WO2021242253A1

WO2021242253A1 - Wi-fi and radio access network selection based on link quality

Info

Publication number: WO2021242253A1
Application number: PCT/US2020/035169
Authority: WO
Inventors: Ning Zhang; Mingguang Xu; kai SHI
Original assignee: Google LLC
Current assignee: Google LLC
Priority date: 2020-05-29
Filing date: 2020-05-29
Publication date: 2021-12-02
Anticipated expiration: 2022-11-29

Abstract

A user equipment (UE) measures one or more link quality characteristics, such as throughput, for one or more of the plurality of available wireless network connections. Based on the measurements, the UE can select the network connection associated with an expected higher link quality and can communicate with a packet data network using the selected network connection. Further, the UE can update the link quality characteristic measurements over time and make corresponding changes in the selected network connection.

Description

WI-FI AND RADIO ACCESS NETWORK SELECTION BASED ON LINK QUALITY BACKGROUND

Wireless communication devices such as cellular phones, tablets, and computers frequently are configured to wirelessly connect to a packet data network (PDN), such as the Internet. To enhance the reliability, flexibility, and geographical scope of PDN connections, a wireless communication device can support multiple types of wireless connections. For example, some wireless devices include both a Wi-Fi network interface that supports connections complying with the Institute of Electrical and Electronics Engineers (IEEE) 802.11 standard or its successors, and a Radio Access Network (RAN) interface supporting connections with one or more networks associated with a specified Radio Access Technology, such as a Third Generation Partnership Project (3GPP) Long Term Evolution (LTE) RAT implementation, or a 3GPP Fifth Generation (5G) New Radio (NR) RAT implementation.

BRIEF DESCRIPTION OF THE DRAWINGS The present disclosure may be better understood, and its numerous features and advantages made apparent to those skilled in the art by referencing the accompanying drawings. The use of the same reference symbols in different drawings indicates similar or identical items.

FIG. 1 is a block diagram of a communication network including a user equipment (UE) that selects between different network interfaces based on an identified link quality for each network interface in accordance with some embodiments.

FIG. 2 is a block diagram illustrating an example of the UE of FIG. 1 selecting between a Wi-Fi network and a RAN connection based on a comparison of throughput characteristics for each network in accordance with some embodiments. FIG. 3 is a block diagram illustrating a hardware configuration of the UE of FIGs. 1 and 2 in accordance with some embodiments.

FIG. 4 is a block diagram illustrating an example of network selection criteria of FIG.

3 in accordance with some embodiments. FIG. 5 is a flow diagram of a method of a UE selecting between available network connections based on link quality in accordance with some embodiments.

FIG. 6 is a flow diagram of a method of a UE selecting a RAN connection for data communication based on the predicted link quality of an available Wi-Fi connection in accordance with some embodiments.

DETAILED DESCRIPTION

While a wireless communication device can support different types of network connections, the relative quality of each available network connection can vary based on a number of factors, such as the location of the wireless communication device relative to network access points or base stations, the strength of wireless communication signals, the presence of potentially interfering objects or other wireless communication devices, and the like. Conventionally, in order to communicate (send and receive) data with a PDN or other network, a wireless communication device selects between available network connections based on a fixed set of criteria. For example, some wireless communication devices always select an available Wi-Fi connection over an available RAN connection to communicate data, as the Wi-Fi connection is assumed to provide a better user experience, for instance because it is assumed to have better link quality than the RAN connection. However, in some circumstances, the quality of the RAN connection can exceed that of the Wi-Fi connection and therefore provide a better user experience. Further, because the quality of each connection can vary based on the above-referenced factors, a fixed selection order for available network connections can lead to a poor user experience in different circumstances.

Disclosed herein are techniques for facilitating the selection of a wireless network connection at a UE (i.e. , a wireless communication device) based on an identified link quality for each of a plurality of available wireless network connections. In some embodiments, the UE measures one or more link quality characteristics, such as throughput, for one or more of the plurality of available wireless network connections. Based on the measurements, the UE can select the network connection associated with an expected higher link quality and can communicate with the PDN using the selected network connection. Further, the UE can update the link quality characteristic measurements over time and make corresponding changes in the selected network connection. The UE thereby enhances the likelihood that the higher-quality network connection is selected for the particular circumstances (e.g., the location) of the UE at a given point in time, thereby enhancing the overall user experience with the UE.

