CN117715130A

CN117715130A - Network switching method, device, terminal and storage medium

Info

Publication number: CN117715130A
Application number: CN202311713934.5A
Authority: CN
Inventors: 袁正; 宋锐星; 薛学儒; 赵星; 向祺鑫; 李雄; 杨庆; 吴咸樾; 李销; 王毛; 刘得煌; 何坚
Original assignee: Guangdong Oppo Mobile Telecommunications Corp Ltd
Current assignee: Guangdong Oppo Mobile Telecommunications Corp Ltd
Priority date: 2023-12-13
Filing date: 2023-12-13
Publication date: 2024-03-15

Abstract

The embodiment of the application discloses a network switching method, a device, a terminal and a storage medium, and belongs to the technical field of network switching. The method comprises the following steps: acquiring current radio frequency fingerprint data, wherein the current radio frequency fingerprint data represents radio frequency characteristics of a current connected network; determining a target subspace where the current scene is located based on the spatial attribute of each subspace in the current scene and the current radio frequency fingerprint data; generating a network recommendation result based on a communication semantic knowledge graph corresponding to the target subspace, wherein the communication semantic knowledge graph represents the association relationship between network data of different dimensions in the target subspace; based on the current network connection condition and the network recommendation result, performing network switching; by adopting the scheme provided by the embodiment of the application, the network switching efficiency can be improved, and the network experience is optimized.

Description

Network switching method, device, terminal and storage medium

Technical Field

The embodiment of the application relates to the technical field of network switching, in particular to a network switching method, a device, a terminal and a storage medium.

Background

With the development of the mobile internet, there is an increasing demand for network connectivity. For different application scenarios, in the process of selecting different network connection modes, the problems of network bandwidth, signal strength, stability and the like are often faced, so that the user network experience is affected.

In the related art, network switching is performed by detecting the network quality of the currently connected network according to the terminal. For example, the network quality is evaluated by a series of indexes such as signal strength, signal-to-noise ratio, bit error rate and the like, so that when the signal strength is lower than the strength threshold value, the network scanning is triggered, and after the available network with stronger signal strength is scanned, the network switching is performed.

Obviously, in the related art, network switching attempt is only performed under the condition of poor triggering network quality, so that the communication network switching is lagged, and obvious network blocking or even unusable conditions easily occur.

Disclosure of Invention

The embodiment of the application provides a network switching method, a network switching device, a terminal and a storage medium. The technical scheme is as follows:

in one aspect, an embodiment of the present application provides a network handover method, where the method includes:

acquiring current radio frequency fingerprint data, wherein the current radio frequency fingerprint data represents radio frequency characteristics of a current connected network;

determining a target subspace in which the current is positioned based on the spatial attribute of each subspace in the current scene and the current radio frequency fingerprint data, wherein the subspace is obtained by spatial clustering based on the historical radio frequency fingerprint data, and the spatial attribute comprises the historical radio frequency fingerprint characteristics of the subspace;

Generating a network recommendation result based on a communication semantic knowledge graph corresponding to the target subspace, wherein the communication semantic knowledge graph represents the association relationship between network data of different dimensions in the target subspace;

and switching the network based on the current network connection condition and the network recommendation result.

In another aspect, an embodiment of the present application provides a network switching device, where the device includes:

the data acquisition module is used for acquiring current radio frequency fingerprint data, wherein the current radio frequency fingerprint data represents radio frequency characteristics of a current connected network;

the space positioning module is used for determining a current target subspace based on the space attribute of each subspace in the current scene and the current radio frequency fingerprint data, wherein the subspace is obtained by spatial clustering based on the historical radio frequency fingerprint data, and the space attribute comprises the historical radio frequency fingerprint characteristics of the subspace;

the result generation module is used for generating a network recommendation result based on the communication semantic knowledge graph corresponding to the target subspace, wherein the communication semantic knowledge graph represents the association relationship between network data of different dimensions in the target subspace;

And the network switching module is used for switching the network based on the current network connection condition and the network recommendation result.

In another aspect, embodiments of the present application provide a terminal, where the terminal includes a processor and a memory, where the memory stores at least one computer instruction, where the at least one computer instruction is loaded and executed by the processor to implement a network handover method as described in the above aspect.

In another aspect, embodiments of the present application provide a computer-readable storage medium having stored therein at least one computer instruction that is loaded and executed by a processor to implement a network handover method as described in the above aspects.

In another aspect, embodiments of the present application provide a computer program product comprising computer instructions stored in a computer-readable storage medium. The processor of the terminal reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions so that the terminal performs the network handover method provided in various alternative implementations of the above aspect.

In the embodiment of the application, the current target subspace is determined by acquiring the radio frequency fingerprint data at the current moment and performing space positioning according to the space attribute of each subspace in the current scene and the current radio frequency fingerprint data, so that the network recommendation result is generated according to the communication semantic knowledge graph corresponding to the target subspace, and network switching is performed based on the current network connection condition and the network recommendation result. By adopting the scheme provided by the embodiment of the application, the network switching efficiency in the complex subspace can be improved, and the network experience is optimized.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 illustrates a flow chart of a network switching method provided by an exemplary embodiment of the present application;

FIG. 2 illustrates an ontology structure diagram of a communication semantic knowledge graph provided by an exemplary embodiment of the present application;

FIG. 3 illustrates a block diagram of a network entry type provided by an exemplary embodiment of the present application;

FIG. 4 illustrates a block diagram of network quality metrics provided by an exemplary embodiment of the present application;

FIG. 5 illustrates a flowchart for constructing a communication semantic knowledge graph, as provided by an exemplary embodiment of the present application;

FIG. 6 illustrates a flowchart of generating a subspace database provided by an exemplary embodiment of the present application;

FIG. 7 illustrates a schematic view of a subspace partition scenario provided by an exemplary embodiment of the present application;

FIG. 8 illustrates a flowchart of a radio frequency fingerprint data clustering provided in an exemplary embodiment of the present application;

fig. 9 is a flowchart of a network handover method according to another exemplary embodiment of the present application;

FIG. 10 illustrates a flowchart for correcting network recommendations provided by an exemplary embodiment of the present application;

FIG. 11 illustrates a flow chart of network switching provided by an exemplary embodiment of the present application;

FIG. 12 illustrates a flowchart of generating a path list provided by an exemplary embodiment of the present application;

FIG. 13 illustrates a flow chart for generating a network recommendation list and model training provided by an exemplary embodiment of the present application;

FIG. 14 illustrates a flowchart for performing network handoff provided by one exemplary embodiment of the present application;

fig. 15 shows a flowchart for performing network handover according to another exemplary embodiment of the present application;

fig. 16 is a flowchart of a network switching method according to another exemplary embodiment of the present application;

FIG. 17 illustrates a flow chart of predicting network handoffs provided by one exemplary embodiment of the present application;

fig. 18 is a block diagram illustrating a network switching device according to an exemplary embodiment of the present application;

fig. 19 shows a schematic structural diagram of a terminal according to an exemplary embodiment of the present application.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the present application more apparent, the embodiments of the present application will be described in further detail below with reference to the accompanying drawings.

Therefore, in the related art, network switching is only attempted when the network quality of the currently connected network is detected to be low, so that the user can obviously feel the problem of network blocking before the network switching, thereby reducing the network use experience of the user. For complex multi-subspace and multi-network scenes, as the user moves in the scene, the terminal needs to switch the network in time according to the current position, i.e. the condition of network switching cannot be just the network quality of the current connected network.

