Automotive network switch with anomaly detection

By using embedded CPUs and machine learning accelerators in network switches to analyze packet flows, the problem of difficulty in detecting anomalies in electronic subsystems in automotive, industrial, and smart home systems in existing technologies is solved, enabling efficient fault identification and performance monitoring.

CN113994641BActive Publication Date: 2024-11-01INFINEON TECHNOLOGIES AMERICAS CORP
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202080045227.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2019-06-25
Filing Date
2020-06-24
Publication Date
2024-11-01
Estimated Expiration
2040-06-24

AI Technical Summary

Technical Problem

Existing communication systems struggle to efficiently detect and predict anomalies in electronic subsystems in automotive, industrial, and smart home systems, especially in short-range, high-data-rate communication environments, where they cannot promptly identify potential faults and performance degradation.

Method used

By deploying embedded CPUs and machine learning accelerators in network switches, anomalies can be identified by analyzing packet flows, generating health scores, and notifying the central processing unit, thus enabling distributed anomaly detection.

Benefits of technology

It improves the efficiency and predictive ability of detecting anomalies in electronic subsystems, reduces the burden on the central computer, simplifies switch operation, and integrates naturally with the architecture of modern automotive systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113994641B_ABST
    Figure CN113994641B_ABST
Patent Text Reader

Abstract

An automotive network switch (32) includes a plurality of ports (48), a switch core (52), and one or more processors (54, 56). The ports are configured to receive packets from an electronic subsystem (28) of a vehicle (24) through a computer network (20) deployed in the vehicle, and transmit packets to other electronic subsystems (28) of the vehicle through the computer network. The switch core is configured to receive packets from one or more of the ports, forward packets to at least one of the ports, and transmit packets through a network link of the computer network. The processor is configured to obtain at least some of the packets processed by the switch, analyze the obtained packets to identify anomalies in one or more of the electronic subsystems of the vehicle, and send a notification of the anomaly to a central processor (60) outside the switch through the computer network.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of U.S. Provisional Patent Application No. 62 / 866,285, filed on June 25, 2019, the disclosure of which is incorporated herein by reference. Technical Field

[0003] The present disclosure relates generally to communication systems, and in particular to methods and systems for anomaly detection in network switches. Background Art

[0004] Various applications such as automotive in-vehicle communication systems, certain industrial communication systems, and smart home systems require communication at high data rates over relatively short distances. Several types of protocols and communication media have been proposed for such applications. For example, in March 2015, "IEEE 802.3bw-2015-IEEE Standard for Ethernet Amendment 1: Physical Layer Specifications and Management Parameters for 100Mb / s Operation over a Single Balanced Twisted Pair Cable (100BASE-T1)" specifies Ethernet communication over a twisted pair copper wire medium.

[0005] The above description is presented as a general overview of the relevant art in the field and should not be construed as an admission that any of the information contained therein constitutes prior art to the present patent application. Summary of the invention

[0006] One embodiment described herein provides an automotive network switch, which includes multiple ports, a switch core, and one or more processors. The port is configured to receive packets from an electronic subsystem of a vehicle through a computer network deployed in the vehicle, and transmit packets to other electronic subsystems of the vehicle through the computer network. The switch core is configured to receive packets from one or more of the ports, forward packets to at least one of the ports, and transmit packets through a network link of the computer network. The one or more processors are configured to obtain at least some of the packets processed by the switch, analyze the obtained packets to identify anomalies in one or more of the electronic subsystems of the vehicle, and send a notification of the anomaly to a central processor outside the switch through the computer network.

[0007] In one embodiment, the one or more processors are configured to generate a health score that quantifies the severity of the anomaly and report the health score in the notification. In an example embodiment, when analyzing the acquired packets, the one or more processors are configured to identify an existing anomaly or predict an upcoming anomaly. In a disclosed embodiment, at least one of the electronic subsystems includes a sensor, the port is configured to receive at least some of the packets from the sensor, and the one or more processors are configured to identify abnormal function of the sensor.

[0008] In some embodiments, the one or more processors are configured to identify an anomaly in response to detecting a statistical deviation pattern of an attribute of a packet. In various embodiments, the statistical deviation pattern includes one or more of a statistical deviation packet drop, a statistical deviation packet rate, a statistical deviation delay, and a statistical deviation inter-packet interval.

