, thus affecting its ability to quickly respond to abnormal road conditions. When a vehicle is hijacked, due to sensor hijacking or reduced computing performance, the autonomous driving system alone cannot determine the extent of the hijacking. Third-party systems such as V2X systems are needed for auxiliary detection to determine the normal operation of the vehicle's computing and control systems. To reduce road accidents caused by hacking, V2X systems need to have a mechanism to handle different hijacking scenarios for autonomous vehicles. The autonomous driving anti-hacking detection system proactively compares vehicle network monitoring data and vehicle operation data during autonomous vehicle operation to detect abnormal autonomous vehicles and assign an abnormal vehicle rating. For detected hijacked vehicles, appropriate countermeasures must be taken, and hijacking vulnerabilities must be promptly investigated. For vehicles under normal control, priority must be given to ensuring passenger safety and completing the driving mission. For vehicles out of control, emergency response departments must be notified to intercept the vehicle and prevent accidents. Summary of the Invention

[0003] This invention provides a monitoring method for anti-hacking attacks on autonomous driving systems, mainly including:

[0004] Detect abnormal data from autonomous vehicles; activate the vehicle's safety mode; issue a hacker hijacking detection for vehicles with abnormal driving data; analyze the level of hacker hijacking if the hijacking detection fails; issue warnings and dispatch information to surrounding vehicles based on the hijacking level; request interceptor vehicles to intercept vehicles deemed particularly dangerous during hijacking; notify relevant personnel to assist in intercepting the vehicles.

[0005] Further, optionally, the detection of abnormal autonomous vehicle data includes:

[0006] When an autonomous vehicle begins driving, it proactively uploads data to the vehicle-to-everything (V2X) system via the network. The system acquires information such as the vehicle's license plate, brand and model, current location, destination, and route planning, and sets up traffic light data collection points along the route based on the route plan. During the journey, the V2X system monitors the vehicle's location, records changes in speed and acceleration, and other driving data via GPS or BeiDou satellite navigation. The vehicle also uses onboard sensors to measure speed, acceleration, and other driving data. When the autonomous vehicle passes a traffic light data collection point, it sends its driving data to the collection point. The system compares the data acquired by the V2X system with the data proactively uploaded by the vehicle. If there is a conflict between the V2X data and the data returned by the vehicle, or if there are anomalies in the returned data, a safety hazard is identified, including: abnormal sensor data of the autonomous vehicle; abnormal control system of the autonomous vehicle; and abnormal network data interaction of the autonomous vehicle.

[0007] The abnormal sensor data from the autonomous vehicle specifically includes:

[0008] The vehicle-to-everything (V2X) system compares sensor data from autonomous vehicles with standard driving data from vehicles of the same model to determine if the sensor data is within the normal range. If the comparison indicates that the sensor data is outside the normal range, the system first retrieves the vehicle's diagnostic report in real time to rule out vehicle malfunctions. If the diagnostic report shows the vehicle is normal, the system determines that there is a risk of hijacking in the vehicle's sensor system or control system; otherwise, the system must remind the owner to have the vehicle inspected and repaired promptly.

[0009] The abnormality in the autonomous vehicle control system specifically includes:

[0010] Analyze vehicle driving data and compare it with road driving regulation data to determine whether any illegal driving behavior occurred during the vehicle's operation. Input the data into a Naive Bayes classifier and use minimum error rate Bayesian decision-making for the determination.

[0011] If a vehicle exhibits behaviors such as continuous lane changes, multiple emergency accelerations and decelerations, and decreased braking capacity during operation, it is suspected that the vehicle control system may be malfunctioning. Comparing the acceleration and deceleration curves of the vehicle's current driving data with those of past driving data, if the curves show significant differences exceeding a set threshold, it is determined that the vehicle control system may be at risk of hijacking.

