A method and apparatus for network positioning

Abstract

The application discloses a network positioning method and device, comprising the following steps: obtaining at least one grid covered by a wireless access point and offline fingerprint features corresponding to the grid from a preset fingerprint feature library according to fingerprint information of at least one wireless access point carried by a positioning request of a terminal device; obtaining real-time fingerprint features of the wireless access point according to the fingerprint information; inputting the offline fingerprint features and the real-time fingerprint features of the wireless access point covering the same grid into a first sorting model to score and sort the grid; selecting a preset number of grids from the sorted grids as candidate grids; inputting fingerprint feature data of the candidate grids into a second sorting model to score and sort the candidate grids; and selecting a grid where the terminal device is located from the sorted candidate grids. The application expands the application range of the positioning service from indoor to indoor and outdoor integration by establishing a sorting model, and makes the architecture of the positioning easy to maintain and expand, and meets different scene requirements.

Description

Technical Field

[0001] This invention belongs to the field of positioning technology, and more specifically, relates to a method and apparatus for network positioning. Background Technology

[0002] With the development of mobile communication technology and mobile terminal hardware, network positioning services are playing an increasingly important role in people's production and life. Existing positioning technologies have incorporated deep learning algorithms, but currently suffer from the following shortcomings: (1) Limited applicability: Existing deep learning-based positioning algorithms are generally applicable to indoor environments. They extract deep features from RSSI big data and establish a fingerprint database based on these features to locate users, without considering the needs of outdoor positioning. (2) The fingerprint database uses a single feature, only RSSI features, making it susceptible to environmental changes. (3) Training algorithms mostly use RBM or BP algorithms, which have insufficient generalization ability for positioning problems using classification or regression models. Given these shortcomings, providing engineering-applicable network positioning that also considers indoor and outdoor positioning has become an urgent problem for the industry. Summary of the Invention

[0003] One of the technical problems solved by this invention is to provide a method and apparatus for network positioning.

[0004] According to one embodiment of the present invention, a method for network positioning is provided, comprising:

[0005] Based on the fingerprint information of at least one wireless access point carried in the location request of the terminal device, at least one geographical grid covered by the wireless access point and the offline fingerprint features corresponding to the geographical grid are obtained from a preset fingerprint feature library; based on the fingerprint information, the real-time fingerprint features of the wireless access point are obtained; the fingerprint compression features and real-time fingerprint features of wireless access points covering the same geographical grid are organized together as the fingerprint feature data of the geographical grid; the fingerprint feature data of all geographical grids are input into a first sorting model to score and sort the geographical grids, and a preset number of geographical grids are selected from the sorted geographical grids as candidate geographical grids; the fingerprint feature data of the candidate geographical grids are input into a second sorting model to score and sort the candidate geographical grids, and the geographical grid where the terminal device is located is selected from the sorted candidate geographical grids.

[0006] According to another embodiment of the present invention, a network positioning device is provided, the device comprising: an offline feature calculation module, a real-time feature calculation module, a first sorting module, and a second sorting module; wherein, the offline feature calculation module, based on fingerprint information of at least one wireless access point carried in a positioning request of a terminal device, obtains from a preset fingerprint feature database at least one geographic grid covered by the wireless access point and offline fingerprint features corresponding to the geographic grid; the real-time feature calculation module, based on the fingerprint information, obtains real-time fingerprint features of the wireless access point; the offline fingerprint features and real-time fingerprint features of wireless access points covering the same geographic grid are organized together as fingerprint feature data of the geographic grid; the fingerprint feature data of all geographic grids are input into the first sorting module to score and sort the geographic grids, and a preset number of geographic grids are selected from the sorted geographic grids as candidate geographic grids; the fingerprint feature data of the candidate geographic grids are input into the second sorting module to score and sort the candidate geographic grids, and the geographic grid where the terminal device is located is selected from the sorted candidate geographic grids.

[0007] According to one embodiment of the present invention, a server is provided, the server comprising: a storage device; one or more processors; wherein the storage device is configured to store one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement the aforementioned network positioning method.

[0008] According to one embodiment of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed, implements the aforementioned network positioning method.