[0009] In some embodiments, the one or more processors are configured to run a machine learning model configured to detect anomalies. In one embodiment, the one or more processors are configured to count events associated with the packet and detect anomalies based on the counted events. In the disclosed embodiments, at least one of the one or more processors includes a machine learning (ML) accelerator separate from a switch central processing unit (CPU) or a switch graphics processing unit (GPU) disposed in the switch.

[0010] In some embodiments, the one or more processors are configured to select one or more packet flows from among packets processed by the switch according to predefined selection criteria, and identify anomalies by analyzing packets in the selected packet flows. In an example embodiment, the one or more processors are configured to select the one or more packet flows by performing one or more of the following: (i) inspecting one or more header fields of one or more packets in the packets, and (ii) applying deep packet inspection (DPI) to one or more packets in the packets.

[0011] According to the embodiments described herein, a method for anomaly detection in a vehicle is further provided. The method includes: in an automotive network switch in the vehicle, receiving packets from an electronic subsystem of the vehicle through a computer network deployed in the vehicle, and transmitting packets to other electronic subsystems of the vehicle through the computer network. At least some of the packets processed by the switch are analyzed by the switch to identify anomalies in one or more of the electronic subsystems of the vehicle. A notification of the anomaly is sent from the switch to a central processor outside the switch through the computer network.

[0012] The present disclosure will be more fully understood through the following detailed description of its embodiments in conjunction with the accompanying drawings: BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 is a block diagram schematically illustrating an automotive communication system according to embodiments described herein;

[0014] Figure 2 is a diagram schematically illustrating the collection of packet-related events per port in a network switch according to an embodiment described herein; and

[0015] Figure 3 is a schematic diagram showing a method of performing a Figure 1 Flow chart of a method for anomaly detection using machine learning (ML) in a network switch of a communication system. DETAILED DESCRIPTION

[0016] Embodiments described herein provide improved methods and systems for condition monitoring for use in electronic systems that include subsystems that exchange communication packets with each other. The disclosed techniques identify existing or impending faults in electronic subsystems by detecting anomalies in the packets.

[0017] The embodiments disclosed herein are described in the context of automotive applications (e.g., systems that collect data from sensors or otherwise transmit data within a vehicle). However, this choice is merely for clarity. The disclosed techniques are equally applicable to other applications, such as industrial and / or smart home networks.

[0018] In some disclosed embodiments, a vehicle includes multiple electronic subsystems, such as sensors, various types of electronic control units (ECUs), advanced driver assistance systems (ADAS), in-vehicle infotainment (IVN) systems, a main central computer, etc. The electronic subsystems communicate via a computer network (e.g., an Ethernet network). The network includes multiple network switches that communicate with each other and the various subsystems via network links (e.g., twisted pair copper wire links).

[0019] In many practical scenarios, existing or impending faults in electronic subsystems can be identified by detecting anomalies in network packets transmitted or received over a network link. For example, a faulty sensor can be identified by detecting an excessively high rate of dropped packets or a statistically deviant packet rate.

[0020] In some embodiments described herein, the task of detecting anomalies in communication traffic is delegated to one or more network switches in the network. Typically, although not necessarily, anomaly detection is distributed among multiple network switches.

[0021] In one example embodiment, the network switch includes, among other components, one or more processors configured to analyze at least some of the packets processed by the switch and identify anomalies in one or more of the electronic subsystems of the vehicle based on the analyzed packets. When an anomaly is identified, for example, by statistical analysis of network traffic, the one or more processors are configured to send a notification of the anomaly to a central processor external to the switch via the network. The notification may include, for example, a "health score" that quantifies the severity of the anomaly. The central processor is configured to receive health scores from various network switches and take appropriate actions in response to the health scores.

[0022] In one non-limiting embodiment, each network switch includes an embedded CPU and a machine learning (ML) accelerator. The embedded CPU receives selected packet flows processed by the switch, performs certain preparations on the packet flows, and sends the packet flows to the ML accelerator. The ML accelerator runs a suitable ML inference model that identifies faults in the subsystem by analyzing the packet flows.

