US-12621220-B2 - Virtual network assistant with location input

US12621220B2US 12621220 B2US12621220 B2US 12621220B2US-12621220-B2

Abstract

Techniques are described in which a network management system (NMS) is configured to determine a root cause of degraded network performance based on SLE metrics and the locations associated with network devices providing the SLE metrics. The NMS can determine service level experience (SLE) metrics associated with each client device on a network and location data for each client device of the plurality of client devices. The NMS can generate a time series of parameter vectors, where each parameter vector includes SLE metrics corresponding to each client device of the plurality of client devices. Each parameter vector is associated with the location of the client device corresponding to the SLE metrics. The NMS can determine, based on the time series of parameter vectors and associated locations, a root cause for a degradation in SLE metrics associated with the one or more of the client devices.

Inventors

Mohammad Zohoorian
Ebrahim Safavi
Shmuel Shaffer

Assignees

JUNIPER NETWORKS, INC.

Dates

Publication Date: 20260505
Application Date: 20210916

Claims (20)

1 . A system comprising: a plurality of access point (AP) devices configured to provide a wireless network at a site; a location engine configured to determine location data for a plurality of client devices, wherein the location data indicates a location associated with each client device of the plurality of client devices; and a network management system comprising: memory, and one or more processors coupled to the memory and configured to: receive, from the plurality of AP devices, network data collected by the plurality of AP devices or a plurality of client devices that is associated with the wireless network, generate a time series of parameter vectors, each parameter vector of the time series of parameter vectors comprising service level expectation (SLE) metrics determined from the network data of a corresponding client device of the plurality of client devices, wherein each parameter vector is associated with the location data indicating a corresponding location of the corresponding client device, determine, based on the time series of parameter vectors, a set of impacted client devices from the plurality of client devices, wherein the set of impacted client devices are experiencing a degradation in the SLE metrics, and wherein at least one impacted client device of the set of impacted client devices is associated with an AP device of the plurality of AP devices, determine, based on the location data indicating locations of the set of impacted client devices that are experiencing the degradation in the SLE metrics, whether any impacted client device in the set of impacted client devices is in close proximity relative to another impacted client device in the set of impacted client devices, and based on determining that there are impacted client devices in the set of impacted client devices within close proximity of each other, determine that the AP device associated with the at least one impacted client device is not a root cause for the degradation in the SLE metrics associated with the set of impacted client devices.
2 . The system of claim 1 , wherein to determine that the AP device is not the root cause for the degradation in the SLE metrics associated with the set of impacted client devices, the one or more processors are configured to: determine that the set of impacted client devices is correlated with the degradation in the SLE metrics; and determine that the root cause for the degradation in the SLE metrics associated with the set of impacted client devices is due to electrical noise in a vicinity of the set of impacted client devices.
3 . The system of claim 1 , wherein the one or more processors are further configured to determine an initial root cause, and wherein to determine that the AP device is not the root cause for the degradation in the SLE metrics, the one or more processors are configured to determine that there are impacted client devices in the set of impacted client devices within close proximity of each other based on clustering impacted client devices with degradation in SLE metrics that contributed to the determination of the initial root cause.
4 . The system of claim 1 , wherein to determine that the AP device is not the root cause for the degradation in the SLE metrics associated with the set of impacted client devices, the one or more processors are configured to: determine that there are impacted client devices in the set of impacted client devices within close proximity of each other based on a characteristic of each client device of the plurality of client devices.
5 . The system of claim 4 , wherein the characteristic comprises a wireless frequency band.
6 . The system of claim 1 , wherein the AP device is a first AP device of the plurality of AP devices, and wherein the degradation in the SLE metrics continues for a time period longer than a threshold amount of time, and wherein to determine that the first AP device is not the root cause for the degradation in the SLE metrics associated with the set of impacted client devices, the one or more processors are configured to determine the root cause is a gap in wireless coverage, and wherein the one or more processors are configured to provide a suggested location to place a second AP device.
7 . The system of claim 1 , wherein to determine that the AP device is not the root cause for the degradation in the SLE metrics associated with the set of impacted client devices, the one or more processors are configured to determine the root cause based on an output of a machine learning model.
8 . The system of claim 1 , wherein to determine that the AP device is not the root cause for the degradation in the SLE metrics associated with the set of impacted client devices, the one or more processors are configured to determine the root cause is transient noise.
9 . The system of claim 1 , wherein the parameter vector corresponding to the corresponding client device includes the corresponding location of the corresponding client device.
10 . The system of claim 1 , wherein the parameter vector corresponding to the corresponding client device includes an identifier of the AP device of the plurality of AP devices with which the corresponding client device is associated.