To illustrate via an example, in some embodiments the UE can support both a Wi-Fi connection with the PDN (that is, a wireless network connection via a Wi-Fi access point) and a RAN connection with the PDN (that is, a wireless network connection via a Radio Access Network). The UE can identify the throughput for each of the Wi-Fi connection and the RAN connection and can select the network connection having the higher identified throughput to communicate with the PDN. In response to one or more specified criteria, such as a change in location of the UE, expiration of a specified amount of time, a change in application executing at the UE, and the like, or any combination thereof, the UE can update the identified throughput for each network connection, and again select the network connection having the higher identified throughput. The UE can thereby change the selected network based on the changing circumstances of the UE, such that it is more likely that the network associated with greater throughput is selected at any given point in time, and thus improving the user experience.

In some embodiments, to reduce overhead the UE identifies the link quality, such as throughput, for only a subset of the available network connections, and selects an available connection based on a comparison of the link quality to a corresponding threshold. For example, the UE can identify only the throughput associated with the Wi-Fi network connection. In response to the identified throughput being below a specified threshold, the UE can select the RAN connection for data communication.

In some embodiments, the UE identifies the throughput for the Wi-Fi connection based on link-layer characteristics, such as one or more of a link speed (e.g., one or both of a transmit link speed and a receive link speed, a channel utilization, and a packet error rate) thereby enhancing the accuracy of the identified throughput.

In some cases, the throughput for a network connection at a given point in time can be a lagging indicator of the link quality associated with the network connection. Accordingly, in some embodiments the identified throughput for one or more of the available network connections is based on a prediction of the throughput for the corresponding connection at some future point in time. In some embodiments, the UE employs a machine learning (ML) model to generate the predicted throughput for a network connection, wherein the ML model is trained based on a specified set of link quality characteristics and employs one or more ML prediction techniques such as one or more of a random forest technique, a linear regression technique, a neural network, and the like.

In some embodiments, the UE can select from the available wireless networks based on both the identified link quality and additional criteria, such as whether the available RAN connection is a metered connection, the throughput requirements for an executing application, the location of the UE, and the like, or any combination thereof. To illustrate via an example, the UE can identify 1) that an available RAN connection has a higher throughput than an available Wi-Fi connection; 2) that the RAN connection is a metered connection; and 3) that the throughput requirements for an executing app fall below a threshold, such that the lower throughput of the Wi-Fi connection is expected to provide a satisfactory user experience. Accordingly, the UE selects the Wi-Fi connection for communicating with the PDN, thereby avoiding potential network charges to the user. In response to the user executing a new application having throughput requirements that exceed the threshold, the UE can switch to the RAN to ensure that the new application’s throughput requirements are met.

FIGs. 1 illustrates a communication network 100 implementing selectable network connections based on identified link quality in accordance with some embodiments. As shown in FIG. 1 , the communication network 100 includes a user equipment (UE) 102 that is capable of connecting to a packet data network (PDN) 104 via multiple wireless network connections. The UE 102 can include any of a variety of electronic wireless communication devices, such as a cellular phone, a cellular-enabled tablet computer or cellular-enabled notebook computer, an automobile or other vehicle employing cellular services (e.g., for navigation, provision of entertainment services, in-vehicle mobile hotspots, etc.), and the like. The PDN 104 is the Internet, one or more private interconnecting data networks, or a combination thereof. The UE 102 is configured to establish network connections with the PDN 104 via at least two different connections: a Wi-Fi connection via a wireless access point (WAP) 112 and radio access network (RAN) connection via a RAN 115. The WAP 112 is a device that supports wireless connection to the wired PDN 104 according to a specified Wi-Fi protocol. For example, in some embodiments the WAP 112 is embedded within a Wi-Fi router and complies with an IEEE 802.11 protocol or its successors, such as one or more of the IEEE 802.11a, 802.11b, 802.11 g, 802.11h, and 802.11ac protocols.