[0012] The abnormal network data interaction of the autonomous vehicle specifically includes:

[0013] A vehicle passes a traffic light data collection point, but fails to send data to the collection point as required, and the vehicle's response continuously times out, indicating an anomaly in the vehicle's data interaction. At this point, the vehicle-to-everything (V2X) system determines that the network connection between the system and the vehicle has been lost, classifying the vehicle as being in an abnormal network state, potentially indicating a network system malfunction and a risk of network hijacking. The V2X system obtains passenger information through vehicle registration, contacts passengers to activate safe mode, checks for damage to the vehicle's network modules, and performs vehicle hijacking detection. If the vehicle transmits data normally, but the data transmission packet loss rate exceeds a preset threshold or the vehicle's data integrity fails verification, the vehicle's network transmission is deemed unstable. At this point, the V2X system determines that the network connection between the system and the vehicle is unstable, classifying the vehicle as being in an abnormal network state and a risk of DDoS attack / network hijacking. The V2X system sends a data request to the vehicle to attempt to rebuild data exchange, and simultaneously issues a command to the vehicle to activate safe mode for hijacking detection.

[0014] Further optionally, activating the vehicle safety mode includes:

[0015] When the vehicle receives a command from the vehicle-to-everything (V2X) system or a passenger to switch to safe mode, it issues an abnormal warning to surrounding vehicles, indicating danger. The vehicle's autonomous driving system gradually engages the braking system, reduces the vehicle's power system, and decreases the vehicle's speed. It closes the vehicle's entertainment data port, opens the V2X data interaction port and emergency port, and refuses to accept access from unknown IP addresses and multimedia access requests. It monitors the vehicle's network communication channel data transmission and continuously returns vehicle data to the V2X system.

[0016] Further, optionally, the method of issuing a hacker hijacking detection for vehicles with abnormal driving data includes:

[0017] Upon detecting a vehicle with abnormal data, the vehicle-to-everything (V2X) network sends a hijacking detection command to the vehicle. Upon receiving the command, the autonomous vehicle notifies passengers of a potential security risk requiring hijacking detection and sends a temporary parking warning to surrounding vehicles. The V2X system then switches the vehicle to safe mode, pulls it over, and activates its hazard lights. If the vehicle does not comply with the command, it is determined that the vehicle has been hijacked, the hijacking level is escalated, and the emergency response team is notified. Once the vehicle stops as instructed and is in a safe state, the hijacking detection system checks for suspicious processes not belonging to the system's normal processes. If the system finds no suspicious processes, the abnormal situation is attributed to a DDoS attack on the multimedia network, causing network congestion, and the vehicle continues its mission in safe mode. If the system finds suspicious processes, it automatically terminates the abnormal processes and restarts the detection system. If the vehicle passes the re-detection, it continues its mission in safe mode.

[0018] Further, optionally, the hijacking detection failure analysis of the hacker hijacking level includes:

[0019] After the vehicle network system determines that the vehicle has been hijacked, it alerts the driver to a security risk and reports the hijacking detection results to the passengers. Based on the test results, the vehicle network system determines the level of hijacking and issues corresponding instructions to the autonomous vehicle. The autonomous vehicle operates its modules according to the detection level. For modules whose hijacking level does not affect safe driving, the system deactivates the automatic control of that module, switches it to manual control, disconnects the hijacking module from other modules, and prioritizes the passenger transport task. For modules whose hijacking level affects safe driving, the system completely stops the automatic control of the vehicle control system, disables the vehicle's power system, activates the vehicle's braking system, orders the vehicle to remain stationary, switches other vehicle modules to manual control mode, unlocks the doors, notifies the passengers of the hijacking level determination, reminds them to disembark to ensure their safety, and uploads the hijacking results to the emergency response department, prioritizing passenger safety and alerting the police.

[0020] Further optionally, the step of providing early warning and dispatching of surrounding vehicles based on the hijacking level includes:

[0021] The vehicle-to-everything (V2X) system initiates vehicle hijacking response measures based on the level of vehicle hijacking, and reports detailed information about the hijacked vehicle to the vehicle management department. If the hijacking level does not affect the safe driving of the vehicle, the driving task is completed first, and the hijacking is dealt with later. At the same time, a notification is broadcast to surrounding V2X vehicles, reminding them to keep a certain distance from the vehicle. If the hijacking level affects the safe driving of the vehicle and the vehicle control is lost, an emergency safety warning is broadcast to surrounding vehicles, requiring vehicles in the same lane and adjacent lanes of the hijacked vehicle to replan their routes and quickly move away from the hijacked vehicle.