11 . A network management system (NMS) that manages a plurality of access point (AP) devices in a wireless network, the NMS comprising: memory; and one or more processors coupled to the memory and configured to: receive, from the plurality of AP devices, network data collected by the plurality of AP devices or a plurality of client devices that is associated with the wireless network, generate a time series of parameter vectors, each parameter vector of the time series of parameter vectors comprising service level expectation (SLE) metrics determined from the network data of a corresponding client device of the plurality of client devices, wherein each parameter vector is associated with location data indicating a corresponding location of the corresponding client device, determine, based on the time series of parameter vectors, a set of impacted client devices from the plurality of client devices, wherein the set of impacted client devices are experiencing a degradation in the SLE metrics, and wherein at least one impacted client device of the set of impacted client devices is associated with an AP device of the plurality of AP devices, determine, based on the location data indicating locations of the set of impacted client devices that are experiencing the degradation in the SLE metrics, whether any impacted client device in the set of impacted client devices is in close proximity relative to another impacted client device in the set of impacted client devices, and based on determining that there are impacted client devices in the set of impacted client devices within close proximity of each other, determine that the AP device associated with the at least one impacted client device is not a root cause for the degradation in the SLE metrics associated with the set of impacted client devices.
12 . The NMS of claim 11 , wherein to determine that the AP device is not the root cause for the degradation in the SLE metrics associated with the set of impacted client devices, the one or more processors are configured to: determine that the set of impacted client devices is correlated with the degradation in the SLE metrics; and determine that the root cause for the degradation in the SLE metrics associated with the set of impacted client devices is due to electrical noise in a vicinity of the set of impacted client devices.
13 . The NMS of claim 11 , wherein the one or more processors are further configured to determine an initial root cause, and wherein to determine that the AP device is not the root cause for the degradation in the SLE metrics, the one or more processors are configured to determine that there are impacted client devices in the set of impacted client devices within close proximity of each other based on clustering impacted client devices with degradation in SLE metrics that contributed to the determination of the initial root cause.
14 . The NMS of claim 11 , wherein to determine that the AP device is not the root cause for the degradation in the SLE metrics associated with the set of impacted client devices, the one or more processors are configured to determine that there are impacted client devices in the set of impacted client devices within close proximity of each other based on a wireless frequency band of each client device of the plurality of client devices.
15 . The NMS of claim 11 , wherein the AP device is a first AP device of the plurality of AP devices, and wherein the degradation in the SLE metrics continues for a time period longer than a threshold amount of time, and wherein to determine that the first AP device is not the root cause for the degradation in the SLE metrics associated with the set of impacted client devices, the one or more processors are configured to determine the root cause is a gap in wireless coverage, and wherein the one or more processors are configured to provide a suggested location to place a second AP device.
16 . The NMS of claim 11 , wherein to determine that the AP device is not the root cause for the degradation in the SLE metrics associated with the set of impacted client devices, the one or more processors are configured to determine the root cause based on an output of a machine learning model.
17 . The NMS of claim 11 , wherein to determine that the AP device is not the root cause for the degradation in the SLE metrics associated with the set of impacted client devices, the one or more processors are configured to determine the root cause is transient noise.
18 . The NMS of claim 11 , wherein the parameter vector corresponding to the corresponding client device includes the corresponding location of the corresponding client device and an identifier of the AP device of the plurality of AP devices with which the corresponding client device is associated.
19 . A method comprising: receiving, from a plurality of AP devices, network data collected by the plurality of AP devices of a wireless network or a plurality of client devices associated with the wireless network; determining, based on the network data, service level expectation (SLE) metrics of a corresponding client device of the plurality of client devices; determining location data for each client device of the plurality of client devices, wherein the location data indicates a location associated with each client device of the plurality of client devices; generating a time series of parameter vectors, each parameter vector of the time series of parameter vectors comprising SLE metrics determined from the network data of the corresponding client device of the plurality of client devices, wherein each parameter vector is associated with the location data indicating a corresponding location of the corresponding client device; determining, based on the time series of parameter vectors, a set of impacted client devices from the plurality of client devices, wherein the set of impacted client devices are experiencing a degradation in the SLE metrics, and wherein at least one impacted client device of the set of impacted client devices is associated with an AP device of the plurality of AP devices; determining, based on the location data indicating locations of the set of impacted client devices that are experiencing the degradation in the SLE metrics, whether any impacted client device in the set of impacted client devices is in close proximity relative to another impacted client device in the set of impacted client devices; and based on determining that there are impacted client devices in the set of impacted client devices within close proximity of each other, determining that the AP device associated with the at least one impacted client device is not a root cause for the degradation in the SLE metrics associated with the set of impacted client devices.
20 . The method of claim 19 , wherein determining that the AP device is not the root cause for the degradation in the SLE metrics associated with the set of impacted client devices comprises: determining that the set of impacted client devices corresponds to the degradation in the SLE metrics; and determining that the root cause for the degradation in the SLE metrics associated with the set of impacted client devices is due to electrical noise in a vicinity of the set of impacted client devices.