The RAN 115 is a radio access network that supports wireless connection with a backhaul infrastructure including a core network 117, with the core network 117 providing a further connection to the PDN 104. To provide connections to the core network 117, the RAN 115 includes base stations, such as base station 113, operable to wirelessly communicate with UEs within signal range based on one or more radio access technologies (RATs). Examples of the base station 113 include, for example, a NodeB (or base transceiver station (BTS)) for a Universal Mobile Telecommunications System (UMTS) RAT implementation (also known as “3G”), an enhanced NodeB (eNodeB) for a Third Generation Partnership Project (3GPP) Long Term Evolution (LTE) RAT implementation, a 5G node B (“gNB”) for a 3GPP Fifth Generation (5G) New Radio (NR) RAT implementation, and the like. As is well known in the art, the base station 113 operates as an “air interface” that establishes radio frequency (RF) wireless connections with UEs, and these wireless connections (or “links”) then serve as data and voice paths between the UEs and the core networks 117 for providing various services to the UEs, including voice services via circuit-switched networks or packet-switched networks, messaging services such as simple messaging service (SMS) or multimedia messaging service (MMS), multimedia content delivery, presence services, and the like.

As noted above, the UE 102 can connect to the PDN 104 via the WAP 112, referred to as a Wi-Fi connection, via the RAN 115, referred to as a RAN connection, or concurrently via both a Wi-Fi connection and a RAN connection. The UE 102 is capable of communicating (sending and receiving) data with the PDN 104 via either connection, and at different times can select a different one of the Wi-Fi connection and the RAN connection for data communication. For example, in some embodiments the UE 102 includes separate wireless modems to establish the Wi-Fi connections and the RAN connection and can select between the connections by placing the modem for the non-selected connection in a low-power or other inactive mode.

In some embodiments, the UE 102 monitors link quality characteristics associated with each of the Wi-Fi connection and the RAN connection, and selects the connection having the better link quality for data transfer. Examples of the link quality characteristics that can be monitored by the UE 102 include connection throughput, connection latency, and the like, or any combination thereof. In some embodiments, the UE 102 identifies the link quality characteristics based on link-layer information provided by the wireless modem associated with each connection, such as one of link speed, channel utilization, packet error rate, and the like, or any combination thereof. In some embodiments, the UE 102 can monitor the link quality characteristics for both the transmission of data and the reception of data, based on corresponding link-layer information for both an uplink and a downlink portion of the corresponding connection.

To illustrate, in the example of FIG. 1 the UE 102 monitors a throughput 110 for the Wi-Fi connection via the WAP 112 and a throughput 111 for the RAN connection via the base station 113. In some embodiments, the UE 102 selects the connection having the higher throughput to communicate data. As the relative value of the throughput 110 and throughput 111 change over time, the UE 102 can update the connection selected for data communication. For example, in some embodiments the UE 102 updates the selected network connection in response to a detected change in network conditions, such as the UE 102 changing cells of the RAN 115, or in response to detecting a change in signal strength of a wireless signal associated with either the RAN 115 or the WAP 112.

In some embodiments, monitoring the link quality for the RAN connection can require an undesirable amount of processing overhead, or can require monitoring devices not present at the UE 102. Accordingly, the UE 102 can select between the Wi-Fi connection and the RAN connection based only on the link quality of the Wi-Fi connection. For example, the UE 102 can compare the throughput 110 to a threshold and, in response to the throughput 110 being above the threshold indicating sufficient throughput for the Wi-Fi connection for a satisfactory user experience, the UE 102 selects the Wi-Fi connection for data transfer. In response to the throughput 110 being below the threshold, indicating insufficient throughput for the Wi-Fi connection for a satisfactory user experience, the UE 102 selects the RAN connection.

In some cases, the monitored throughput or other link quality characteristic at a given point in time is a lagging indicator of the link quality for the Wi-Fi and RAN connections. Accordingly, in some embodiments the UE 102 can predict the link quality for each of the Wi-Fi and RAN connections based on the link quality characteristics. For example, in some embodiments the UE 102 predicts the throughput for a given network connection based on detected trends or patterns in the throughput for the network connection, such as detecting that the throughput has changed linearly over a specified amount of time. In other embodiments, the UE 102 can predict the link quality for a network connection based on an ML model, such as a neural network that has been trained to identify patterns in link quality characteristics such as throughput. In some embodiments, the ML model is trained based on a specified training set of patterns. In other embodiments, the ML model can be trained, or adjusted, based on link quality information collected by the UE 102 as it is used.

In some embodiments, the RAN 115 is a 5G network that is separated into slices, or virtualized logical networks, with different slices of the RAN 115 supporting different service level requirements. In addition, the RAN 115 can assign different amounts of network resources to different network slices, such that the link quality of a given slice of the RAN 115 can differ from the link quality of another slice. The UE 102 can determine the link quality, such as the throughput or latency, for an individual network slice assigned to the UE 102, and determine whether to employ the RAN 115 for data communication based on the link quality for an individual network slice, rather than based on the link quality of all slices of the network.