[0022] Further optionally, the step of requesting an interceptor to intercept a vehicle deemed particularly dangerous during hijacking includes:

[0023] For autonomous vehicles that do not accept hijacking detection and ignore the stop command, the vehicle is identified as a hijacked dangerous vehicle, and surrounding connected vehicles are notified and an alarm is triggered; passengers in the hijacking danger situation are notified and guided to activate the braking system to reduce the vehicle speed; the connected vehicle system identifies and tracks the hijacked vehicle through GPS, GIS and road cameras, and at the same time sends a request to the nearest emergency response center to intercept the dangerous vehicle, requesting the dispatch of large vehicles for intercepting out-of-control autonomous vehicles; the out-of-control vehicle's driving route is analyzed, the interception destination is determined, the shortest route is planned, and sent to the interception vehicle center.

[0024] A monitoring method for anti-hacking attacks on autonomous driving systems is characterized in that the system comprises:

[0025] After receiving a request to intercept a vehicle, the interception center personnel determine the vehicle to be dispatched based on available vehicles and the interception distance. The vehicle-to-everything (V2X) system designs an interception plan based on road congestion and the hijacked vehicle's expected route, and establishes a voice channel between the intercepting vehicle and the hijacking vehicle to inform the passenger of the interception plan. The V2X system also notifies other vehicles on the planned interception route to give way to the emergency interception vehicle, ensuring that the intercepting vehicle arrives at the interception location as quickly as possible and begins the interception operation.

[0026] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0027] This invention can identify anomalies in autonomous vehicle data and assess whether there are potential safety hazards. It uses a hacker hijacking detection system to determine if the autonomous driving system has been hijacked and assesses the level of hijacking. Different response strategies are determined based on the level of hijacking, allowing for timely intervention and ensuring the safety of passengers and other road vehicles. A vehicle-to-everything (V2X) system is used to provide safety warnings to surrounding connected vehicles, adjust driving plans, and control information interaction. This prevents more serious road consequences caused by untimely assessments or inaccurate notifications. Attached Figure Description

[0028] Figure 1 This is a flowchart of a monitoring method for anti-hacking attacks on autonomous driving according to the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0030] Figure 1 This is a flowchart of a monitoring method for anti-hacking attacks on autonomous vehicles according to the present invention. Figure 1 As shown, this embodiment of a method for monitoring anti-hacking attacks on autonomous driving may specifically include:

[0031] Step 101: Detect abnormal data from autonomous vehicles.

[0032] When an autonomous vehicle begins operation, it proactively uploads data to the vehicle-to-everything (V2X) system via the network. The system acquires information such as the vehicle's license plate, brand and model, current location, destination, and route planning, and sets up traffic light data collection points along the route based on the plan. During operation, the V2X system monitors the vehicle's position, records changes in speed and acceleration, and other driving data via GPS or BeiDou satellite navigation. The vehicle's speed, acceleration, and other driving data are measured by onboard sensors. When the autonomous vehicle passes a traffic light data collection point, it sends its driving data to the collection point. The data acquired by the V2X system is compared with the data proactively uploaded by the vehicle. If there is a conflict between the V2X data and the vehicle's returned data, or if the returned data contains anomalies, a safety hazard is identified. For example, data receivers can be installed at traffic light intersections to receive the data uploaded by the vehicle. Alternatively, the vehicle can collect data through its built-in sensors, uploading this data to the V2X system each time it passes a traffic light, including sensor data and driving plans. If a vehicle actively uploads data showing that it is located at intersection A, but the GPS positioning system shows that the vehicle is at intersection B, then there is a data conflict between the two systems, and the vehicle may be abnormal.

[0033] The sensor data of the autonomous vehicle is abnormal.