Therefore, the embodiment of the application provides a network switching method, which is used for determining the current target subspace according to the current radio frequency fingerprint data and the spatial attribute of each subspace in the current scene, so that a network recommendation result is generated based on the communication semantic knowledge graph corresponding to the target subspace, and further, network switching is performed based on the current network connection condition and the network recommendation result, thereby improving the network switching efficiency and optimizing the network experience.

Referring to fig. 1, a flowchart of a network switching method according to an exemplary embodiment of the present application is shown, where the method includes the following steps:

Step 101, acquiring current radio frequency fingerprint data, wherein the current radio frequency fingerprint data characterizes radio frequency characteristics of a current connected network.

In some embodiments, the terminal may acquire current radio frequency fingerprint data by initiating a WIFI scan or a Cellular (Cellular) scan.

In the case of acquiring current rf fingerprint data based on WIFI scanning, the current rf fingerprint data is WIFI rf fingerprint data, which may include a media access control (Media Access Control, MAC) address, a service set identifier (Service Set Identifier, SSID), a basic service set identifier (Basic Service Set ID, BSSID), a received signal strength (Received Signal Strength Indicator, RSSI), and a data transmission round trip time of the WIFI device.

In the case of acquiring current rf fingerprint data based on cellular scanning, the current rf fingerprint data is cellular network rf data, which may include a radio access technology (Radio Access Technology, RAT) corresponding to the cellular network, a Cell ID, a physical Cell ID (Physical Cell Identifier, PCI), a frequency point (Absolute Radio Frequency Channel Number, ARFCN), a cellular network code (Public Land Mobile Network Code), a network Bandwidth (Bandwidth), a signal received power (Reference Signal Receiving Power, RSRP), a received signal quality (Reference Signal Receiving Quality, RSRQ), a signal-to-noise-and-interference ratio (Signal to Interference plus Noise Ratio, SINR), and the like.

Step 102, determining a target subspace in which the current target subspace is located based on the spatial attribute of each subspace in the current scene and the current radio frequency fingerprint data, wherein the subspace is obtained by spatial clustering based on the historical radio frequency fingerprint data, and the spatial attribute comprises the historical radio frequency fingerprint characteristics of the subspace.

Optionally, considering that a plurality of candidate networks are generally arranged in a complex scene and are used for meeting the network requirements of users located in different spaces in the scene, in order to improve the network switching efficiency, accurate positioning of the current space where the terminal is located is required to be achieved.

In one possible implementation manner, the terminal may determine the target subspace where the terminal is currently located according to the spatial attribute of each subspace in the current scene and the current rf fingerprint data.

Alternatively, each subspace in the scene may be obtained by spatial clustering based on the historical radio frequency fingerprint data in the scene, and the spatial attribute of the subspace includes the historical radio frequency fingerprint feature of the subspace.

Optionally, the spatial attribute of the subspace may include, in addition to the historical radio frequency fingerprint feature, a historical network usage condition, such as a historical network usage period, a historical network usage application, and the like in the subspace.

Alternatively, the terminal may store the subspace information in the current scene through a subspace database. For example, the number of subspaces, the subspace number, and the spatial attribute, etc., within the current scene may be included in the subspace database.

In one possible implementation manner, after acquiring the current rf fingerprint data, the terminal may acquire spatial attributes of each subspace in the current scene from the subspace database, where the spatial attributes include historical rf fingerprint features, so that the terminal may determine the target subspace where the terminal is currently located based on the current rf fingerprint data and each historical rf fingerprint data.

And step 103, generating a network recommendation result based on the communication semantic knowledge graph corresponding to the target subspace, wherein the communication semantic knowledge graph represents the association relation between the network data of different dimensions in the target subspace.

In some embodiments, considering that there may be multiple candidate networks or multiple network connection manners in the subspace, in order to improve accuracy of network switching, after determining the current target subspace, the terminal may generate a network recommendation result according to the communication semantic knowledge graph corresponding to the target subspace, and recommend each candidate network according to the priority through the network recommendation result.

Optionally, the communication semantic knowledge graph characterizes the association relationship between the network data of different dimensions in the target subspace. The network data may include network time, network application type, network type, signal strength, signal bandwidth, data transmission rate, network feedback for application, and the like, which is not limited in the embodiments of the present application.

In some embodiments, the terminal may find currently available candidate networks from the communication semantic knowledge graph corresponding to the target subspace according to the current network requirement, and generate the network recommendation result based on each candidate network.

Optionally, the network recommendation result may be a network quality of each candidate network, such as a current network signal strength of each candidate network, or may be a network recommendation score of each candidate network, or may be another network recommendation form, which is not limited in the embodiment of the present application.

And 104, switching the network based on the current network connection condition and the network recommendation result.

In some embodiments, after obtaining the network recommendation result, the terminal needs to determine whether to perform network switching according to the current network connection condition, for example, if the current connected network is better than the recommended candidate network, the terminal does not perform network switching; in case the recommended candidate network is better than the currently connected network, a network handover is performed.

In one possible implementation, the terminal may determine whether to perform a network handover by comparing the signal strength of the currently connected network with the signal strengths of the respective candidate networks. Under the condition that the signal intensity of the current connected network is lower than that of one of the candidate networks, the terminal can switch the network into the candidate network with higher signal intensity.

In summary, in the embodiment of the present application, by acquiring the radio frequency fingerprint data at the current moment, performing spatial positioning according to the spatial attribute of each subspace in the current scene and the current radio frequency fingerprint data, determining the current target subspace, generating the network recommendation result according to the communication semantic knowledge graph corresponding to the target subspace, and performing network switching based on the current network connection condition and the network recommendation result. By adopting the scheme provided by the embodiment of the application, the network switching efficiency in the complex subspace can be improved, and the network experience is optimized.

In some embodiments, in order to improve the accuracy of generating the network recommendation result, the accuracy of the communication semantic knowledge graph corresponding to each subspace needs to be improved first.

Optionally, the terminal may collect a large amount of historical network data corresponding to each subspace, and construct a communication semantic knowledge graph based on the historical network data.

In one possible implementation manner, the terminal may collect historical network data corresponding to each subspace, where the historical network data may include historical network time, historical network application, application type of the historical network application, historical radio frequency fingerprint data, historical protocol measurement data and historical network experience data.

Optionally, the historical radio frequency fingerprint data may include network radio frequency fingerprints and network Layer2 data, the network radio frequency fingerprints may be divided into cellular radio frequency fingerprints and WIFI radio frequency fingerprints, and the network Layer2 data may be divided into cellular Layer2 data and WIFI Layer2 data.

The cellular radio frequency fingerprint may include a radio access technology (Radio Access Technology, RAT) corresponding to the cellular network, a Cell ID, a physical Cell ID (Physical Cell Identifier, PCI), a frequency point (Absolute Radio Frequency Channel Number, ARFCN), a cellular network code (Public Land Mobile Network Code), a network Bandwidth (Bandwidth), a signal received power (Reference Signal Receiving Power, RSRP), a signal received quality (Reference Signal Receiving Quality, RSRQ), a signal-to-noise-and-interference ratio (Signal to Interference plus Noise Ratio, SINR), and the like.

The cellular Layer2 data may include a download (RX) transmission Rate, an upload (TX) transmission Rate, an Uplink Grant (Uplink Grant), an amount of buffered data to be uploaded, a buffer status report (Buffer Status Report, BSR), an Uplink Error Rate (BLER), a downlink BLER, and the like.