Description

RELATED APPLICATION This application claims the benefit of U.S. Provisional Application No. 63/202,942, filed Jun. 30, 2021, the entire contents of which are incorporated herein by reference. FIELD The disclosure relates generally to computer networks and, more specifically, machine learning-based diagnostics of computer networks and network systems using device location data. BACKGROUND Wireless access networks make use of network of wireless access points (APs), which are physical, electronic devices that enable other devices to wirelessly connect to a wired network using various wireless networking protocols and technologies, such as wireless local area networking protocols conforming to one or more of the IEEE 802.11 standards (i.e., “WiFi”), Bluetooth/Bluetooth Low Energy (BLE), mesh networking protocols such as ZigBee or other wireless networking technologies. Many different types of wireless client devices, such as laptop computers, smartphones, tablets, wearable devices, appliances, and Internet of Things (IoT) devices, incorporate wireless communication technology and can be configured to connect to wireless access points when the device is in range of a compatible wireless access point in order to access a wired network. Wireless access networks, and computer networks in general, are complex systems which may experience transient and/or permanent issues. Some of the issues may be caused by sources external to the wireless network, such as noise introduced by electrical devices. SUMMARY In general, this disclosure describes techniques determining root causes of degradations in performance of networks based on location data of client devices in the network. A network management system (NMS) receives network data associated with a plurality of client devices in a wireless network at a site. The network data is indicative of one or more aspects of wireless network performance. The NMS can determine that the network data indicates degradation in performance metrics, and, based on the network data, can determine a root cause for the degradation. In some aspects, the NMS can use location data to facilitate determining the root cause. For example, the network data may indicate that some, but not all of the client devices associated with one or more APs have poor SLE metrics. In response to the indication of poor SLE metrics, the NMS can cluster the client devices experiencing poor SLE metrics based on location. For example, the NMS can cluster client devices based on the location of the client devices. If all of the client devices in the cluster have poor SLE metrics, the NMS can determine that there is not a fault in the AP, and can determine that some other factor is the cause of the poor SLE metrics of client devices the cluster. For example, transient noise (e.g., temporary electrical interference with a network signal) may be the cause of the poor SLE metrics for client devices in those clusters where the client devices exhibits poor SLE metrics. The techniques of the disclosure provide one or more technical advantages and practical applications. For example, the techniques enable the NMS to automatically and accurately determine root causes for network degradation that may be due to sources external to the network and/or network devices, such as transient noise. The ability of an NMS to identify such external causes as root causes provided by the techniques disclosed herein can avoid performance of remedial actions that may be unnecessary and/or fail to address the actual cause of poor SLE metrics of some of the client devices. This can be advantageous because it can avoid unnecessary resource costs and customer inconvenience associated with performing the unnecessary remedial actions. As an example, transient noise may be the actual root cause for degradation in SLE metrics. In the absence of location data of client devices, the root cause may appear to be a fault in an AP, potentially resulting in a reset of the AP. In the case where transient noise is the actual root cause, resetting the AP does not address the actual root cause, and there can be wasted resources and unnecessary downtime involved in resetting the AP. Additionally, the techniques facilitate detection of root causes not previously detectable in an automated manner. Further, the more accurate detection of root causes facilitated by the techniques disclosed herein can result in more rapid resolution of issues in a network, leading to greater user and network operator satisfaction. In one example, a system includes a plurality of AP devices configured to provide a wireless network at a site; a location engine configured to determine location data for a plurality of client devices, wherein the location data comprises a location associated with each client device of the plurality of client devices; and a network management system comprising: a memory storing network data received from the plurality of AP devices, the network data collec