[0034] The vehicle-to-everything (V2X) system compares sensor data from autonomous vehicles with standard driving data from vehicles of the same model to determine if the sensor data is within the normal range. If the comparison indicates that the sensor data is outside the normal range, the system first retrieves the vehicle's diagnostic report in real time to rule out vehicle malfunctions. If the diagnostic report shows the vehicle is normal, the system assesses a potential hijacking risk in the vehicle's sensor or control system; otherwise, it alerts the owner for timely inspection and repair. For example, if the vehicle's high beams are on, but the road is in urban areas with good lighting conditions, the V2X system will detect an anomaly. If the diagnostic report reveals a low beam malfunction, the system will promptly alert the owner to repair the headlights. Similarly, if the tire pressure sensor data shows abnormal tire pressure while the vehicle is in motion, and the sensor also indicates that a door is not closed, the V2X system will determine that the vehicle is malfunctioning.

[0035] The autonomous vehicle's control system is malfunctioning.

[0036] Analyze vehicle driving data and compare it with road driving regulation data to determine if any illegal driving behavior occurred during the vehicle's operation. The data is input into a Naive Bayes classifier, and a minimum error rate Bayesian decision is used for judgment. If the vehicle exhibits behaviors such as continuous lane changes, multiple emergency accelerations and decelerations, or decreased braking ability, it is judged that the vehicle control system may be abnormal. Compare the acceleration and deceleration curves of the current driving data with those of past driving data. If the curves show significant differences exceeding a set threshold, it is judged that the vehicle control system may be at risk of hijacking. For example, frequent abnormal acceleration and deceleration operations, or continuous speed exceeding the road speed limit, which do not conform to safe driving regulations, are also considered as potentially indicating a problem with the vehicle control system. Alternatively, if the system detects a significant increase in braking distance and a decrease in braking ability under the same level of braking command, it may also indicate a problem.

[0037] Anomalies occurred in the network data interaction of autonomous vehicles.

[0038] A vehicle passes a traffic light data collection point, but fails to send data as required, and its response times out continuously, indicating an anomaly in data interaction. At this point, the vehicle-to-everything (V2X) system determines that the network connection with the vehicle has been lost, classifying the vehicle as having a network anomaly, a potential network system malfunction, and a risk of network hijacking. The V2X system obtains passenger information through vehicle registration, contacts passengers to activate safe mode, checks for damage to the vehicle's network modules, and performs vehicle hijacking detection. If the vehicle transmits data normally, but the data transmission packet loss rate exceeds a preset threshold or the vehicle's data integrity fails verification, the system determines that the vehicle's network transmission is unstable. The V2X system then determines that the network connection between the system and the vehicle is unstable, classifying the vehicle as having a network anomaly and a risk of DDoS attack / network hijacking. The V2X system sends a data request to the vehicle to attempt to rebuild data exchange and simultaneously instructs the vehicle to activate safe mode for hijacking detection. For example, if a vehicle is at a traffic intersection or other vehicle data collection node, and the V2X system sends a data request but the vehicle's response times out for one minute, it indicates a network problem. The V2X system continues to send data requests to the vehicle, attempting to rebuild the data connection. Meanwhile, the vehicle-to-everything (V2X) system identifies the vehicle as having a network malfunction and continuously requests it to pull over for inspection. If the vehicle cannot receive the pull-over instruction due to a network malfunction, the V2X system uses passenger information to connect with the passengers by phone and requests them to pull over for inspection.

[0039] Step 102: Activate the vehicle's safety mode.

[0040] Upon receiving a command from the vehicle-to-everything (V2X) system or a passenger to switch to safe mode, the vehicle issues an anomaly warning to surrounding vehicles, indicating a potential hazard. The vehicle's autonomous driving system gradually engages the braking system, reducing the vehicle's power and speed. It closes the vehicle's entertainment data ports while opening the V2X data interaction ports and emergency ports, rejecting access from unknown IP addresses and multimedia access requests. It monitors vehicle network communication channel data transmission and continuously returns vehicle data to the V2X system. For example, upon receiving a vehicle safety mode command from the V2X system, the vehicle activates the braking module, reducing its speed. The vehicle alerts surrounding vehicles to the anomaly via Bluetooth or broadcast. The vehicle closes multimedia ports such as music and video, leaving only the V2X data ports and emergency ports open to reduce network data exchange overhead and data processing complexity. The vehicle remains connected to the V2X system, proactively sending vehicle information to it.