The WIFI radio frequency fingerprint may include, among other things, a media access control (Media Access Control, MAC) address of the WIFI device, a service set identification (Service Set Identifier, SSID), a basic service set identification (Basic Service Set ID, BSSID), a received signal strength (Received Signal Strength Indicator, RSSI), and a data transmission round trip time.

The WIFI Layer2 data may include an RX transmission rate, a TX transmission rate, a channel utilization (Channel Utilization Ratio), a channel load strength, a total TX success number, a total TX retry number, a total TX failure number, a total RX success number, a total current channel busy time, a total current channel working time, and the like.

Optionally, the historical protocol measurement data may be transmission control protocol (Transmission Control Protocol, TCP)/user datagram protocol (User Datagram Protocol, UDP) measurement data, which may include an application packet name, a Round-Trip Time (RTT), a number of packets, a packet size, a number of retransmission packets, a retransmission rate, a domain name system (Domain Name System, DNS) query delay value, and the like.

Optionally, the historical network experience data may be user experience data, and may include application package names, user experience data of application feedback, user experience scoring data, and the like.

In a possible implementation manner, after the historical network data corresponding to each subspace is collected, in order to improve the construction efficiency of the knowledge graph, the terminal may further perform data cleaning and statistics on the historical network data, so as to obtain the processed historical network data.

Optionally, the terminal may filter the historical network data according to the length of the network usage time, for example, filter the historical network data generated in the network connection process in a shorter time. The terminal can also filter other historical network using data according to the network using experience data, for example, the historical network using data with low user experience score is filtered.

Optionally, the terminal may further classify the historical network data with network time periods, so that the statistical values of different network data in each network time period are calculated by taking the network time period as a unit, including but not limited to a mean value, a maximum value, a minimum value, a standard deviation, a sample number, and the like.

Further, after the historical network data is processed, the terminal can generate a communication semantic knowledge graph corresponding to each subspace according to the processed historical network data.

In one possible implementation, the terminal may first determine a feature dimension corresponding to each historical network data, for example, the feature dimension may be classified into a demand feature, a network feature, an evaluation feature, and the like. And the terminal classifies the historical network data according to the characteristic dimension corresponding to each historical network data, and determines the association relation between the historical network data with different dimensions, so that the communication semantic knowledge graph corresponding to each subspace can be obtained by taking each historical network data as a node and the association relation between the historical network data as an edge to generate association paths between the different historical network data.

In an illustrative example, the demand features may include network time, network application, and network application type, the network features may include radio frequency fingerprint data, protocol measurement data, and the evaluation features may include network experience data, so that the terminal may determine each feature entity and relationship according to knowledge graph construction principles.

Illustratively, as shown in fig. 2, the terminal may set the body structure of the communication semantic knowledge graph to be centered on network requirements and network connection portals, where each network requirement (Connection Demand) corresponds to a set of network time (Service Time Slot), network application type, subspace (Subspace), and each network connection portal corresponds to a set of network time, subspace, network portal type (Network Access Type), and network quality index (Network Access KPI).

Illustratively, as shown in fig. 3, the network entry types include three types, which are a 5G network (5 gaccesstype), a 4G network (4G Access Type), and a WIFI network (WIFI Access Type), respectively. Different network parameters are respectively corresponding to different network types, for example, a 5G network and a 4G network correspond to a cell identifier, a physical cell identifier, a frequency point, a cell network code and the like; the WIFI network corresponds to a basic service set identifier, a media access control address, a service set identifier, and the like.

Schematically, as shown in fig. 4, the network quality indexes are also corresponding to three types, namely, 5G network quality (5G Access KPI), 4G network quality (4G Access KPI) and WIFI network quality (WIFI Access KPI). Different index parameters are respectively corresponding to each network quality index, for example, the 5G network quality and the 4G network quality are respectively corresponding to a cellular signal strength, a technical service quality index (Quality of Service, qoS) and a user experience index (Quality of Experience, qoE), and the WIFI network quality is corresponding to a WIFI signal strength, a technical service quality index and a user experience index.

In some embodiments, considering that there may be a plurality of terminals in the same scene, that is, there are a plurality of network data for the terminals, for example, a family is taken as an example, each family member corresponds to at least one terminal device, and different terminal devices correspond to different historical network data, so in order to improve the integrity of constructing the communication semantic knowledge graph, the server may also construct the communication semantic knowledge graph corresponding to each subspace in the current scene.

In a possible implementation manner, the server acquires historical network data corresponding to each terminal in the scene, and integrates and counts the historical network data so as to construct a communication semantic knowledge graph corresponding to each subspace. And after the terminal determines the target subspace, the communication semantic knowledge graph of the subspace can be acquired from the server, so that a network recommendation result is generated.

Referring to fig. 5, a flowchart for constructing a communication semantic knowledge graph according to an exemplary embodiment of the present application is shown. Firstly, a terminal acquires historical network data corresponding to each subspace in a current scene, builds communication semantic features, and further obtains the processed historical network data through filtering, screening and statistical processing of the historical network data, so that a communication semantic knowledge graph is generated based on the processed historical network data.

In the embodiment, the historical network data with different dimensions are obtained, so that the association relation between the historical network data is determined according to different feature dimensions, the communication semantic knowledge graph is generated, and the knowledge graph construction efficiency is improved. And the server builds the communication semantic knowledge graph according to the historical network data corresponding to each terminal in the current scene, so that the integrity of knowledge graph construction is improved, and the network recommendation result can be generated more accurately.

In some embodiments, in order to improve the accuracy of generating the network recommendation result, besides optimizing the communication semantic knowledge graph, the accuracy of spatial positioning needs to be improved, that is, the terminal needs to accurately divide each subspace in the scene first.

In a possible implementation manner, the terminal needs to collect the historical radio frequency fingerprint data corresponding to different scanning conditions under the current scene, so that under the condition that the data volume of the collected historical radio frequency fingerprint data reaches a quantity threshold, the terminal can calculate correlation coefficients between every two historical radio frequency fingerprint data to obtain second correlation coefficients corresponding to each historical radio frequency fingerprint data, and further perform spatial clustering according to the second correlation coefficients corresponding to each historical radio frequency fingerprint data to obtain each subspace corresponding to the current scene.

Alternatively, the second correlation coefficient calculating method may be a Pearson (Pearson) correlation coefficient calculating method, a Cosine (Cosine) similarity calculating method, an Adjusted Cosine (Adjusted Cosine) similarity calculating method, or the like, which is not limited in the embodiment of the present application.

Optionally, the collection mode of the historical radio frequency fingerprint data can be set according to different scanning conditions, wherein the scanning conditions can be set according to a terminal state, a screen state, a network connection state and a charging and discharging state.

For example, in the mobile and bright screen state, and the connected WIFI network, the terminal may acquire WIFI physical layer and data link layer (L2) data every 3 seconds, and acquire 1 second cellular radio frequency fingerprint data after receiving the WIFI scanning result, where the acquisition time is no longer than 3 minutes.

For another example, under the condition that the terminal is in a static and bright screen state and is connected with a WIFI network, the terminal can acquire WIFI L2 data once every 60 seconds, acquire 1 second of cellular radio frequency fingerprint data after receiving a WIFI scanning result, and the acquisition time is not longer than 5 minutes at maximum.

