# US 2019/0171510 A1 · Monitoring and analyzing watchdog messages in an internet of things network environment

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2019/0171510 A1
- Document ID: US20190171510A1
- Status: Published application
- Kind: A1
- Application number: 16/260,923
- Filed: 2019-01-29
- Published: 2019-06-06
- Earliest claimed: 2016-09-28
- Assignees: McAfee, Inc.
- Applicants: McAfee, LLC
- Inventors: Ned M. Smith; Thiago Jose Macieira; Zheng Zhang; Tobias M. Kohlenberg; Igor G. Muttik
- Web page: https://patentgrep.com/patent/US20190171510A1

## CPC

- G06F 11/0757: {by exceeding a time limit, i.e. time-out, e.g. watchdogs}
- G06F 11/0751: {Error or fault detection not based on redundancy (power supply failures G06F1/30; network fault management H04L41/06)}
- G06F 11/079: {Root cause analysis, i.e. error or fault diagnosis (in a hardware test environment G06F11/22; in a software test environment G06F11/36)}
- H04L 43/04: Processing captured monitoring data, e.g. for logfile generation
- H04L 43/0805: by checking availability
- H04L 43/10: Active monitoring, e.g. heartbeat, ping or trace-route
- H04L 43/16: Threshold monitoring
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

There is disclosed in one example a network device, including: a hardware platform including at least a processor and a memory; a communication interface; and stored instructions on the memory to instruct the processor to provide a health monitoring engine (HME) configured to: communicatively couple to a network via the network interface; construct a reference template during a training period; observe watchdog behavior on the network during an observation period; identify an abnormality in the watchdog behavior including a substantial variance from the reference template; and trigger a resilience response to the substantial variance.

## Claims

### Claim 1

A network device, comprising:
a hardware platform comprising at least a processor and a memory;
a communication interface; and
stored instructions on the memory to instruct the processor to provide a health monitoring engine (HME) configured to:
communicatively couple to a network via the network interface;
construct a reference template during a training period;
observe watchdog behavior on the network during an observation period;
identify an abnormality in the watchdog behavior comprising a substantial variance from the reference template; and
trigger a resilience response to the substantial variance.

### Claim 2 (depends on 1)

The network device of claim 1, wherein observing watchdog behavior comprises processing heartbeat messages.

### Claim 3 (depends on 2)

The network device of claim 2, wherein the heartbeat messages comprise information related to a network node or subnet that produced the heartbeat message, the information comprising at least one of a device identifier (ID), a local time stamp, a next heartbeat message time, node health information, and node or subnet diagnostic information.

### Claim 4 (depends on 1)

The network device of claim 1, wherein observing watchdog behavior on the network comprises operating a machine learning analysis engine to construct the reference template based at least in part on heartbeat messages that nodes or subnets in the network are configured to report.

### Claim 5 (depends on 4)

The network device of claim 4, wherein the machine learning analysis engine is further to measure differences between actual characteristics of heartbeat messages from a node or subnet and expected heartbeat patterns for the node or subnet.

### Claim 6 (depends on 5)

The network device of claim 5, wherein the characteristics include at least one of heartbeat message arrival time, latency, anomalies due to work patterns, time of day, and site specific patterns.

### Claim 7 (depends on 1)

The network device of claim 1, wherein the HME comprises a machine learning analysis engine to identify a threshold of substantial variance from the reference template.

### Claim 8 (depends on 7)

The network device of claim 7, wherein identifying an abnormality further comprises identifying a threshold exceeding the variance.

### Claim 9 (depends on 1)

The network device of claim 1, wherein the resilience response includes at least one of notification, repair, and deployment of network redundancy.

### Claim 10 (depends on 1)

The network device of claim 1, wherein constructing the reference template during the training period comprises applying a data model description of expected behavior of a message traffic system.

### Claim 11 (depends on 1)

The network device of claim 1, further comprising logic for the device to act as a network gateway to an internet of things network.

### Claim 12

One or more tangible, non-transitory computer-readable mediums having stored thereon executable instructions to provide a health monitoring engine (HME) configured to:
communicatively couple to an internet of things (IoT) network;
train a machine learning algorithm, comprising constructing a reference template;
monitor watchdog behavior on the IoT network;
identify watchdog behavior on the IoT network varying, beyond a threshold, from the reference template; and
activate a resilience response to protect IoT network health.

### Claim 13 (depends on 12)

The one or more tangible, non-transitory computer-readable mediums of claim 12, wherein monitoring watchdog behavior comprises processing heartbeat messages.

### Claim 14 (depends on 13)

The one or more tangible, non-transitory computer-readable mediums of claim 13, wherein the heartbeat messages comprise information related to an IoT device or subnet that produced the heartbeat message, the information comprising at least one of a device identifier (ID), a local time stamp, a next heartbeat message time, node health information, and node or subnet diagnostic information.

### Claim 15 (depends on 12)

The one or more tangible, non-transitory computer-readable mediums of claim 12, wherein monitoring watchdog behavior on the IoT network comprises operating a machine learning analysis engine to construct the reference template based at least in part on heartbeat messages that IoT devices or subnets in the network are configured to report.

### Claim 16 (depends on 15)

The one or more tangible, non-transitory computer-readable mediums of claim 15, wherein the machine learning analysis engine is further to measure differences between actual characteristics of heartbeat messages from an IoT device or subnet and expected heartbeat patterns for the IoT device or subnet.

### Claim 17 (depends on 16)

The one or more tangible, non-transitory computer-readable mediums of claim 16, wherein the characteristics include at least one of heartbeat message arrival time, latency, anomalies due to work patterns, time of day, and site-specific patterns.

### Claim 18

A computer-implemented method of providing a health monitoring engine (HME), comprising:
communicatively coupling to an internet of things (IoT) network;
training a machine learning algorithm, comprising constructing a reference template;
monitoring behavior on the IoT network;
identifying behavior on the IoT network that does not conform to the reference template; and
activating a resilience response.

### Claim 19 (depends on 18)

The method of claim 18, wherein monitoring behavior comprises processing heartbeat messages.

### Claim 20 (depends on 19)

The method of claim 19, wherein the heartbeat messages comprise information related to an IoT device or subnet that produced the heartbeat message, the information comprising at least one of a device identifier (ID), a local time stamp, a next heartbeat message time, node health information, and node or subnet diagnostic information.