[0041] Step 103: Issue a hacker hijacking detection for vehicles with abnormal driving data.

[0042] Upon detecting a vehicle with abnormal data, the vehicle-to-everything (V2X) network sends a hijacking detection command to the vehicle. Upon receiving the command, the autonomous vehicle notifies passengers of a potential security risk requiring hijacking detection and sends a temporary parking alert to surrounding vehicles. The V2X system then switches the vehicle to safe mode, pulls it over, and activates its hazard lights. If the vehicle does not comply with the command, it is determined that the vehicle has been hijacked, the hijacking level is escalated, and the emergency response team is notified. Once the vehicle stops as instructed and is in a safe state, the hijacking detection system checks for suspicious processes not belonging to the system's normal processes. If the system finds no suspicious processes, the abnormal situation is attributed to a DDoS attack on the multimedia network, causing network congestion, and the vehicle continues its mission in safe mode. If the system finds suspicious processes, it automatically terminates the abnormal processes and restarts the detection system. If the vehicle passes the re-detection, it continues its mission in safe mode. For example, if an autonomous vehicle deviates from its pre-set path without notifying the vehicle network or passengers, the hijacking detection system will issue a hijacking detection request. The vehicle's autonomous driving system will then automatically pull over to the side of the road to conduct the hijacking detection. After ensuring the vehicle safely pulls over, data from both connected and disconnected states will be analyzed to determine if any data exceeds a reasonable threshold.

[0043] Step 104: If the hijacking detection fails, analyze the level of hacker hijacking.

[0044] After the vehicle-to-everything (V2X) system detects a hijacking, it alerts the driver to a security risk and reports the hijacking detection results to the passengers. Based on the test results, the V2X system determines the level of hijacking and issues corresponding instructions to the autonomous vehicle. The autonomous vehicle then operates its modules according to the detection level. For modules whose hijacking level does not affect safe driving, its automatic control is deactivated, switching to manual control. The hijacking module is disconnected from other modules, and the passenger transport mission is prioritized. For modules whose hijacking level affects safe driving, the automatic control of the vehicle's control system is completely stopped, the vehicle's power system is disabled, the braking system is activated, the vehicle is ordered to remain stationary, other vehicle modules are switched to manual control, the doors are unlocked, and the passengers are notified of the hijacking level assessment. Passengers are advised to disembark to ensure their safety, and the hijacking results are uploaded to emergency response departments, prioritizing passenger safety and triggering an alarm. For example, if the system detects that the hijacked functions are those that do not affect driving, such as the audio system or windows, it determines that the hijacking level is low and does not affect the safe operation of the vehicle. However, if the system detects that the hijacked functions are sensors, the vehicle's power system, the braking system, or functions that seriously affect driving safety, the autonomous driving system will forcibly activate the braking system, disconnect from the vehicle's power system, switch other modules of the vehicle to manual control mode, automatically unlock the doors, report the hijacking determination result via voice, guide passengers to get off the vehicle, and ensure passenger safety.

[0045] Step 105: Issue warnings and dispatch vehicles in the vicinity based on the hijacking level.

[0046] The vehicle-to-everything (V2X) system initiates vehicle hijacking response measures based on the level of vehicle hijacking, reporting detailed information about the hijacked vehicle to the vehicle management department. If the hijacking level does not affect safe driving, the driving task is completed first, and the hijacking is addressed. Simultaneously, a notification is broadcast to surrounding V2X vehicles, advising them to maintain a safe distance. If the hijacking level affects safe driving and control of the vehicle is lost, an emergency safety alert is broadcast to surrounding vehicles, requiring vehicles in the same and adjacent lanes to reroute and quickly move away from the hijacking vehicle. For example, if a vehicle's window is determined to be hijacked but does not affect safe driving, the driving task is completed first, while surrounding vehicles are alerted to the hijacking vehicle. If the vehicle's powertrain is hijacked, it is determined to severely impact safe driving. The V2X system then broadcasts an emergency notification to all vehicles near the hijacking vehicle on the current road, requiring them to quickly move away from the hijacking vehicle to ensure safety.