[0009] This invention considers not only wireless LAN access points but also cellular base stations, expanding the applicability of positioning from indoors to integrated indoor and outdoor environments. The offline neural network is input not only with features such as signal strength or channel parameters but also with various other features including data collection, association, and IP address. This information is fused and refined through an offline feature compression layer (using a deep neural network model), improving positioning accuracy and expanding application scenarios. Furthermore, an online network is introduced, fusing offline and online features to improve positioning accuracy and robustness while balancing computation and storage. Finally, a learning-to-rank (LTR) ranking algorithm is used for positioning, combining coarse and fine ranking to improve accuracy while reducing computational load. This also makes the overall positioning algorithm architecture easy to maintain and extend, with parameters flexibly adjustable in different scenarios.

[0010] Those skilled in the art will understand that although the following detailed description will be made with reference to the embodiments and drawings, the invention is not limited to these embodiments, but rather the scope of the invention is broad and is intended to be defined only by the appended claims. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0012] Figure 1 This is a system framework diagram according to the present invention.

[0013] Figure 2 This is a flowchart of a network positioning method according to the present invention.

[0014] Figure 3 This is a schematic diagram of a network positioning device according to the present invention. Detailed Implementation

[0015] To facilitate understanding and implementation of the present invention by those skilled in the art, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] To facilitate understanding and implementation of the present invention by those skilled in the art, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Example 1

[0018] See Figure 1 Embodiment 1 of the present invention provides a system framework diagram for network localization using neural networks and sorting algorithms.

[0019] To facilitate understanding, let's first describe the overall structure of this system:

[0020] Figure 1This is a system structure diagram of the present invention. The system consists of a server and a mobile terminal. The server is responsible for storing offline feature fingerprint information and calculating positioning results, and feeding the positioning results back to the user's mobile terminal. The server includes an offline computing layer and an online computing layer. By inputting the original features of wireless access points, such as W WIFI access points and C base stations, into the offline feature compression model in the offline computing layer, the server calculates the offline compressed features of the wireless access points and stores the offline compressed features of all wireless access points in a pre-set fingerprint feature library in the offline computing layer. The online computing layer includes a coarse ranking model and a fine ranking model. When the server receives a positioning request containing access point fingerprint information sent by the mobile terminal, it obtains the real-time fingerprint features of the wireless access points based on the fingerprint information. The server organizes the offline fingerprint features and real-time fingerprint features of wireless access points covering the same geographic grid together as the fingerprint feature data of the geographic grid. This fingerprint feature data is input into the coarse ranking model to obtain a preset number of candidate geographic grids, and then the preset number of candidate geographic grids are input into the fine ranking model to finally determine the geographic grid where the mobile terminal is located. A mobile terminal consists of a display module, a processor module, a communication module, and a signal acquisition module. The display module shows the location results to the user. The processor module controls wireless signal acquisition and communication with the server. The communication module communicates with the server and the wireless access point. The signal acquisition module collects relevant information from the wireless access point. Mobile terminal devices are typically carried by users who require location services. When a user needs location services, they send a location request to the server using the mobile terminal, and subsequently receive the location estimation information returned by the server and display it to the user in a visual manner. Currently popular smartphones, tablets, in-vehicle terminals, or personal digital assistants can all serve as mobile terminals.

[0021] Example 2

[0022] See Figure 2 Embodiment 2 of the present invention provides a method for network localization using neural networks and ranking algorithms, comprising the following steps:

[0023] Step S10: Based on the fingerprint information of at least one wireless access point carried in the location request of the terminal device, obtain at least one geographical grid covered by the wireless access point and the offline fingerprint features corresponding to the geographical grid from the preset fingerprint feature database.

[0024] Before this step, the original features of the wireless access points need to be input into the offline feature compression layer model to calculate the offline compressed features of the wireless access points, and the offline compressed features of all wireless access points need to be stored in a pre-set fingerprint feature library.

[0025] Wireless access points include access points and base stations. The raw characteristics of an access point include at least one of the following: number of collection points, access point type, number of grids covered by the collection point, the percentage of page views (PV) in each grid relative to the total PV of the access point, the percentage of PV in each grid relative to the total number of collection points in the current grid, and the RSSI (Radio Signal Strength Index) distribution vector. The raw characteristics of a base station include at least one of the following: number of collection points, base station type, number of grids covered by the collection point, the percentage of PV in each grid relative to the total PV of the base station, the percentage of PV in each grid relative to the total number of collection points in the current grid, and the RSSI distribution vector. The acquisition of the raw characteristics of access points and base stations does not need to be performed on mobile devices; the server can calculate and store them offline. The grid division method can be a geographical division method such as geohash or Mercator projection, as shown in Table 1.