[0023] The ML inference model may be trained, for example, to detect patterns of statistical deviations of packets, based, for example, on one or more packets being corrupted in some manner, deviating from a predetermined policy rule, being dropped, not being dropped despite being corrupted in some manner, etc. In some models, the absolute number of packets is considered, while in other models, the rate and / or percentage of packets exhibiting different characteristics are considered. In some models, one or more characteristic signatures are generated, such as may be seen in a scatter plot defined by a plot of two or more packet characteristics, and the characteristic signatures are then analyzed to determine the presence of an anomaly.

[0024] Analyzing packet flows and identifying anomalies at the network switch is very effective for several reasons. First, the network switch is located at the "edge of the network", close to the sensors or other subsystems being monitored. In addition, the disclosed solution enables the distribution of analysis tasks among multiple switches, thereby reducing the burden on the central computer and simplifying the operation of each individual switch. The disclosed distributed solution integrates naturally with the architecture of modern automotive systems, where the vehicle is divided into zones served by different "zone ECUs" and different switches.

[0025] Figure 1 is a block diagram schematically illustrating an automotive communication system deployed in a vehicle 24 according to embodiments described herein.

[0026] The vehicle 24 includes a plurality of electronic subsystems 28 of various types. Some of the subsystems 28 include sensors such as cameras, speed sensors, accelerometers, audio sensors, infrared sensors, radar sensors, lidar sensors, ultrasonic sensors, range finders or other proximity sensors, and / or any other suitable sensor types. Other subsystems 28 include, for example, advanced driver assistance systems (ADAS) and / or in-vehicle infotainment (IVN) systems. Other subsystems 28 include electronic control units (ECUs) that control vehicle elements such as the engine, body, steering, and the like. Additionally or alternatively, the vehicle 24 may include any other suitable type of electronic subsystem 28.

[0027] In some embodiments, the vehicle 24 is divided into multiple zones, and the subsystems 28 of each zone are controlled by a corresponding “zone ECU” 40. Each zone ECU 40 communicates with a central computer 44 of the vehicle 24.

[0028] The electronic subsystems 28, ECU 40, and central computer 44 communicate with each other by sending and receiving communication packets over computer network 20. In this example, network 20 operates in accordance with one of the IEEE 802.3 Ethernet standards cited above (e.g., IEEE 802.3bw-2015). Network 20 includes a plurality of automotive network switches 32, which in this example are Ethernet switches. Communication between switches 32, between switches 32 and ECU 40, and between ECU 40 and subsystems 28 is performed via network links 36. Links 36 may include any suitable physical media, such as twisted pair copper wire links, optical links, waveguides, etc., in accordance with applicable Ethernet standards.

[0029] In one embodiment, Figure 1 The inset at the bottom of the is a diagram showing the internal structure of an example switch 32. In some embodiments, all switches 32 of the network 20 have a similar internal structure. In other embodiments, only a subset of switches 32 (perhaps only a single switch 32) will have such a structure.

[0030] In this example, the switch 32 includes a plurality of ports 48, a switch core 52, an embedded CPU 54, and a machine learning (ML) accelerator 56. The ports 48 are connected to respective links 36 that connect the switch 32 to the subsystem 28, to other switches 32, and / or to a higher-level computer 60 (e.g., a CPU or GPU), such as a regional ECU 40 or a central computer 44. Thus, the ports 48 are configured to transmit and receive packets to and from other system components over the network 20. The switch core 52 is configured to forward packets between ports so that each received packet is transmitted to its appropriate destination. The embedded CPU 54 is configured to configure and control the operation of the switch 32. The embedded CPU 54 may include, for example, an ARM processor or other suitable processor.

[0031] Figure 1 The configuration of the system and its components (such as Figure 1 The internal structure of the network switch 32 shown is an example configuration depicted for clarity only. In alternative embodiments, any other suitable configuration may be used. For example, the network 20 may include any suitable number of switches 32, possibly a single switch 32, and the network 20 may have any other suitable interconnection topology. The number and type of subsystems 28 and ECUs 40 in the vehicle 24 may also vary. For clarity, elements that are not mandatory for understanding the disclosed technology are omitted from the figure.

[0032] Figure 1 Different elements of the system and its various components can be implemented using dedicated hardware or firmware, such as using hardwired or programmable logic, for example, using an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA). In some embodiments, for example, the entire automotive network switch 32 is implemented using a system on a chip (SoC). Additionally or alternatively, some functions (e.g., functions of the embedded CPU 54 and / or the ML accelerator 56) can be implemented in software and / or using a combination of hardware and software elements.