FIG. 2 illustrates an example of the UE 102 selecting between the Wi-Fi connection and the RAN connection based on relative throughput in accordance with some embodiments. In the depicted example, the UE 102 has identified a higher throughput for the RAN connection (represented as a throughput value of 100 for the throughput 111) than for the Wi-Fi connection (represented as a throughput value of 50 for the throughput 110). In response, the UE 102 selects the RAN connection for data communication, and therefore communicates data 220 with the PDN 104 via the base station 113.

In some embodiments, the UE 102 selects between the RAN connection and the Wi Fi connection based on criteria in addition to the link quality of each connection, such as based on one or more of whether one of the connections is a metered connection, application quality (e.g., throughput) requirements, user-adjusted settings at the UE 102, and the like. For example, in some cases the RAN connection is a metered connection, wherein the user or owner of the UE 102 is charged based on the amount of data transferred via the RAN 115. Accordingly, the UE 102 can select the RAN connection only when the difference between the throughput 111 and the throughput 111 exceeds a specified threshold. The UE 102 thereby avoids selecting the metered RAN connection until the difference in connection throughput is relatively large, therefore selecting the metered RAN connection is likely to have a particularly large impact on the user experience.

As another example, in some cases a particular application (e.g., a video streaming application) can have relatively large throughput requirements to provide a satisfactory user experience, whereas another application (e.g., a simple video game) can have relatively small throughput requirements. Accordingly, when the video streaming application is executed, the UE 102 can select the metered RAN connection when the RAN connection has larger bandwidth and can select the un metered Wi-Fi connection when the video game application is executed.

FIG. 3 illustrates an example hardware configuration for the UE 102 in support of selecting among available network connections based on link quality in accordance with some embodiments. In the depicted example, the UE 102 includes a central processing unit (CPU) or other general processor 302, a system memory 304, a Wi-Fi modem 310, and an RF modem 316. In some embodiments, the UE 102 includes additional circuitry to support wireless network connections, such as a Wi-Fi transceiver, at least one Wi-Fi antenna suitable for RF signaling and signal processing in one or more frequency bands typically associated with Wi-Fi connections, an RF transceiver, and at least one RF antenna suitable for RF signaling and signal processing in frequency bands associated with cellular RATs. Further, it will be appreciated that the UE 102 can include a number of additional components omitted from FIG. 3 for ease of illustration, including, for example, one or more displays, one or more touchscreens, keypads, mice, touchpads, microphones, speakers, and other user input/output devices, one or more sensors, batteries or other power sources, graphical processing units (GPUs) or other coprocessors, and the like.

As a general operational overview, the general processor 302 executes executable instructions from a software stack that includes an operating system (OS) and one or more user software applications and which further can include the protocol stacks executed by processors of the RF modem 316 and WLAN modem 310. The OS manages the general operation of the various hardware components of the UE 102 as well as supports the execution of the one or more user software applications, with the user software applications typically accessed from system memory 304 for execution by the general processor 302. During execution, one or more processes of the OS or the user software application (referred to generally as “local processes”) may seek to wirelessly communicate data with the PDN 104.

In the event that a local process is seeking to communicate data with the PDN 104, the general processor 302 can employ one of a Wi-Fi connection, communicating data via the Wi-Fi modem 310, and a cellular RAN connection, communicating data via the RF modem 316. For either of these connections the corresponding modem can handle lower level operations associated with the corresponding network protocol, such as some or all of the physical, data link, and network layers, while the OS and the user software application executing at the general processor 302 support the higher-level layers of the network protocol, such as the transport, session, presentation, and application layers.

In some embodiments, one or both of the Wi-Fi modem 310 and the RF modem can report link-layer information or other information indicative of a link quality for the corresponding network connection. Thus, for example, the Wi-Fi modem 310 can report link-layer information to the operating system indicative of a throughput or other aspect of link quality for the Wi-Fi connection with the PDN 104. Examples of such link-layer information include a link speed, a channel utilization, and a packet error rate for one or both of an uplink portion and a downlink portion of the network connection. Based on the link-layer or other information the OS can generate a throughput score for the corresponding connection. In some embodiments, the throughput score is generated by determining a weighted sum of the link speed, channel utilization, and packet error rate for both the uplink and downlink portions of the network connection, where the weights for each characteristic are determined during a design or testing phase of the UE 102.