For another example, under the condition that the signal intensity of the WIFI L2 is smaller than the intensity threshold (for example, -67 dbm), the terminal can acquire WIFI L2 data once every 3 seconds, acquire cellular radio frequency fingerprint data once every 5 seconds, and the acquisition time is not longer than 3 minutes.

For another example, in the state of being in a bright screen, and in the process that the WIFI switch is turned on from off, the terminal can obtain cellular radio frequency fingerprint data for 1 second after receiving the WIFI scanning result according to the system WIFI scanning mechanism.

For another example, in the state of being in a bright screen, and in the process that the WIFI switch is turned on to off, the terminal can perform one WIFI scan, and obtain cellular radio frequency fingerprint data of 1 second after receiving the WIFI scan result.

For another example, in case of triggering a lighting-off event or a charging event and connecting to a WIFI network, the terminal acquires connection and disconnection time points of the WIFI network and basic service set identification (Basic Service Set ID, BSSID), and acquires connection and disconnection time points of the cellular network and Cell identification (Cell ID) during connection to the WIFI network.

It should be noted that, the present application only schematically illustrates the scanning condition and the acquisition mode of the acquired rf fingerprint data, and the present application is not limited thereto.

In a possible implementation manner, in the process of collecting the historical radio frequency fingerprint data, the terminal can also utilize a filtering algorithm to filter the data and store the processed historical radio frequency fingerprint data in the radio frequency fingerprint database in sequence, so that under the condition that the data volume in the radio frequency fingerprint database reaches a quantity threshold, the terminal can cluster the radio frequency fingerprint data through a clustering algorithm, thereby forming each subspace, and storing the subspace quantity, the space attribute and the space number in the current scene into the subspace database.

Referring to fig. 6, a flowchart of generating a subspace database according to an exemplary embodiment of the present application is shown. Firstly, a terminal acquires radio frequency fingerprint data by starting WIFI scanning or honeycomb scanning, performs preprocessing and filtering on the radio frequency fingerprint data, and stores the processed radio frequency fingerprint data into a radio frequency fingerprint database, and further under the condition that the data volume in the radio frequency fingerprint database reaches a threshold value, namely a clustering condition is met, the terminal calculates correlation coefficients between the radio frequency fingerprint data through a clustering algorithm, so that the radio frequency fingerprint data is clustered according to the correlation coefficients, namely spatial clustering is performed, a spatial division result of a current scene is obtained, and all subspaces in the current scene are stored into a subspace database. Illustratively, as shown in FIG. 7, the current scene 701 is clustered through space, resulting in 7 subspaces.

Referring to fig. 8, a flowchart of radio frequency fingerprint data clustering is shown, according to an exemplary embodiment of the present application. First, to ensure that each rf fingerprint data is processed, each rf fingerprint data may be first sample marked, thereby cycling through each rf fingerprint data and re-marking the accessed rf fingerprint data. When traversing the first radio frequency fingerprint data, the first radio frequency fingerprint data can be firstly assumed to be a core point of a subspace, so that the radio frequency fingerprint data is taken as a class C, other radio frequency fingerprint data with correlation with the radio frequency fingerprint data is obtained as a sample set N, in the sample set N, similarly, traversing is carried out on each radio frequency fingerprint data in the set, the radio frequency fingerprint data with higher correlation with the core point are classified as the class C until all the radio frequency fingerprint data in the sample set N are traversed, then new core points are redetermined, and the like until all the radio frequency fingerprint data are traversed, so that the space division is carried out on the current scene according to the clustering result of the radio frequency fingerprint data, and each subspace in the current scene can be obtained.

In the above embodiment, by collecting the historical radio frequency fingerprint data in the scene, spatial clustering is performed based on the correlation degree between the historical radio frequency fingerprint data, so as to obtain a plurality of subspaces corresponding to the scene, and efficiency and accuracy of spatial clustering and scene division are improved.

In some embodiments, after dividing and spatially clustering the scene to obtain each subspace in the scene and generating a communication semantic knowledge graph corresponding to each subspace, the terminal can determine a target subspace and generate a network recommendation result according to the current radio frequency fingerprint data.

Referring to fig. 9, a flowchart of a network switching method according to another exemplary embodiment of the present application is shown, where the method includes the following steps:

step 901, acquiring current radio frequency fingerprint data, wherein the current radio frequency fingerprint data characterizes radio frequency characteristics of a current connected network.

The specific implementation of this step may refer to step 101, and this embodiment is not described herein.

Step 902, determining a first correlation coefficient between each subspace and the current radio frequency fingerprint data based on the spatial attribute of each subspace in the current scene and the current radio frequency fingerprint data, wherein the first correlation coefficient represents the correlation degree between the current radio frequency fingerprint data and the subspace.

In some embodiments, in order to improve accuracy of spatial positioning, after acquiring the current rf fingerprint data, the terminal may further calculate a first correlation coefficient between the current rf fingerprint data and each subspace according to the spatial attribute of each subspace in the current scene and the current rf fingerprint data, so as to quantify the degree of correlation between the current rf fingerprint data and each subspace.

Optionally, the first correlation coefficient characterizes a degree of association between the current radio frequency fingerprint data and the subspace, and the higher the degree of association, the greater the first correlation coefficient.

In one possible implementation manner, the terminal may calculate a correlation coefficient between the historical radio frequency fingerprint data and the current radio frequency fingerprint data according to the historical radio frequency fingerprint data corresponding to each subspace, so as to serve as a first correlation coefficient between each subspace and the current radio frequency fingerprint data.

Alternatively, the calculating method of the first correlation coefficient may be a Pearson (Pearson) correlation coefficient calculating method, a Cosine (Cosine) similarity calculating method, an Adjusted Cosine (Adjusted Cosine) similarity calculating method, or the like, which is not limited in the embodiment of the present application.

In step 903, a target subspace where the current rf fingerprint is located is determined based on the first correlation coefficients between each subspace and the current rf fingerprint data, where the first correlation coefficient corresponding to the target subspace is higher than the first correlation coefficients corresponding to other subspaces.

In some embodiments, after obtaining the first correlation coefficients between each subspace and the current rf fingerprint data, the terminal may determine the target subspace in which the current subspace is located according to the degree of correlation represented by each first correlation coefficient.

In one possible implementation manner, the terminal may determine, as the target subspace, a subspace corresponding to the maximum value of the first correlation coefficients according to the sorting result of the first correlation coefficients, that is, the first correlation coefficient corresponding to the target subspace is higher than the first correlation coefficients corresponding to the other subspaces.

Step 904, based on the communication semantic knowledge graph, generating a path list corresponding to the target subspace through graph traversal, wherein the path list comprises a plurality of connection relation pairs formed by different network requirements and network connection inlets.

In some embodiments, after determining the current target subspace and acquiring the communication semantic knowledge graph corresponding to the target subspace, the terminal may traverse the communication semantic knowledge graph in a graph traversal manner, so as to generate a path list corresponding to the target subspace according to the traversal result.

Optionally, the path list includes a plurality of connection relation pairs formed by different network requirements and network connection portals. For example, the network requirements include network time, network application type, and the like, and the network parameters corresponding to the network connection entry may include network type, network quality index, and the like. Under the condition that the network utilization is the same and the network utilization time is different, different network connection inlets can be corresponding, namely different network types and network quality indexes.

