Methods to detect spoofing attacks on automated driving systems

Patent metadata

Field	Value
Publication number	US-11310269-B2
Application number	US-201916653711-A
Country	US
Kind code	B2
Filing date	Oct 15, 2019
Priority date	Oct 15, 2019
Publication date	Apr 19, 2022
Grant date	Apr 19, 2022

Abstract

Systems and methods are disclosed for an ADV to leverage pre-defined static objects along a planned route of travel to detect and counter attacks that attempt to change the destination or the planned route. The ADV may detect updates to the static objects if the planned route is changed. Based on the updated static objects, the ADV determines if there is an abnormal re-routing of the planned route or if there is a new route due to a suspicious destination change. The ADV may also leverage the static objects to detect spoofing attacks against the sensor system. The ADV may evaluate if sensors of the sensor system are able to detect and identify the static objects to identify an impaired sensor. The ADV may perform cross-check on the ability of the sensors to detect and identify dynamic objects to gain confidence that the impaired sensor is due to spoofing attacks.

First claim

What is claimed is:

1. A computer-implemented method for detecting a spoofing attack against an autonomous driving vehicle (ADV), the method comprising:
determining a planned route from a starting location to a destination, the planned route to the destination being updatable as the ADV travels on the planned route;
obtaining a plurality of pre-defined static objects along the planned route from a map database, the plurality of pre-defined static objects being updatable when the planned route is updated;
detecting an update to the plurality of pre-defined static objects;
determining that a re-routing of the planned route is associated with the spoofing attack based on a frequency of past re-routings of the planned route, and a distance of the planned route in response to detecting the update to the plurality of pre-defined static objects;
determining a change to the destination by analyzing a new route traced by a plurality of updated pre-defined static objects;
confirming that a passenger in the ADV does not request the change to the destination to determine that the re-routinq is associated with the spoofinq attack;
confirming that there is the passenger in the ADV;
prompting the passenger in the ADV to request stopping the ADV; and
sending a report that the re-routing of the planned route is associated with the spoofing attack.

2. The method of claim 1 , further comprising : analyzing one or more of a traffic condition, a difference in distances between the abnormal re-routing and the planned route, a travel time of the planned route, or a difference in travel times between the abnormal re-routing and the planned route to determine that the re-routing of the planned route is associated with the spoofing attack.

3. The method of claim 1 , further comprising: verifying that an updated pre-defined static object is identified to confirm that the planned route is re-routed.

4. The computer-implemented of claim 1 , further comprising:
evaluating if each of a plurality of sensors of the ADV is able to identify one of the plurality of pre-defined static objects from the map when the ADV is at a location corresponding to a pre-defined static object;
evaluating if each of the plurality of sensors is able to identify a dynamic object along the planned route; and
identifying one or more impaired sensors based on evaluations of the one or more impaired sensors not being able to identify the pre-defined static object or the dynamic object; and
excluding the one or more impaired sensors from a perception function of the ADV.

5. The computer-implemented method of claim 1 , wherein the determining that the re-routing of the planned route is associated with the spoofing attack is based on a difference in the distance from a previous route.

6. The computer-implemented method of claim 1 , wherein the determining that the re-routing of the planned route is associated with the spoofing attack is based on a travel time of the new route.

7. The computer-implemented method of claim 1 , wherein the determining that the re-routing of the planned route is associated with the spoofing attack is based on a traffic condition.

8. A non-transitory machine-readable medium having instructions stored therein, which when executed by a processor, cause the processor to perform operations for detecting a spoofing attack against an autonomous driving vehicle (ADV), the operations comprising:
determining a planned route from a starting location to a destination, the planned route to the destination being updatable as the ADV travels on the planned route;
obtaining a plurality of pre-defined static objects along the planned route from a map database, the plurality of pre-defined static objects being updatable when the planned route is updated;
detecting an update to the plurality of pre-defined static objects;
determining that a re-routing of the planned route is associated with the spoofing attack based on a frequency of past re-routings of the planned route, and a distance of the planned route in response to detecting the update to the plurality of pre-defined static objects;
determining a change to the destination by analyzing a new route traced by a plurality of updated pre-defined static objects; and
confirming that a passenger in the ADV does not request the change to the destination to determine that the re-routing is associated with the spoofing attack;
confirming that there is the passenger in the ADV;
prompting the passenger in the ADV to request stopping the ADV; and
sending a report that the re-routing of the planned route is associated with the spoofing attack.

9. The non-transitory machine-readable medium of claim 8 , further comprising instructions, which when executed by the processor, cause the processor to perform operations comprising:
analyzing one or more of a traffic condition, a difference in distances between the abnormal re-routing and the planned route, a travel time of the planned route, or a difference in travel times between the abnormal re-routing and the planned route to determine that the re-routing of the planned route is associated with the spoofing attack.

10. The non-transitory machine-readable medium of claim 8 , further comprising instructions, which when executed by the processor, cause the processor to perform operations comprising :
verifying that an updated pre-defined static object is identified to confirm the planned route is re-routed.

11. The non-transitory machine-readable medium of claim 8 , further comprising instructions, which when executed by the processor, cause the processor to perform operations comprising:
evaluating if each of a plurality of sensors of the ADV is able to identify one of the plurality of pre-defined static objects from the map when the ADV is at a location corresponding to a pre-defined static object;
evaluating if each of the plurality of sensors is able to identify a dynamic object along the planned route; and
identifying one or more impaired sensors based on evaluations of the one or more impaired sensors not being able to identify the pre-defined static object or the dynamic object; and
excluding the one or more impaired sensors from a perception function of the ADV.

12. The non-transitory machine-readable medium of claim 8 , wherein the determining that the re-routing of the planned route is associated with the spoofing attack is based on a difference in the distance from a previous route.

13. The non-transitory machine-readable medium of claim 8 , wherein the determining that the re-routing of the planned route is associated with the spoofing attack is based on a difference in the distance from a previous route.

14. The non-transitory machine-readable medium of claim 8 , wherein the determining that the re-routing of the planned route is associated with the spoofing attack is based on a traffic condition.

15. A system to detect a spoofing attack against an autonomous driving vehicle (ADV), comprising:
a memory; and
a processor coupled to the memory, wherein the processor is configured to: determine a planned route from a starting location to a destination, the planned route to the destination being updatable as the ADV travels on the planned route;
obtain a plurality of pre-defined static objects along the planned route from a map database, the plurality of pre-defined static objects being updatable when the planned route is updated;
detect an update to the plurality of pre-defined static objects;
determine that a re-routing of the planned route is associated with the spoofing attack based on a frequency of past re-routings of the planned

Assignees

• Baidu Usa Llc

Inventors

• Liu Xiaodong
• Qu Ning

Classifications

• H04L67/52

  specially adapted for the location of the user terminal · CPC title

• G06Q10/047

  Optimisation of routes or paths, e.g. travelling salesman problem · CPC title

• H04L67/12

  specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks · CPC title

• G06F21/554

  involving event detection and direct action · CPC title

• G06V20/58

  Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads · CPC title