[0026] Grid 1 AP 1 AP 2 .. BS 1 BS 2 BS 3 Grid 2 AP 3 AP 4 .. BS 4 BS 5 BS 6 … Grid N AP N AP N+1 .. BS N BS N+1 BS N+2

[0027] Table 1

[0028] The offline feature compression layer model employs a deep neural network. During computation, the original features within each grid are first extracted, and then the offline compressed features of the wireless access point within a specific grid are obtained through the offline feature compression layer model. If any grid within the coverage area of ​​the wireless access point does not have offline fingerprint features stored in the pre-defined fingerprint feature database, the default value is used as the original feature for that grid, and the default offline compressed feature for that grid is also obtained through the offline feature compression layer model. The final offline compressed feature format is shown in Table 2.

[0029]

[0030]

[0031] Table 2

[0032] Step S20: Obtain the real-time fingerprint features of the wireless access point based on the fingerprint information;

[0033] In this step, the server receives a location request from the mobile terminal. The location request includes a list of wireless access points, IP information, and contextual information about the user's location. The list of wireless access points includes: the MAC address of the access point, the key of the base station, the name of the access point and the base station, and the signal strength of the access point and the base station. Based on the above fingerprint information, the real-time fingerprint characteristics of the wireless access points are obtained.

[0034] Step S30: Organize the offline and real-time fingerprint features of wireless access points covering the same geographic grid together as the fingerprint feature data of the geographic grid.

[0035] Step S40: Input the fingerprint feature data of all geographic grids into the first sorting model, score and sort the geographic grids, and select a preset number of geographic grids as candidate geographic grids from the sorted geographic grids; input the fingerprint feature data of the candidate geographic grids into the second sorting model, score and sort the candidate geographic grids, and select the geographic grid where the terminal device is located from the sorted candidate geographic grids.

[0036] In this step, both the first and second ranking models are neural network models, with the second ranking model having more neural network layers than the first. The output of the first ranking model is the score of each candidate grid. Ranking is done based on these scores, completing the coarse ranking. The coarse screening (or preliminary selection) is the process of selecting a subset of grids from all candidates based on the coarse ranking results. These selected grids are then scored by the second ranking model. Finally, during positioning, the optimal grid is selected based on the score calculated by the second ranking model. To further improve positioning accuracy, after obtaining the optimal grid, precise location coordinates can be calculated using clustering algorithms, weighted average centroid algorithms, or Fermat point algorithms.

[0037] Example 3

[0038] See Figure 3 Embodiment 3 of the present invention discloses a network positioning device, characterized in that the device comprises: an offline feature calculation module, a real-time feature calculation module, a first sorting module, and a second sorting module; wherein,

[0039] The offline feature calculation module obtains at least one geographic grid covered by the wireless access point and the corresponding offline fingerprint features from a preset fingerprint feature library based on the fingerprint information of at least one wireless access point carried by the terminal device's location request.

[0040] The real-time feature calculation module obtains the real-time fingerprint features of the wireless access point based on the fingerprint information;

[0041] The offline and real-time fingerprint features of wireless access points covering the same geographic grid are organized together as the fingerprint feature data of the geographic grid.

[0042] The fingerprint feature data of all geographic grids are input into the first sorting module to score and sort the geographic grids, and a preset number of geographic grids are selected from the sorted geographic grids as candidate geographic grids.

[0043] The fingerprint feature data of the candidate geographic grids are input into the second sorting module to score and sort the candidate geographic grids, and the geographic grid where the terminal device is located is selected from the sorted candidate geographic grids.

[0044] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the above-described apparatus and system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0045] In addition, embodiments of the present invention also disclose a server, which includes a storage device and one or more processors. The storage device is used to store one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement the method as described in Embodiment 1.

[0046] This invention also discloses a computer-readable storage medium storing a computer program thereon, which, when executed, implements the method as described in Embodiment 1.