[0047] Step 106: If the hijacked vehicle is deemed to be relatively dangerous, request an interceptor to intercept it.

[0048] For autonomous vehicles that refuse hijacking detection and ignore stop commands, the vehicle is identified as a hijacked dangerous vehicle, and surrounding connected vehicles are notified and an alarm is triggered. Passengers in the hijacking danger situation are notified and guided to activate the braking system to reduce speed. The connected vehicle system identifies and tracks the hijacked vehicle using GPS, GIS, and road cameras, while simultaneously sending a request to the nearest emergency response center to intercept the dangerous vehicle, requesting the dispatch of large vehicles for intercepting out-of-control autonomous vehicles. The system analyzes the out-of-control vehicle's route, determines the interception destination, plans the shortest route, and sends it to the interception vehicle center. For example, if autonomous vehicle data shows hijacking anomalies, but the hijacking detection request cannot command the vehicle to stop, an emergency danger warning is broadcast to surrounding vehicles, alerting them to avoid the vehicle. The system first connects with the passengers in the hijacked vehicle, guiding them to manually activate the braking system to reduce speed. Simultaneously, the connected vehicle system continues to track the hijacked vehicle, reporting the hijacking information to the nearest emergency response center and issuing an interception request.

[0049] Step 107: Notify relevant personnel to assist in intercepting the vehicle.

[0050] Upon receiving a request to intercept a vehicle, the interception center determines the vehicle to dispatch based on available vehicles and the interception distance. The vehicle-to-everything (V2X) system designs an interception plan based on road congestion and the hijacked vehicle's projected route, establishing a voice channel between the intercepting vehicle and the hijacking vehicle to inform the passenger of the interception plan. The V2X system notifies other vehicles on the planned interception route to give way to the emergency interceptor, ensuring the interceptor reaches the interception location as quickly as possible to begin the interception operation. For example, after receiving a request to intercept a vehicle, the interception center calculates that the hijacked vehicle may continue straight for 500 meters. The system detects an available interceptor ahead of the hijacked vehicle and selects it for the interception mission. The interceptor plans its interception route and notifies all drivers on the predicted route, prompting them to replan their routes and give way. The intercepting personnel and the passenger maintain voice communication, and the passenger cooperates to complete the interception mission.