In some embodiments, the operating system executing at the general processor 302 selects between the Wi-Fi connection and the RAN connection for data communication, based on the relationship between the respective throughput scores for each connection. Thus, for example, the operating system can select the network connection having the higher throughput score and communicate data with the PDN 104 via the selected network connection by sending and receiving data via the corresponding one of the modems 310 and 316.

In other embodiments, the operating system calculates the throughput score for the Wi-Fi connection only, based upon the link-layer information provided by the Wi-Fi modem 310, and selects between the Wi-Fi connection and the RAN connection based on a comparison of the throughput score to a specified threshold. For example, if the throughput score is less than the specified threshold, indicating a poor-quality Wi-Fi connection, the operating system can select the RAN connection for data communication, thereby enhancing the user experience.

In some embodiments, in addition to the throughput score, the operating system selects between the Wi-Fi connection and the RAN connection based on connection selection criteria 330 stored at the memory 304. An example of the connection selection criteria 330 is illustrated at FIG. 4 in accordance with some embodiments.

In the depicted example, the connection selection criteria 330 includes application requirements 440, meter characteristics 441 , and location information 442. The application requirements 440 are a data structure storing information indicative of link quality requirements for applications to be executed at the general processor 302. The application requirements 440 can indicate link quality requirements for individual applications, for different application types, for particular processes invoked by an individual application, and the like, or any combination thereof. For example, the application requirements 440 can indicate that a video streaming application requires relatively high throughput, while an audio streaming program or game program requires relatively low throughput. When selecting between the Wi-Fi network connection and the RAN connection, the general processor 302 can select the network that more closely satisfies the link quality requirements for the executing application.

To illustrate via an example, in some cases the Wi-Fi network connection can have lower throughput than the RAN connection, but can have other benefits, such as a lower cost to the user, lower power consumption, and the like. Accordingly, in response to executing an application associated with low throughput requirements, such as a game program, the general processor 302 can select the Wi-Fi network connection for data communication. In response to the user initiating an application with higher throughput requirements, such as a video streaming program the general processor 302 can change the selected network connection to the RAN connection. The GP 302 can thereby satisfy the different link quality requirements for different applications while conserving system resources, reducing user costs, and the like, or a combination thereof.

The meter characteristics 441 is a data structure storing information indicating whether and how each type of network connection is metered by a corresponding service provider. For example, the meter characteristics 441 can indicate that the RAN connection is a metered connection such that a service provider charges an owner of the UE 102 to use the RAN connection. In some embodiments, the meter characteristics 441 can indicate more granular metering information, such as a metering level currently associated with the RAN connection. For example, the meter characteristics can indicate whether an owner of the UE 102 has exceeded a data limit associated with the RAN connection, and the current cost to the owner of communicating via the RAN connection. When selecting between the Wi-Fi network connection and the RAN connection, the general processor 302 can employ the meter characteristics 441 to reduce costs imposed on the owner of the UE 102. For example, when the meter characteristics 441 indicate that the owner is below the data limit (and therefore using the RAN imposes no charges on the owner), the general processor 302 can select between the Wi-Fi network connection and the RAN connection based only on the relative link quality identified for each network connection. In response to the meter characteristics 441 indicating that the owner has exceeded the data limit, the general processor 302 can prevent the RAN connection from being selected unless the identified link quality for the Wi-Fi connection is below a specified threshold.

The location information 442 is a data structure storing historical link quality information for different locations, such as different cells of the RAN 115. The GP 302 can record the location information 442 over time and can use the location information 442 when selecting between the Wi-Fi network connection and the RAN connection. For example, in some cases the location information 442 can indicate that the link quality associated with the RAN connection in a given location, such as in a given network cell, has historically been relatively low. In response, the general processor 302 can select the Wi-Fi connection for data communication, even if the current link quality for the RAN connection is identified as higher than the current link quality for the Wi-Fi connection. The general processor 302 thereby avoids selecting a network connection that may temporarily have better link quality but is likely to degrade overtime.