[0033] In some embodiments, CPU 54 and / or ML accelerator 56 include a programmable processor that is programmed using software to perform the functions described herein. For example, the software may be downloaded to any processor in electronic form over a network, or may alternatively or additionally be disposed and / or stored on a non-transitory tangible medium such as a magnetic, optical or electronic memory.

[0034] In some embodiments, the embedded CPU 54 and the ML accelerator 56 are configured to analyze at least some of the packets processed by the switch 32. By analyzing the packets, the embedded CPU 54 and the ML accelerator 56 identify anomalies that indicate an existing or impending failure in a subsystem of the vehicle 24.

[0035] Note that while the embodiments described herein involve a certain "division of labor" between the embedded CPU 54 and the ML accelerator 56, such a division is by no means mandatory. In alternative embodiments, packet analysis and anomaly detection may be divided between the embedded CPU 54 and the ML accelerator 56 in any other suitable manner. Further, alternatively, packet analysis and anomaly detection may be performed by any other suitable configuration of one or more processors. For example, in some embodiments, such as when a relatively simple inference model is used, packet analysis and anomaly detection are performed only by the embedded CPU 54, and the ML accelerator is omitted.

[0036] In this context, the term "anomaly in a packet" refers to any pattern, attribute, content, and / or other one or more attributes of a packet or packet stream that deviates from a baseline. Some example anomalies include statistical deviation patterns in one or more attributes of packets originating from a subsystem 28, such as:

[0037] Statistical deviation packet drops, for example, the percentage of dropped packets exceeds a certain threshold percentage or the deviation from the expected baseline drop rate exceeds a certain threshold percentage.

[0038] Statistical deviation packet rate, for example, the packet rate is below a certain threshold rate or deviates from the expected baseline rate by more than a certain threshold percentage.

[0039] Statistical deviation packet latency, for example, latency exceeding a certain threshold latency or deviation from an expected baseline latency exceeding a certain threshold percentage.

[0040] Statistical deviations from the inter-group interval, for example, an interval exceeding a certain threshold interval or a deviation from the expected baseline inter-group interval exceeding a certain threshold percentage.

[0041] In alternative embodiments, embedded CPU 54 and ML accelerator 56 may identify any other suitable anomalies in packets processed by switch 32 that may indicate an existing or impending failure.

[0042] In an exemplary embodiment, the embedded CPU 54 and ML accelerator 56 monitor and establish expected baseline packet statistics (eg, packet rate per source subsystem, packet drop rate per source subsystem or per port, etc.) This baseline is then used to identify deviations.

[0043] In various embodiments, the embedded CPU 54 and ML accelerator 56 detect various anomalies that may indicate various existing or impending failures. For example, a sensor may become dirty, a sensor may be unplugged or a connector may become loose over time, a vehicle may pass through an unexpected area of ​​persistent electromagnetic interference (e.g., an extreme thunderstorm), or an engine problem may cause interference that affects grouping. Some failures are "hard failures" such as a complete failure of a sensor. Other failures, such as a dirty camera or a slightly loose connector, are "soft" failures, meaning that they exhibit some performance degradation that corresponds to a significant change in grouping, but are not catastrophic failures.

[0044] In some embodiments, a "soft" fault develops over time and becomes more severe, and this development is detected by analyzing packets by the embedded CPU 54 and / or ML accelerator 56. For example, a trend of increasing latency, decreasing throughput, and / or increasing packet drops may indicate a developing fault.

[0045] In some embodiments, the embedded CPU 54 counts various packet-related events using appropriate counters. Typically, although not necessarily, the counters accumulate events individually per port 48. The embedded controller 54 and / or ML accelerator 56 may use the counted events to detect anomalies.

[0046] Figure 2 is a diagram schematically illustrating the collection of packet-related events per port in a network switch 32 according to an embodiment described herein. In this example, the embedded controller 54 maintains appropriate registers that count the occurrence of the following events per port 48 of the switch 32:

[0047] InDiscards: A count of frames that were received as good but could not be forwarded due to lack of buffer memory.

[0048] InFiltered: A count of frames that were received as good but not forwarded due to filter policy rules.

[0049] InAccepted: A count of frames that were received well and were not filtered by the policy and were not dropped due to errors.

