Automotive network switch with anomaly detection

1. An automotive network switch, comprising:

a plurality of ports, configured to receive packets from electronic subsystems of a vehicle over a computer network deployed in the vehicle, and to transmit the packets to other electronic subsystems of the vehicle over the computer network;

a switch core, configured to receive the packets from one or more of the ports, to forward the packets to at least one of the ports, and to transmit the packets over network links of the computer network; and

one or more processors, configured to:

obtain at least some of the packets processed by the switch core;

analyze the obtained packets to identify an anomaly in one or more of the electronic subsystems of the vehicle;

generate a health score, comprising a value that indicates a severity of the anomaly in a range of severity levels; and

send a notification of the anomaly, including the health score, over the computer network to a central processor that is external to the automotive network switch.

2. The automotive network switch according to claim 1 , wherein, in analyzing the obtained packets, the one or more processors are configured to identify an existing anomaly or to predict an imminent anomaly.

3. The automotive network switch according to claim 1 , wherein at least one of the electronic subsystem comprises a sensor, wherein the ports are configured to receive at least some of the packets from the sensor, and wherein the one or more processors are configured to identify an abnormal functioning of the sensor.

4. The automotive network switch according to claim 1 , wherein the one or more processors are configured to identify the anomaly in response to detecting a statistically-deviant pattern of attributes of the packets.

5. The automotive network switch according to claim 4 , wherein the statistically-deviant pattern comprises one or more of:

a statistically-deviant dropping of packets;

a statistically-deviant packet rate;

a statistically-deviant latency; and

a statistically-deviant inter-packet spacing.

6. The automotive network switch according to claim 1 , wherein the one or more processors are configured to run a Machine-Learning model configured to detect the anomaly.

7. The automotive network switch according to claim 1 , wherein the one or more processors are configured to count events relating to the packets, and to detect the anomaly based on the counted events.

8. The automotive network switch according to claim 1 , wherein at least one of the one or more processors comprises a Machine-Learning (ML) accelerator that is separate from a switch Central Processing Unit (CPU) or switch Graphics Processing Unit (GPU) disposed in the switch.

9. The automotive network switch according to claim 1 , wherein the one or more processors are configured to select one or more packet streams from among the packets processed by the switch, in accordance with a predefined selection criterion, and to identify the anomaly by analyzing the packets in the selected packet streams.

10. The automotive network switch according to claim 9 , wherein the one or more processors are configured to select the one or more packet streams by performing on one or more of (i) examining one or more header fields of one or more of the packets, and (ii) applying Deep Packet Inspection (DPI) to one or more of the packets.

11. The automotive network switch according to claim 1 , wherein the health score comprises a numerical value that quantifies the severity of the anomaly.

12. A method for anomaly detection in a vehicle, the method comprising:

in an automotive network switch in the vehicle, receiving packets from electronic subsystems of the vehicle over a computer network deployed in the vehicle, and transmitting the packets to other electronic subsystems of the vehicle over the computer network;

analyzing, by the automotive network switch, at least some of the packets processed by the automotive network switch to identify an anomaly in one or more of the electronic subsystems of the vehicle;

generating a health score, comprising a value that indicates a severity of the anomaly in a range of severity levels; and

sending from the automotive network switch a notification of the anomaly, including the health score, over the computer network to a central processor that is external to the automotive network switch.

13. The method according to claim 12 , wherein analyzing the packets comprises identifying an existing anomaly or predicting an imminent anomaly.

14. The method according to claim 12 , wherein receiving the packets comprises receiving at least some of the packets from a sensor, and wherein analyzing the packets comprises identifying an abnormal functioning of the sensor.

15. The method according to claim 12 , wherein analyzing the packets comprises identifying the anomaly in response to detecting a statistically-deviant pattern of attributes of the packets.

16. The method according to claim 12 , wherein analyzing the packets comprises running a Machine-Learning model configured to detect the anomaly.

17. The method according to claim 12 , wherein analyzing the packets comprises counting events relating to the packets, and detecting the anomaly based on the counted events.

18. The method according to claim 12 , wherein analyzing the packets is performed, at least partially, by a Machine-Learning (ML) accelerator that is separate from a switch Central Processing Unit (CPU) or switch Graphics Processing Unit (GPU) disposed in the switch.

19. The method according to claim 12 , wherein analyzing the packets comprises selecting one or more packet streams from among the packets processed by the switch, in accordance with a predefined selection criterion, and identifying the anomaly by analyzing the packets in the selected packet streams.

20. The method according to claim 12 , wherein the health score comprises a numerical value that quantifies the severity of the anomaly.