Claims

1. A monitoring method for anti-hacking attacks on autonomous driving systems, characterized in that, The method includes: To detect anomalies in autonomous vehicle data, when an autonomous vehicle begins driving, it proactively uploads data to the vehicle-to-everything (V2X) system via the network. The system acquires information such as the vehicle's license plate, brand and model, current location, destination, and route planning, and sets up traffic light data collection points along the route based on the plan. During operation, the V2X system monitors the vehicle's position, records changes in speed and acceleration, and other driving data via GPS or BeiDou satellite navigation. The vehicle also uses onboard sensors to measure speed, acceleration, and other driving data. When the autonomous vehicle passes a traffic light data collection point, it sends its driving data to the collection point. The system then compares the data acquired by the V2X system with the data proactively uploaded by the vehicle. If there is a conflict between the vehicle network data and the vehicle-returned data, or if there are anomalies in the vehicle-returned data, it is determined that the vehicle has a safety hazard. The detection of abnormal autonomous vehicle data specifically includes: abnormal autonomous vehicle sensor data, abnormal autonomous vehicle control system, and abnormal autonomous vehicle network data interaction. The abnormal autonomous vehicle sensor data specifically includes: the vehicle network system comparing the autonomous vehicle sensor data with the standard driving data of the same model vehicle to determine whether the vehicle sensor data is within the normal range; if the comparison data determines that the vehicle sensor data is outside the normal range, the vehicle's detection report is retrieved in real time to rule out the possibility of vehicle malfunction. Activate the vehicle's safety mode; If a vehicle with abnormal driving data is detected as a hacker, and the vehicle does not pull over as instructed, the vehicle is determined to be hacked and the hijacking level is increased. If the hijacking detection fails to analyze the level of hacker hijacking, the vehicle-to-everything (V2X) system determines the level of hijacking based on the test results and issues corresponding instructions to the autonomous vehicle. The autonomous vehicle then operates the vehicle modules according to the detection level. For modules whose hijacking level does not affect safe driving, the automatic control of the module is deactivated and switched to manual control. The hijacking module is disconnected from other modules, and the task of transporting passengers is then prioritized. For modules whose hijacking level affects safe driving, the automatic control of the vehicle control system is completely stopped, the vehicle's power system is prohibited from operating, the vehicle's braking system is activated, and the vehicle is ordered to remain stationary. Based on the hijacking level, provide early warnings and dispatch nearby vehicles; If a hijacked vehicle is deemed particularly dangerous, a request is made to intercept it. The vehicle-to-everything (V2X) system uses GPS, GIS, and road cameras to jointly identify and track the hijacked vehicle. Notify relevant personnel to assist in intercepting the vehicle, establish a voice channel between the intercepting vehicle and the hijacked vehicle, and inform the passengers of the interception plan.

2. The method according to claim 1, wherein, The detection of abnormal data from autonomous vehicles includes: Abnormal sensor data in autonomous vehicles; abnormal control system in autonomous vehicles; abnormal network data interaction in autonomous vehicles. The abnormal sensor data from the autonomous vehicle specifically includes: If the inspection report shows that the vehicle is normal, it is determined that there is a risk of hijacking in the vehicle's sensor system or vehicle control system; otherwise, the system should remind the owner to have the vehicle inspected and repaired in time. The abnormality in the autonomous vehicle control system specifically includes: Analyze vehicle driving data and compare it with road driving regulations data to determine whether there are any violations of driving rules during the vehicle's operation; input the data into a Naive Bayes classifier and use minimum error rate Bayes decision-making to make a judgment; If a vehicle exhibits behaviors such as continuous lane changes, multiple emergency accelerations and decelerations, and decreased braking ability while in motion, it is determined that there is an abnormality in the vehicle control system. Compare the acceleration and deceleration curves of the vehicle's current driving data with those of the past driving data. If there is a significant difference between the curves and the difference exceeds the set threshold, it is determined that the vehicle control system is at risk of hijacking. The abnormal network data interaction of the autonomous vehicle specifically includes: The vehicle passes through the traffic light data collection point, but fails to send data to the collection point as required, and the vehicle response times out continuously, indicating an anomaly in the vehicle's data interaction. At this point, the vehicle network system determines that the network connection between the vehicle and the vehicle has been lost, indicating that the vehicle is in an abnormal network state, the vehicle network system has malfunctioned, and the vehicle is at risk of network hijacking. The vehicle network system obtains passenger information through vehicle information registration, contacts the passengers to activate the safety mode, detects whether the vehicle network module is damaged, and performs vehicle hijacking detection. If the vehicle is transmitting data normally, but the data transmission packet loss rate exceeds a preset threshold or the vehicle data integrity fails the verification, the vehicle network transmission is determined to be unstable. At this time, the vehicle network system determines that the network connection between the system and the vehicle is unstable, and determines that the vehicle is in an abnormal network state. The vehicle is at risk of being attacked and hijacked by hackers using DDoS attacks. The vehicle network system sends a data request to the vehicle to attempt to rebuild the data exchange, and at the same time issues an instruction to the vehicle to enable safe mode and perform hijacking detection.