[0050] InBadAccepted: A count of frames received with CRC errors that were not discarded or filtered.

[0051] The counters listed above are described by way of example only. Additionally or alternatively, the embedded CPU 54 may accumulate any other suitable statistical data useful for detecting anomalies.

[0052] In some cases, an anomaly in a packet stream indicates a fault in the subsystem that generates the packet stream. For example, an abnormally low packet rate from a camera can indicate a fault in the camera. In other cases, an anomaly in a packet stream indicates a fault in another component, such as a loose or faulty cable or connector elsewhere in the network. Therefore, in this context, the term "fault in a subsystem of the vehicle 24" refers generally to any component that may cause an anomaly in the packets, such as a subsystem 28, an ECU, a cable, a connector, etc.

[0053] In some embodiments, the ML accelerator 56 runs a trained ML inference model in software and / or hardware that identifies anomalies in one or more packet streams. In other embodiments, the ML inference model can run on the embedded CPU 54 without the need for an additional accelerator. In other embodiments, the embedded CPU 54 or other processor can use any suitable criteria to analyze packets and identify anomalies without the need for an ML model.

[0054] Figure 3 is a flow chart schematically illustrating a method for using machine learning (ML) for anomaly detection in a network switch 32 of a communication system 20 according to an embodiment described herein. The method begins at a packet receiving operation 70, where the switch core 52 receives a packet from a sensor and / or other subsystem 28 of a vehicle 24 via a port 48.

[0055] At the selection and mirroring operation 74, the switch core 52 selects one or more relevant packet flows for analysis and mirrors the packets of the selected packet flows to the embedded CPU 54. In some embodiments, the switch core 52 uses a suitable address-based policy or uses other packet header fields to select packet flows for mirroring. For example, the switch core 52 can select packets with a predefined source address that corresponds to a corresponding sensor or other subsystem that has been pre-designated for monitoring. In other embodiments, the switch core 52 can perform deep packet inspection (DPI), for example, inspecting the packet payload or header located above the Ethernet layer, to select packet flows for mirroring. Further, alternatively, the switch core 52 can use any other suitable technology or criteria to select packets for transmission to the embedded CPU 54.

[0056] At check operation 78, the embedded controller 54 checks whether the required ML analysis is within its own processing capabilities or whether the ML analysis requires the use of the ML accelerator 56. As described above, in some cases, the required ML analysis uses a simple model that can be run on the embedded controller 56. In this case, at CPU analysis operation 82, the embedded CPU 54 applies the required ML model to the packet in order to identify an existing or impending fault.

[0057] In some embodiments, when a fault is identified, the model outputs a "health score" - a numerical value that gives a quantitative measure of the severity of the fault. In an example embodiment, a very low health score (e.g., zero) indicates a hard fault, a higher health score indicates a partial fault that degrades performance but is not catastrophic, and an even higher health score indicates an impending fault that is expected but has not yet occurred. Alternatively, any other suitable way to quantify the severity of a fault may be used.

[0058] At reporting operation 86, embedded CPU 54 reports the health score to a higher-level CPU / GPU 60, for example, to a regional ECU 40 or central computer 44. In one embodiment, embedded CPU 54 generates a dedicated Ethernet frame including the health score and sends the frame to the higher-level CPU / GPU 60 via switch core 52. Alternatively, embedded CPU 54 may report the health score in any other suitable manner.

[0059] On the other hand, if the result of operation 78 is that the ML analysis is complex and requires the use of the ML accelerator 56, then at a transmission operation 90, the embedded CPU 54 transmits the packet stream provided by the switch core 52 to the ML accelerator 56. At an accelerator analysis operation 94, the ML accelerator 56 applies an ML model (e.g., a deep learning model) to the packet in order to identify an existing or impending fault. As described above, in some embodiments, when a fault is identified, the model outputs a health score that gives a quantitative measure of the severity of the fault. The ML accelerator 56 sends the health score to the embedded CPU 54. At a reporting operation 86, the embedded CPU 54 reports the health score to a higher-level CPU / GPU 60, such as to a regional ECU 40 or a central computer 44.

[0060] In various embodiments, the higher-level CPU / GPU 60 may perform any suitable action in response to receiving the health score. For example, the higher-level CPU / GPU 60 may issue a real-time alert if the health score indicates a serious failure, or record the health score in memory for offline analysis. In some embodiments, the higher-level CPU / GPU 60 may make a decision or initiate a response action based on a combination of health scores received from the same switch 32 or from multiple different switches 32.