FIG. 5 is a flow diagram of a method 500 of selecting a network connection based on link quality characteristics in accordance with some embodiments. The method 500 is described with respect to an example implementation and the communication network 100 of FIG. 1. At block 502, the UE 102 determines one or more link quality characteristics, such as throughput or latency, of the Wi-Fi connection via the WAP 112. At block 504, the UE 102 determines the link quality characteristics for the RAN connection via the base station 113. At block 506, the UE 102 selects the network connection having the higher link quality, as indicated by the different link quality characteristics, and communicates data with the PDN 104 via the selected network connection.

FIG. 6 is a flow diagram of a method 600 of a UE selecting a RAN connection for data communication based on predicted link quality of an available Wi-Fi connection in accordance with some embodiments. The method 600 is described with respect to an example implementation at the communication network 100 of FIG. 1. At block 602, the UE 102 monitors link-layer characteristics of the Wi-Fi network connection via the WAP 112, such as one or more of a link speed, channel utilization, and packet error rate for the Wi-Fi network connection. At block 604, the UE 102 predicts link quality, such as a throughput, for the Wi-Fi network connection based on the monitored link-layer characteristics. In some embodiments, the UE 102 predicts the link quality by accessing a table indicating a historical relationship between link-layer characteristics and link quality. The table can include previously identified relationships between a particular link-layer characteristic measured at the UE 102, or at a test device and a corresponding measurement of link quality. The UE 102 selects the link quality from the table entry that most closely matches the monitored link-layer characteristic.

At block 606, the UE 102 determines whether the predicted link quality for the Wi-Fi network connection is less than a specified threshold. If not, the UE 102 selects, or continues to select, the Wi-Fi network connection to communicate data with the PDN 104. If, at block 606, the UE determines that the predicted link quality for the Wi-Fi network connection is less than the threshold, the method flow proceeds to block 608 and the UE 102 selects the RAN connection, via the base station 113, to communicate data with the PDN 104.

In some embodiments, certain aspects of the techniques described above may be implemented by one or more processors of a processing system executing software. The software comprises one or more sets of executable instructions stored or otherwise tangibly embodied on a non-transitory computer readable storage medium. The software can include the instructions and certain data that, when executed by the one or more processors, manipulate the one or more processors to perform one or more aspects of the techniques described above. The non-transitory computer readable storage medium can include, for example, a magnetic or optical disk storage device, solid state storage devices such as Flash memory, a cache, random access memory (RAM) or other non-volatile memory device or devices, and the like. The executable instructions stored on the non-transitory computer readable storage medium may be in source code, assembly language code, object code, or other instruction format that is interpreted or otherwise executable by one or more processors. A computer readable storage medium may include any storage medium, or combination of storage media, accessible by a computer system during use to provide instructions and/or data to the computer system. Such storage media can include, but is not limited to, optical media (e.g., compact disc (CD), digital versatile disc (DVD), Blu-Ray disc), magnetic media (e.g., floppy disc , magnetic tape, or magnetic hard drive), volatile memory (e.g., random access memory (RAM) or cache), non-volatile memory (e.g., read-only memory (ROM) or Flash memory), or microelectromechanical systems (MEMS)-based storage media. The computer readable storage medium may be embedded in the computing system (e.g., system RAM or ROM), fixedly attached to the computing system (e.g., a magnetic hard drive), removably attached to the computing system (e.g., an optical disc or Universal Serial Bus (USB)-based Flash memory), or coupled to the computer system via a wired or wireless network (e.g., network accessible storage (NAS)).

Note that not all of the activities or elements described above in the general description are required, that a portion of a specific activity or device may not be required, and that one or more further activities may be performed, or elements included, in addition to those described. Still further, the order in which activities are listed are not necessarily the order in which they are performed. Also, the concepts have been described with reference to specific embodiments. However, one of ordinary skill in the art appreciates that various modifications and changes can be made without departing from the scope of the present disclosure as set forth in the claims below. Accordingly, the specification and figures are to be regarded in an illustrative rather than a restrictive sense, and all such modifications are intended to be included within the scope of the present disclosure.

Benefits, other advantages, and solutions to problems have been described above with regard to specific embodiments. However, the benefits, advantages, solutions to problems, and any feature(s) that may cause any benefit, advantage, or solution to occur or become more pronounced are not to be construed as a critical, required, or essential feature of any or all the claims. Moreover, the particular embodiments disclosed above are illustrative only, as the disclosed subject matter may be modified and practiced in different but equivalent manners apparent to those skilled in the art having the benefit of the teachings herein. No limitations are intended to the details of construction or design herein shown, other than as described in the claims below. It is therefore evident that the particular embodiments disclosed above may be altered or modified and all such variations are considered within the scope of the disclosed subject matter. Accordingly, the protection sought herein is as set forth in the claims below.