3. The method according to claim 1, wherein, Activating the vehicle safety mode includes: When the vehicle receives a command from the vehicle-to-everything (V2X) system or a passenger to switch to safe mode, it issues an abnormal warning to surrounding vehicles, indicating danger. The vehicle's autonomous driving system gradually engages the braking system, reduces the vehicle's power system, and decreases the vehicle's speed. It closes the vehicle's entertainment data port, opens the V2X data interaction port and emergency port, and refuses to accept access from unknown IP addresses and multimedia access requests. It monitors the vehicle's network communication channel data transmission and continuously returns vehicle data to the V2X system.

4. The method according to claim 1, wherein, The method of issuing a hacker hijacking detection for vehicles with abnormal driving data includes: After detecting a vehicle with abnormal data, the vehicle network sends a hacking hijacking detection command to the vehicle. After receiving the command, the autonomous vehicle notifies the passengers that there is a safety hazard in the vehicle and that a hijacking detection is required. It also sends information to surrounding vehicles to temporarily stop for abnormal vehicle detection. The vehicle network system controls the vehicle to enter safe mode, pulling it to the side of the road and activating its hazard lights. Once the vehicle is parked safely, the hacker hijacking detection system checks for suspicious processes not belonging to the system's normal processes. If the system detects no suspicious processes, the abnormal situation is attributed to a DDoS attack on the multimedia network, causing network congestion, and the vehicle continues driving in safe mode. If the system detects suspicious processes, it automatically terminates the abnormal processes and restarts the detection system. Once the vehicle passes the re-detection, it continues driving in safe mode.

5. The method according to claim 1, wherein, The hijacking detection fails to analyze the level of hacker hijacking, including: Once the vehicle network system determines that the vehicle has been hijacked, it will alert passengers to potential security risks and report the hijacking detection results to them. For modules where the hijacking level affects safe driving, it will also switch other vehicle modules to manual control mode, unlock the doors, notify passengers of the hijacking level assessment, remind passengers to get off the vehicle to ensure their safety, and upload the hijacking results to the emergency response department to prioritize passenger safety and call the police.

6. The method according to claim 1, wherein, The method of providing early warning and dispatching of surrounding vehicles based on the hijacking level includes: The vehicle-to-everything (V2X) system initiates vehicle hijacking response measures based on the level of vehicle hijacking, and reports detailed information about the hijacked vehicle to the vehicle management department. If the hijacking level does not affect the safe driving of the vehicle, the driving task is completed first, and the hijacking is dealt with later. At the same time, a notification is broadcast to surrounding V2X vehicles, reminding them to keep a certain distance from the vehicle. If the hijacking level affects the safe driving of the vehicle and the vehicle control is lost, an emergency safety warning is broadcast to surrounding vehicles, requiring vehicles in the same lane and adjacent lanes of the hijacked vehicle to replan their routes and quickly move away from the hijacked vehicle.

7. The method according to claim 1, wherein, The request to intercept vehicles deemed particularly dangerous during hijacking includes: For autonomous vehicles that do not accept hijacking detection and ignore the stop command, the vehicle is identified as a hijacked dangerous vehicle, and surrounding connected vehicles are notified and an alarm is triggered; passengers in the hijacking danger situation are notified and guided to activate the braking system to reduce the vehicle speed; at the same time, the connected vehicle system sends a request to the nearest emergency response center to intercept the dangerous vehicle, requesting the dispatch of large vehicles for intercepting out-of-control autonomous vehicles for support; the out-of-control vehicle's driving route is analyzed, the interception destination is determined, the shortest route is planned, and sent to the interception vehicle center.

8. The method according to claim 1, wherein, The notification to relevant personnel to assist in intercepting the vehicle includes: After receiving a request to intercept a vehicle, the personnel at the vehicle interception center will determine which vehicle to dispatch based on the availability of available vehicles and the interception distance. The vehicle-to-everything (V2X) system designs interception plans based on road congestion conditions and the expected route of the hijacked vehicle; The vehicle-to-everything (V2X) system notifies other vehicles on the planned interception route to give way to the emergency interception vehicle, ensuring that the interception vehicle reaches the interception location as quickly as possible and begins the interception operation.

Citation Information

Patent Citations

  • Anti-hacker hijacking method, device and equipment for unmanned vehicle as well as storage medium

    CN107194248A