[0047] The technical solution disclosed in this invention collects raw information that considers both indoor access points and outdoor communication base stations, expanding the applicability of positioning from indoor to integrated indoor and outdoor environments. Compared to existing technologies that rely on signal strength parameters that are susceptible to environmental changes to train models, this invention introduces features that are less affected by environmental changes, such as PV ratio and access point type, when training the model. Multiple pieces of information are fused and refined using the offline feature compression layer of a deep neural network model, improving the robustness of the positioning system. Furthermore, this invention introduces an online network on the basis of offline training models, fusing offline and online features and using the LTR ranking algorithm for sorting. Through the concatenation of coarse and fine ranking, positioning accuracy is improved while reducing computational load.

[0048] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods, apparatus, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowcharts and block diagrams may represent a module, program segment, or portion of code, containing one or more computer-executable instructions for implementing logical functions. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. Furthermore, it should be noted that each block or combination of blocks in the block diagrams and flowcharts may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0049] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0050] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. The embodiments described above are merely illustrative examples and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description, and such variations or modifications are obvious and fall within the scope of protection of the present invention.

Claims

1. A method of network positioning, characterized by, The method comprises the following steps: According to the fingerprint information of at least one wireless access point carried by the positioning request of the terminal device, at least one geographic grid covered by the wireless access point and the offline fingerprint feature corresponding to the geographic grid are obtained from the preset fingerprint feature library; wherein the fingerprint information comprises at least one of the following: wireless access point list, IP information, context information of user positioning; According to the fingerprint information, the real-time fingerprint feature of the wireless access point is obtained; The offline fingerprint feature and the real-time fingerprint feature of the wireless access point covering the same geographic grid are organized together as the fingerprint feature data of the geographic grid; The fingerprint feature data of all geographic grids is input into a first sorting model to score and sort the geographic grids, and a preset number of geographic grids are selected from the sorted geographic grids as candidate geographic grids; The fingerprint feature data of the candidate geographic grid is input into a second sorting model to score and sort the candidate geographic grid, and the geographic grid where the terminal device is located is selected from the sorted candidate geographic grid; The first sorting model and the second sorting model are neural network models, and the number of neural network layers of the second sorting model is more than that of the first sorting model.

2. The method of claim 1, wherein: The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library.

3. The method of claim 2, wherein: The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library.

4. The method of claim 3, wherein: The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library.

5. The method of claim 4, wherein: The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library.

6. The method of claim 5, wherein: The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library.

7. The method of claim 3, wherein: The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library.

8. The method of claim 1, wherein: The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library.

9. The method of claim 1, wherein: The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is stored in an offline feature fingerprint library. The offline fingerprint feature of the wireless access point is obtained through an offline feature compression layer model, and the offline fingerprint feature of the wireless access point is 10. An apparatus for network positioning, the apparatus comprising: The device comprises an offline feature calculation module, a real-time feature calculation module, a fingerprint feature data acquisition module, a first sorting module, and a second sorting module. The offline feature calculation module is configured to acquire at least one geographic grid covered by a wireless access point and offline fingerprint features corresponding to the geographic grid from a preset fingerprint feature library according to fingerprint information of the wireless access point carried by a positioning request of a terminal device, wherein the fingerprint information comprises at least one of the following: a wireless access point list, IP information, and context information of user positioning. The real-time feature calculation module is configured to acquire real-time fingerprint features of the wireless access point according to the fingerprint information. The fingerprint feature data acquisition module is configured to organize offline fingerprint features and real-time fingerprint features of wireless access points covering the same geographic grid together as fingerprint feature data of the geographic grid. The first sorting module is configured to input fingerprint feature data of all geographic grids into a first sorting model, score and sort the geographic grids, and select a preset number of geographic grids from the sorted geographic grids as candidate geographic grids. The second sorting module is configured to input fingerprint feature data of the candidate geographic grids into a second sorting model, score and sort the candidate geographic grids, and select a geographic grid where the terminal device is located from the sorted candidate geographic grids. The first sorting model and the second sorting model are neural network models, and the neural network layers of the second sorting model are more than those of the first sorting model.

11. A server, characterized by The server comprises: a storage device; one or more processors; The storage device is configured to store one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method of network positioning according to any one of claims 1-9.

12. A computer readable storage medium having stored thereon a computer program which, when executed, implements the method of network positioning according to any one of claims 1-9.

Images