In one possible implementation manner, the terminal may first determine a plurality of requirement nodes corresponding to the network requirement in the communication semantic knowledge graph according to a plurality of different network requirements, then, starting from the requirement nodes, traverse other nodes connected with the requirement nodes through a graph in the communication semantic knowledge graph, thereby determining a plurality of network connection inlets conforming to the requirement nodes, further, forming a connection relation pair according to node paths between the network requirement and the network connection inlets, and generating a path list corresponding to the target subspace according to a plurality of connection relation pairs formed by each network requirement and the network connection inlet.

The network demand may include at least one of network time, network application, and network application type. Under the condition that the network utilization is the same and the network utilization time is different, the network utilization system can correspond to different demand nodes, so that different node paths are formed with different network connection inlets.

Alternatively, the process of generating the path list based on the communication semantic knowledge-graph may be performed by the server. In a possible implementation manner, after the server generates the communication semantic knowledge graph corresponding to each subspace based on the historical radio frequency fingerprint data uploaded by each terminal device in the scene, the server may further traverse the communication semantic knowledge graph, and start from each demand node respectively, and form node paths with different network connection inlets, so as to obtain a path list. And in the process of network switching of the terminal, the path list corresponding to each subspace can be directly obtained from the server, or the path list corresponding to the target subspace can be directly obtained.

Step 905, predicting and scoring the candidate networks corresponding to the network connection entries based on each connection relation pair in the path list, to obtain the network confidence and the network recommendation score corresponding to each candidate network.

In some embodiments, after obtaining a plurality of connection relation pairs in the path list, the terminal may predict and score candidate networks corresponding to the network connection entries according to the current network requirement, so as to determine the network confidence level and the network recommendation score corresponding to each candidate network.

The network recommendation score represents the network availability of the candidate network, and the network confidence represents the accuracy of the network recommendation score. In one possible implementation manner, the terminal may determine, according to a target demand node corresponding to the current network demand, a probability from a path of the target demand node from a different node to a network connection entry, so as to determine, based on the path probability corresponding to each connection relation pair, a network confidence level and a network recommendation score corresponding to each candidate network.

In one possible implementation manner, the terminal first obtains a current network demand, where the current network demand may include at least one of a current network time, a current network application, and a network application type, determines a path length corresponding to each connection relation pair in the path list and a network quality of the candidate network, and predicts and scores each candidate network according to the current network demand, the path length, and the network quality of the candidate network, to obtain a network confidence level and a network recommendation score corresponding to each candidate network.

The network confidence coefficient and the path length are in a negative correlation, the network confidence coefficient and the network quality are in a positive correlation, the network recommendation score and the path length are in a negative correlation, and the network recommendation score and the network quality are in a positive correlation.

Step 906, obtaining a current network demand, where the current network demand includes at least one of a current network time, a current network application, and a network application type.

In some embodiments, to improve the efficiency of determining the network recommendation result, the terminal may further determine the network confidence level and the network recommendation score of each candidate network based on the network recommendation model, so that the terminal needs to first obtain the current network usage requirement, where the current network usage requirement includes at least one of the current network usage time, the current network usage application, and the network usage application type.

And step 907, inputting the communication semantic knowledge graph corresponding to the current network demand and the target subspace into a network recommendation model, and outputting the network confidence and the network recommendation score corresponding to each candidate network through the network recommendation model.

In some embodiments, the terminal inputs the current network demand and the communication semantic knowledge graph corresponding to the target subspace into a network recommendation model, and outputs the network confidence and the network recommendation score corresponding to each candidate network through the network recommendation model.

In one possible implementation manner, the terminal may also input the current network requirement and the path list corresponding to the target subspace into the network recommendation model, and output the network confidence degrees and the network recommendation scores corresponding to the candidate networks through the network recommendation model.

Optionally, in order to improve the output efficiency of the network recommendation model, the terminal needs to train the network recommendation model first. In a possible implementation manner, a terminal firstly obtains a sample network requirement, a sample candidate network and application network experience quality corresponding to the sample candidate network, inputs a communication semantic knowledge graph corresponding to the sample network requirement and a target subspace into a network recommendation model, and outputs sample network confidence degrees and sample network recommendation scores corresponding to the sample candidate networks through the network recommendation model, so that the network recommendation model is trained according to the sample network confidence degrees, the sample network recommendation scores and the application network experience quality corresponding to the sample candidate networks, and a trained network recommendation model is obtained.

In a possible implementation manner, in order to reduce the data processing pressure of the terminal in the model training process, the server may be used to train the network recommendation general model based on the network requirement of the general sample, so as to obtain the trained network recommendation general model, so that the terminal directly obtains the path list corresponding to the target subspace and the network recommendation general model from the server side, and based on the network requirement of the individual sample, performs individual training on the network recommendation general model, so as to obtain the corresponding network recommendation model.

Step 908, generating a network recommendation result based on the network confidence level and the network recommendation score corresponding to each candidate network.

In some embodiments, after obtaining the network confidence and the network recommendation score corresponding to each candidate network, the terminal may sort the candidate networks according to the network confidence and the network recommendation score, and generate a network recommendation result.

In one possible implementation manner, considering that the network recommendation result is obtained by predicting and scoring based on objective criteria, in order to make the network recommendation result more fit to the actual historical network usage situation, the terminal may further obtain the historical network connection situation, where the historical network connection situation includes at least one of the historical network connection times, the historical network connection duration and the historical network connection quality, so as to correct the network recommendation score corresponding to the candidate network according to the historical network connection situation, obtain the corrected network recommendation score, and generate the network recommendation result based on the network confidence corresponding to each candidate network and the corrected network recommendation score.

Schematically, as shown in fig. 10, the historical network connection situation can reflect the network usage preference of the user, and the terminal can determine the network usage preference of the user through intention recognition according to the manual network switching operation, the semantic behavior and the semantic scene of the user, so that the network recommendation score of each candidate network is corrected based on the network usage preference, and a final network recommendation result is obtained.

Step 909, performing network screening based on the network recommendation result to obtain a plurality of networks to be switched, wherein the network confidence of the networks to be switched is higher than the confidence threshold, and the network recommendation score of the networks to be switched is higher than the network score of the current connected network.

In some embodiments, considering that the network recommendation result includes the network confidence coefficient and the network recommendation score of each candidate network in the path list, and the network recommendation score of a part of candidate networks is obviously lower, in order to improve the efficiency of subsequent network switching, the terminal may perform network screening according to the network recommendation result to determine a plurality of networks to be switched.

Optionally, the terminal may filter candidate networks with network confidence level lower than the confidence level threshold by setting the confidence level threshold, so that the network confidence level of the network to be switched is higher than the confidence level threshold.

Optionally, the terminal may further determine a network score of the current connected network according to the network quality of the current connected network, and filter candidate networks with a network recommendation score lower than the network score of the current connected network, so that the network recommendation score of the network to be switched is higher than the network recommendation score of the current connected network.

Optionally, the terminal may further determine a score difference between the network recommendation score of each candidate network and the network score of the currently connected network by setting a score difference threshold, and filter the candidate networks with score differences lower than the score difference threshold, so that the network score difference between the network to be switched and the currently connected network is greater than the score difference threshold.

Step 910, performing network switching when the currently connected network is a WIFI network, the network to be switched is a cellular network, the network switch of the cellular network is in an on state, and the network signal strength of the network to be switched meets the signal strength threshold.