[0061] Although the embodiments described herein are primarily directed to automotive network communication systems, the methods and systems described herein may also be used in other applications, such as in industrial network communication systems that use Ethernet links to collect data from sensors and / or control various devices in industrial environments, and in smart home systems that collect data from and control home sensors and appliances.

[0062] Note that the above embodiments are cited by way of example, and the present invention is not limited to what has been specifically shown and described above. On the contrary, the scope of the present invention includes combinations and sub-combinations of the various features described above, as well as changes and modifications that would occur to a person skilled in the art after reading the foregoing description and that are not disclosed in the prior art. The documents incorporated by reference into this patent application should be considered as an integral part of this application, with the exception that if any term is defined in these incorporated documents in a manner that conflicts with the definition made explicitly or implicitly in this specification, only the definition in this specification should be considered.

Claims

1. A communication network switch, comprising: a plurality of ports configured to receive packets from electronic subsystems via a computer network and to transmit the packets to other electronic subsystems via the computer network; a switch core configured to receive the packet from one or more of the ports, to forward the packet to at least one of the ports, and to transmit the packet over a network link of the computer network; as well as One or more processors configured to: obtaining at least some of the packets processed by the switch core; analyzing the acquired packets to identify anomalies in one or more of the electronic subsystems; generating a health score based on the analysis of the acquired packets, the health score quantifying the severity of the anomaly in one or more of the electronic subsystems; as well as Notification of the health score is sent via the computer network to a central processor external to the communications network switch. 2 . The communication network switch of claim 1 , wherein when analyzing the acquired packets, the one or more processors are configured to identify existing anomalies or predict upcoming anomalies.

3. The communication network switch of claim 1, wherein the computer network comprises one of the following: Car Networks; Industrial networks; and Smart home network.

4. A communication network switch according to claim 1 or 2, wherein at least one of the electronic subsystems includes a sensor, wherein the port is configured to receive at least some of the packets from the sensor, and wherein the one or more processors are configured to identify abnormal function of the sensor.

5. The communication network switch of claim 1 or 2, wherein the one or more processors are configured to identify the anomaly in response to detecting a pattern of statistical deviations in an attribute of the packet.

6. The communication network switch of claim 5, wherein the statistical deviation pattern comprises one or more of the following: Statistical deviation packets are discarded; Statistical deviation packet rate; Statistical deviation delay; and Statistical deviation between group intervals.

7. A communication network switch according to claim 1 or 2, wherein the one or more processors are configured to run a machine learning model, and the machine learning model is configured to detect the anomaly.

8. The communication network switch according to claim 1 or 2, wherein the one or more processors are configured to count events related to the packets and detect the anomaly based on the counted events.

9. The communication network switch according to claim 1 or 2, wherein at least one of the one or more processors includes a machine learning ML accelerator, and the machine learning ML accelerator is separated from a switch central processing unit CPU or a switch graphics processing unit GPU provided in the switch.

10. A communication network switch according to claim 1 or 2, wherein the one or more processors are configured to: select one or more packet flows from the packets processed by the switch according to predefined selection criteria, and identify the anomaly by analyzing the packets in the selected packet flows.

11. A method for anomaly detection, the method comprising: In a network switch, packets are received from electronic subsystems via a computer network and the packets are transmitted to other electronic subsystems via the computer network; analyzing, by the network switch, at least some of the packets processed by the network switch to identify anomalies in one or more of the electronic subsystems; generating a health score based on the analysis of the acquired packets, the health score quantifying the severity of the anomaly in one or more of the electronic subsystems; as well as Notification of the health score is sent from the switch through the computer network to a central processor external to the network switch.

12. The method of claim 11, wherein analyzing the grouping comprises identifying an existing anomaly or predicting an upcoming anomaly.

13. The method of claim 11, wherein the computer network comprises one of the following: Car Networks; Industrial networks; and Smart home network.

14. The method of claim 11 or 12, wherein analyzing the packets comprises: The anomaly is identified in response to detecting a pattern of statistical deviations in the attribute of the grouping.

15. The method according to claim 11 or 12, wherein analyzing the packet comprises counting events related to the packet, and detecting the anomaly based on the counted events.