Claims

WHAT IS CLAIMED IS:

1. A method comprising: determining, at a user equipment, a first link quality associated with a first network connection via a Wi-Fi access point; determining, at a user equipment, a second link quality associated with a second network connection via a Radio Access Network (RAN); and selecting one of the first network connection and the second network connection to communicate data based on the first link quality and the second link quality.

2. The method of claim 1 , wherein: determining the first link quality comprises determining a first throughput associated with the first network connection; and determining the second link quality comprises determining a second throughput associated with the second network connection.

3. The method of claim 1 , wherein: determining the first link quality comprises determining the first link quality based on link-layer information associated with the first network connection.

4. The method of claim 3, wherein: the link-layer information comprises at least one of link speed information, channel utilization information, and a packet error rate.

5. The method of claim 1 , wherein determining the first link quality comprises: determining an uplink quality associated with an uplink portion of the first network connection; determining a downlink quality associated with a downlink portion of the first network connection; and determining the first link quality based on the uplink quality and the downlink quality.

6. The method of any preceding claim, wherein the selecting comprises: selecting one of the first network connection and the second network connection based on link quality requirements associated with an application executing at the UE.

7. The method of any preceding claim, wherein the selecting comprises: selecting one of the first network connection and the second network connection based on link quality requirements associated with user settings stored at the UE.

8. The method of any preceding claim, wherein the selecting comprises: selecting one of the first network connection and the second network connection based on connection metering requirements associated with the second network connection.

9. The method of any preceding claim, wherein: determining the first link quality comprises predicting the first link quality.

10. The method of claim 9, wherein predicting the first link quality comprises predicting the first link quality based on a machine learning model trained to predict link quality.

11. The method of any preceding claim, wherein determining the second link quality comprises determining the second link quality for a network slice associated with the second network.

12. A method comprising: determining, at a user equipment, a link quality associated with a first network connection via a Wi-Fi access point; and in response to determining the first link quality is below a threshold, selecting a second network connection to communicate data, wherein the second network connection is via a Radio Access Network (RAN).

13. The method of claim 12, wherein: determining the link quality comprises determining a first throughput associated with the first network connection based on link-layer information associated with the first network connection.

14. The method of claim 13, wherein: the link-layer information comprises at least one of link speed information, channel utilization information, and a packet error rate.

15. A user equipment (UE) comprising: a first network interface to establish a first network connection via a Wi-Fi access point; a first network interface to establish a second network connection via a Radio Access Network (RAN); and a processor to select one of the first network connection and the second network connection to communicate data based on an identified first link quality associated with the first network connection and a second link quality associated with the second network connection.

16. The UE of claim 15, wherein the processor is to: determine the first link quality by determining a first throughput associated with the first network connection; and determine the second link quality by determining a second throughput associated with the second network connection.

17. The UE of claim 15, wherein the processor is to: determine the first link quality based on link-layer information associated with the first network connection.

18. The UE of claim 17, wherein: the link-layer information comprises at least one of link speed information, channel utilization information, and a packet error rate.

19. The UE of claim 15, wherein the processor is to determine the first link quality by: determining an uplink quality associated with an uplink portion of the first network connection; determining a downlink quality associated with a downlink portion of the first network connection; and determining the first link quality based on the uplink quality and the downlink quality.

20. The UE of any of claims 15 to 19, wherein the processor is to: select one of the first network connection and the second network connection based on link quality requirements associated with an application executing at the UE.

21. The UE of any of claims 15 to 20, wherein the processor is to: select one of the first network connection and the second network connection based on link quality requirements associated with user settings stored at the UE.

22. The UE of any of claims 15 to 21 , wherein the processor is to: select one of the first network connection and the second network connection based on connection metering requirements associated with the second network connection.

23. The UE of any of claims 15 to 22, wherein the processor is to: determine the first link quality by predicting the first link quality.

24. The UE of claim 23, wherein predicting the first link quality comprises predicting the first link quality based on a machine learning model trained to predict link quality.

25. The UE of any of claims 15 to 24, wherein determining the second link quality comprises determining the second link quality for a network slice associated with the second network.

26. Computer-executable instructions which, when executed by one or more processors, cause the one or more processors to perform the method of any one of claims 1 to 14.