In some embodiments, after obtaining a plurality of networks to be switched, the terminal may determine whether to perform network switching according to the network type and the network signal strength of the currently connected network and the networks to be switched.

In one possible implementation manner, when the currently connected network is a WIFI network, and the network to be switched is a cellular network, and the network switch of the cellular network is in an on state, and the network signal strength of the network to be switched meets the signal strength threshold, the terminal can perform network switching.

In one possible implementation, in a case where the currently connected network is a WIFI network, and the network to be switched is a cellular network, and the network switch of the cellular network is in an off state, the terminal does not perform network switching.

In one possible implementation manner, when the currently connected network is a WIFI network and the network to be switched is a cellular network, and the network switch of the cellular network is in an on state, and the network signal strength of the network to be switched does not meet the signal strength threshold, the terminal does not perform network switching.

Step 911, performing network switching when the currently connected network and the network to be switched are both WIFI networks and the network signal strength of the network to be switched meets the signal strength threshold.

In one possible implementation manner, the terminal may perform network switching when the currently connected network and the network to be switched are both WIFI networks and the network signal strength of the network to be switched meets the signal strength threshold.

In one possible implementation manner, in the case that the currently connected network and the network to be switched are both WIFI networks and the network signal strength of the network to be switched does not meet the signal strength threshold, the terminal does not perform network switching.

In one possible implementation, the terminal does not perform a network handover in case the currently connected network is a cellular network.

Referring to fig. 11, a flowchart of network switching provided in an exemplary embodiment of the present application is shown. After the network recommendation result is obtained, the terminal firstly screens the candidate networks according to the network confidence degree and the confidence degree threshold value of each candidate network, after the screened candidate networks are obtained, the terminal calculates a score difference value according to the network recommendation score of the candidate networks and the network score of the current connected network, and further screens the candidate networks according to the score difference value and the difference value threshold value, so that the network to be switched is obtained. And judging whether to switch the network according to the network type of the network which is connected and the network to be switched. If the current connected network is a cellular network, network switching is not performed; performing network switching under the condition that the current connected network and the network to be switched are both WIFI networks and the network signal strength of the network to be switched is higher than a signal strength threshold value; if the current connected network and the network to be switched are both WIFI networks and the network signal strength of the network to be switched is lower than the signal strength threshold value, network switching is not performed; when the current connected network is a WIFI network, the network to be switched is a cellular network, a network switch of the cellular network is in an on state, and the network signal strength of the network to be switched meets a signal strength threshold value, network switching is performed; if the current connected network is a WIFI network, the network to be switched is a cellular network, and a network switch of the cellular network is in a closed state, network switching is not performed; and if the current connected network is a WIFI network, the network to be switched is a cellular network, a network switch of the cellular network is in an on state, and the network signal strength of the network to be switched does not meet the signal strength threshold value, network switching is not performed.

In the embodiment, the accuracy of space positioning is improved by calculating the first correlation coefficient between the current radio frequency fingerprint data and each subspace in the scene and performing space positioning according to the first correlation coefficient; through traversing the communication semantic knowledge graph of the target subspace, connection relation pairs between various network requirements and network connection inlets are determined, so that network recommendation scores of candidate networks are determined according to the current network requirements, and the efficiency and accuracy of network recommendation are optimized; in addition, by correcting the network recommendation score based on the historical network connection condition, the rationality and the suitability of the network recommendation result are increased, and the network experience of the user is optimized.

Referring to fig. 12, a flowchart of generating a path list according to an exemplary embodiment of the present application is shown, where the process may be performed by a terminal or a server.

Under the condition that the server executes the process, the server firstly needs to acquire historical network data stored by the terminal, builds communication semantic features based on the historical network data, and then obtains the processed historical network data through filtering, screening and statistical processing of the historical network data, so that a communication semantic knowledge graph is built according to the historical network data and the communication semantic feature data, the terminal can directly inquire and acquire the communication semantic knowledge graph corresponding to the current scene from the server, and the connection relation pair between each network requirement and a network connection inlet is determined through graph traversal of the communication semantic knowledge graph, so that a path list is generated.

Referring to fig. 13, a flowchart of generating a network recommendation list and model training is shown, provided in an exemplary embodiment of the present application.

Considering that a large amount of data is required for model training, the terminal can firstly perform data detection on the communication semantic knowledge graph to judge whether the model training condition is met, and under the condition that the model training condition is met, the execution of the model training process is started. Firstly, performing graph traversal on a communication semantic knowledge graph to generate sample feature data required by model training, further acquiring currently stored model data from a server, training a network recommendation model based on the sample feature data under the condition that the model data is the latest version data, and readjusting training time according to current state machine logic if abnormal conditions occur in the training process. Furthermore, after model training is completed, in the process of recommending the network by the actual application model, the terminal can retrain the model based on the data with feedback of network experience, so that the output accuracy of the model is improved. And after model training is completed, the terminal can directly store the network recommendation list and the model data.

Referring to fig. 14, a flowchart of performing network handover according to an exemplary embodiment of the present application is shown.

In the scene of actually performing network switching, the terminal can judge whether to perform network switching service according to family fence information, subspace change, service state change and time period change. Under the condition of starting network switching, the terminal can directly acquire the network recommendation score of each candidate network from the stored network recommendation list, and correct the network recommendation score corresponding to the candidate network according to the user preference, so as to generate a network recommendation result. Under the condition that the candidate network information cannot be obtained from the network recommendation list, the terminal needs to call the network recommendation model data, and outputs the network recommendation score corresponding to each candidate network through the network recommendation model, so that the network recommendation score corresponding to the candidate network is corrected based on the user preference, and a network recommendation result is generated.

Referring to fig. 15, a flowchart of performing network handover according to another exemplary embodiment of the present application is shown.

Firstly, a terminal determines a current scene, such as a home, through a positioning subsystem, further acquires information required by network recommendation under the condition of determining that the terminal is in a home state, including a subspace positioning result, current system time, a current network recommendation model and a current running application, and further judges whether network recommendation conditions are met under the condition of acquiring all the information, wherein the network recommendation conditions include at least one of subspace change, time period change and foreground application state change, the terminal executes a network recommendation algorithm flow under the condition that the network recommendation conditions are met, outputs network recommendation scores of candidate networks through the network recommendation model, and performs sorting and screening, so that the network recommendation results are transmitted to a recommendation sub-function, and a network switching unit judges whether network switching is performed according to the network recommendation results.

Referring to fig. 16, a flowchart of a network switching method according to another exemplary embodiment of the present application is shown.

The overall flow of network switching can be divided into two phases, training and application verification.

Firstly, in a training stage, a large amount of historical radio frequency fingerprint data are collected, under the condition that the data amount of the historical radio frequency fingerprint data reaches a quantity threshold value, spatial clustering is carried out based on the historical radio frequency fingerprint data, each subspace in a scene is obtained, a result is stored in a subspace database, further, a communication semantic knowledge graph corresponding to each subspace is built based on historical network data and the historical radio frequency fingerprint data, a plurality of connection relation pairs between network requirements and network connection inlets are determined in a graph traversing mode, and a path list is generated and stored in the path list database.

In the application verification stage, after a terminal acquires a positioning request and acquires current radio frequency fingerprint data, the radio frequency fingerprint data is filtered, and under the condition that the current radio frequency fingerprint data is effective data, the current target subspace is determined through a positioning algorithm based on the current radio frequency fingerprint data and the spatial attribute of each subspace, and then a network recommendation result is determined for a corresponding candidate network according to the current network demand and each connection relation in a path list, so that network switching is performed based on the current connected network and the network recommendation result.

Referring to fig. 17, a flowchart of predicting network switch provided in an exemplary embodiment of the present application is shown. In order to optimize network experience, the terminal needs to monitor the network state in real time, predict the service condition of the current network, judge whether to recommend the network, and under the condition that the network recommendation is required, the terminal needs to recommend the network through the communication semantic knowledge graph corresponding to the target subspace and the current network requirement, thereby optimizing the current wireless network communication state.

Referring to fig. 18, a block diagram of a network switching device according to an exemplary embodiment of the present application is shown. The device comprises:

the data acquisition module 1801 is configured to acquire current radio frequency fingerprint data, where the current radio frequency fingerprint data characterizes a radio frequency feature of a current connected network;

the space positioning module 1802 is configured to determine a current target subspace based on spatial attributes of each subspace in a current scene and the current radio frequency fingerprint data, where the subspace is obtained by spatial clustering based on historical radio frequency fingerprint data, and the spatial attributes include historical radio frequency fingerprint features of the subspace;

The result generation module 1803 is configured to generate a network recommendation result based on a communication semantic knowledge graph corresponding to the target subspace, where the communication semantic knowledge graph represents an association relationship between network data in different dimensions in the target subspace;

the network switching module 1804 is configured to perform network switching based on the current network connection situation and the network recommendation result.

Optionally, the result generating module 1803 includes:

the list generation unit is used for generating a path list corresponding to the target subspace through graph traversal based on the communication semantic knowledge graph, wherein the path list comprises a plurality of connection relation pairs formed by different network demands and network connection inlets;

the scoring unit is used for predicting and scoring the candidate networks corresponding to the network connection inlets based on each connection relation pair in the path list to obtain network confidence degrees and network recommendation scores corresponding to each candidate network;

and the result generation unit is used for generating the network recommendation result based on the network confidence degrees and the network recommendation scores corresponding to the candidate networks.

Optionally, the list generating unit is configured to:

Determining a plurality of demand nodes corresponding to the network demand in the communication semantic knowledge graph based on a plurality of network demands, wherein the network demands comprise at least one of network time, network application and network application type;

starting from the demand node, determining a plurality of network connection inlets conforming to the demand node in the communication semantic knowledge graph through graph traversal;

forming the connection relation pair based on a node path between the network demand and the network connection inlet;

and generating the path list corresponding to the target subspace based on a plurality of connection relation pairs formed by the network requirements and the network connection inlets.

Optionally, the scoring unit is configured to:

acquiring current network demand, wherein the current network demand comprises at least one of current network time, current network application and network application type;

determining the path length corresponding to each connection relation pair in the path list and the network quality of the candidate network;

based on the current network demand, the path length and the network quality of the candidate network, predicting and scoring each candidate network to obtain network confidence and network recommendation score corresponding to each candidate network, wherein the network confidence and the path length are in negative correlation, the network confidence and the network quality are in positive correlation, the network recommendation score and the path length are in negative correlation, and the network recommendation score and the network quality are in positive correlation.

Optionally, the result generating module 1803 further includes:

the network management system comprises a demand acquisition unit, a network management unit and a network management unit, wherein the demand acquisition unit is used for acquiring current network demand, and the current network demand comprises at least one of current network time, current network application and network application type;

the output unit is used for inputting the communication semantic knowledge graph corresponding to the current network demand and the target subspace into a network recommendation model, and outputting the network confidence coefficient and the network recommendation score corresponding to each candidate network through the network recommendation model;

the result generating unit is further configured to generate the network recommendation result based on the network confidence degrees and the network recommendation scores corresponding to the candidate networks.

Optionally, the apparatus further includes:

the demand acquisition module is used for acquiring the demand of the sample network, the sample candidate network and the experience quality of the application network corresponding to the sample candidate network;

the output module is used for inputting the sample network demands and the communication semantic knowledge graph corresponding to the target subspace into the network recommendation model, and outputting the sample network confidence coefficient and the sample network recommendation score corresponding to each sample candidate network through the network recommendation model;

And the training module is used for training the network recommendation model based on the sample network confidence coefficient, the sample network recommendation score and the application network experience quality corresponding to the sample candidate network to obtain the trained network recommendation model.

Optionally, the result generating unit is further configured to:

acquiring historical network connection conditions, wherein the historical network connection conditions comprise at least one of historical network connection times, historical network connection duration and historical network connection quality;

correcting the network recommendation score corresponding to the candidate network based on the historical network connection condition to obtain the corrected network recommendation score;

and generating the network recommendation result based on the network confidence corresponding to each candidate network and the corrected network recommendation score.

Optionally, the network switching module 1804 is configured to:

network screening is carried out based on the network recommendation result to obtain a plurality of networks to be switched, wherein the network confidence of the networks to be switched is higher than a confidence threshold, and the network recommendation score of the networks to be switched is higher than the network score of the current connected network;

Performing network switching when the current connected network is a WIFI network, the network to be switched is a cellular network, a network switch of the cellular network is in an on state, and the network signal strength of the network to be switched meets a signal strength threshold;

and performing network switching under the condition that the current connected network and the network to be switched are both WIFI networks and the network signal strength of the network to be switched meets a signal strength threshold.

Optionally, the network switching module 1804 is further configured to:

if the current connected network is a cellular network, network switching is not performed;

if the current connected network and the network to be switched are both WIFI networks and the network signal strength of the network to be switched does not meet the signal strength threshold, network switching is not performed;

if the current connected network is a WIFI network, the network to be switched is a cellular network, and a network switch of the cellular network is in a closed state, network switching is not performed;

and if the current connected network is a WIFI network, the network to be switched is a cellular network, a network switch of the cellular network is in an on state, and the network signal strength of the network to be switched does not meet the signal strength threshold, network switching is not performed.

Optionally, the spatial positioning module 1802 is configured to:

determining a first correlation coefficient between each subspace and the current radio frequency fingerprint data based on the spatial attribute of each subspace in the current scene and the current radio frequency fingerprint data, wherein the first correlation coefficient represents the degree of correlation between the current radio frequency fingerprint data and the subspace;

and determining the target subspace where the current subspace is located based on the first correlation coefficient between each subspace and the current radio frequency fingerprint data, wherein the first correlation coefficient corresponding to the target subspace is higher than the first correlation coefficients corresponding to other subspaces.

Optionally, the apparatus further includes:

the first data acquisition module is used for acquiring historical network data corresponding to each subspace, wherein the historical network data comprises historical network time, historical network application, application type of the historical network application, historical radio frequency fingerprint data, historical protocol measurement data and historical network experience data;

the data processing module is used for carrying out data cleaning and statistics on the historical network data to obtain the processed historical network data;

and the map generation module is used for generating a communication semantic knowledge map corresponding to each subspace based on the processed historical network data.

Optionally, the map generating module is further configured to:

determining characteristic dimensions corresponding to each historical network data based on the processed historical network data;

based on characteristic dimensions corresponding to each historical network data, determining association relations among the historical network data of different dimensions;

and generating association paths among different historical network data by taking the historical network data as nodes and the association relation among the historical network data as edges, so as to obtain the communication semantic knowledge graph corresponding to each subspace.

Optionally, the apparatus further includes:

the second data acquisition module is used for acquiring historical radio frequency fingerprint data corresponding to different scanning conditions in the current scene;

the coefficient calculation module is used for calculating correlation coefficients between every two historical radio frequency fingerprint data under the condition that the data volume of the historical radio frequency fingerprint data reaches a quantity threshold value, so as to obtain second correlation coefficients corresponding to each historical radio frequency fingerprint data;

and the spatial clustering module is used for performing spatial clustering based on the second correlation coefficient corresponding to each historical radio frequency fingerprint data to obtain each subspace corresponding to the current scene.

Referring to fig. 19, a schematic structural diagram of a terminal according to an exemplary embodiment of the present application is shown.

The terminal 1900 may perform the network switching method of the above embodiment, for example, a smart phone, a smart watch, a vehicle-mounted terminal, a tablet computer, a notebook computer, a desktop computer, a bluetooth headset, and the like. Terminal 1900 may also be referred to as a user device, portable terminal, or the like. The terminal 1900 may also include one or more of the following: a processor 1910, and a memory 1920.

Optionally, the processor 1910 utilizes various interfaces and lines to connect various portions of the overall electronic device, perform various functions of the electronic device, and process data by executing or executing instructions, programs, code sets, or instruction sets stored in the memory 1920, and invoking data stored in the memory 1920. Alternatively, the processor 1910 may be implemented in hardware in at least one of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), programmable logic array (Programmable Logic Array, PLA). The processor 1910 may integrate one or a combination of several of a central processing unit (Central Processing Unit, CPU), an image processor (Graphics Processing Unit, GPU), a Neural network processor (Neural-network Processing Unit, NPU), and a baseband chip, etc. The CPU mainly processes an operating system, a user interface, an application program and the like; the GPU is used for rendering and drawing the content required to be displayed by the touch display screen; the NPU is used to implement artificial intelligence (Artificial Intelligence, AI) functionality; the baseband chip is used for processing wireless communication. It will be appreciated that the baseband chip may not be integrated into the processor 1910 and may be implemented by a single chip.

The Memory 1920 may include a random access Memory (Random Access Memory, RAM) or a Read-Only Memory (ROM). Optionally, the memory 1920 includes a non-transitory computer readable medium (non-transitory computer-readable storage medium). Memory 1920 may be used to store instructions, programs, code, sets of codes, or sets of instructions. The memory 1920 may include a stored program area that may store instructions for implementing an operating system, instructions for at least one function, instructions for implementing the various method embodiments described above, and a stored data area; the storage data area may store data created according to the use of the electronic device, etc.

In addition, those skilled in the art will appreciate that the configuration of the electronic device shown in the above-described figures does not constitute a limitation of the electronic device, and the electronic device may include more or less components than illustrated, or may combine certain components, or may have a different arrangement of components.

Embodiments of the present application also provide a computer readable storage medium having at least one instruction stored therein, where the at least one instruction is loaded and executed by a processor to implement the method described in the above embodiments. Alternatively, the computer-readable storage medium may include: ROM, RAM, solid state disk (SSD, solid State Drives), or optical disk, etc. The RAM may include, among other things, resistive random access memory (ReRAM, resistance Random Access Memory) and dynamic random access memory (DRAM, dynamic Random Access Memory).

Embodiments of the present application also provide a computer program product comprising computer instructions stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions to cause the computer device to perform the network switching method provided in the various alternative implementations of the above aspects.

The foregoing description of the preferred embodiments is merely exemplary in nature and is in no way intended to limit the invention, since it is intended that all modifications, equivalents, improvements, etc. that fall within the spirit and scope of the invention.

Claims

1. A method of network switching, the method comprising:

2. The method of claim 1, wherein generating the network recommendation result based on the communication semantic knowledge-graph corresponding to the target subspace comprises:

based on the communication semantic knowledge graph, traversing through a graph to generate a path list corresponding to the target subspace, wherein the path list comprises a plurality of connection relation pairs formed by different network requirements and network connection inlets;

predicting and scoring the candidate networks corresponding to the network connection inlets based on each connection relation pair in the path list to obtain network confidence degrees and network recommendation scores corresponding to each candidate network;

and generating the network recommendation result based on the network confidence coefficient and the network recommendation score corresponding to each candidate network.

3. The method according to claim 2, wherein the generating, based on the communication semantic knowledge graph, the path list corresponding to the target subspace through graph traversal includes:

4. The method according to claim 2, wherein predicting and scoring the candidate networks corresponding to the network connection entries based on each connection relation pair in the path list to obtain the network confidence and the network recommendation score corresponding to each candidate network includes:

5. The method of claim 1, wherein generating a network recommendation result based on the communication semantic knowledge-graph corresponding to the target subspace further comprises:

inputting the current network demand and the communication semantic knowledge graph corresponding to the target subspace into a network recommendation model, and outputting the network confidence and the network recommendation score corresponding to each candidate network through the network recommendation model;

6. The method of claim 5, wherein the method further comprises:

acquiring a sample network requirement, a sample candidate network and network experience quality corresponding to the sample candidate network;

inputting the sample network demands and the communication semantic knowledge graph corresponding to the target subspace into the network recommendation model, and outputting sample network confidence degrees and sample network recommendation scores corresponding to each sample candidate network through the network recommendation model;

and training the network recommendation model based on the sample network confidence, the sample network recommendation score and the application network experience quality corresponding to the sample candidate network to obtain the trained network recommendation model.

7. The method of claim 2 or 5, wherein generating the network recommendation result based on the network confidence corresponding to each candidate network and the network recommendation score comprises:

8. The method according to claim 2 or 5, wherein the performing network handover based on the current network connection situation and the network recommendation result comprises:

9. The method of claim 8, wherein the method further comprises:

10. The method of claim 1, wherein determining the current target subspace based on the spatial attributes of each subspace in the current scene and the current rf fingerprint data comprises:

11. The method according to claim 1, wherein the method further comprises:

collecting historical network data corresponding to each subspace, wherein the historical network data comprises historical network time, historical network application, application type of the historical network application, historical radio frequency fingerprint data, historical protocol measurement data and historical network experience data;

carrying out data cleaning and statistics on the historical network data to obtain the processed historical network data;

and generating a communication semantic knowledge graph corresponding to each subspace based on the processed historical network data.

12. The method of claim 11, wherein generating a communication semantic knowledge graph corresponding to each subspace based on the processed historical network data comprises:

13. The method according to claim 1, wherein the method further comprises:

collecting historical radio frequency fingerprint data corresponding to different scanning conditions in a current scene;

under the condition that the data quantity of the historical radio frequency fingerprint data reaches a quantity threshold value, calculating correlation coefficients between every two historical radio frequency fingerprint data to obtain second correlation coefficients corresponding to each historical radio frequency fingerprint data;

and carrying out spatial clustering based on second correlation coefficients corresponding to the historical radio frequency fingerprint data to obtain each subspace corresponding to the current scene.

14. A network switching device, the device comprising:

15. A terminal, the terminal comprising a processor and a memory; the memory stores at least one computer instruction for execution by the processor to implement the network handover method of any one of claims 1 to 13.

16. A computer readable storage medium having stored therein at least one computer instruction that is loaded and executed by a processor to implement the network handover method of any of claims 1 to 13.

17. A computer program product, the computer program product comprising computer instructions stored in a computer readable storage medium; a processor of a terminal reads the computer instructions from the computer readable storage medium, the processor executing the computer instructions to cause the terminal to perform the network handover method according to any one of claims 1 to 13.