A cell image information determination method, device and electronic equipment

Abstract

Embodiments of the present application disclose a cell portrait information determination method and device and electronic equipment, which are used for more effectively portrait a wireless cell. The method comprises: performing grid division on a three-dimensional electronic map containing a plurality of cells; determining three-dimensional grids in which users in the plurality of cells are located, and processing network performance indexes of the plurality of users in each three-dimensional grid to obtain network performance indexes of the three-dimensional grids; constructing a multi-relationship network model based on at least one of a common coverage relationship and a configuration handover relationship of the plurality of cells; determining Pagerank contribution degrees of the three-dimensional grids covered by the plurality of cells to a network performance index of a certain cell based on the network performance indexes of the three-dimensional grids covered by the plurality of cells and the multi-relationship network model; and aggregating the network performance indexes of the three-dimensional grids covered by the plurality of cells according to the Pagerank contribution degrees to obtain portrait information of a first cell.

Description

Technical Field

[0001] This application relates to the field of community profiling technology, and in particular to a method, apparatus and electronic device for determining community profiling information. Background Technology

[0002] With the development of big data, profiling technology has been widely applied in various industries. Profiling technology mainly refers to the abstract processing of relevant data of a target object, and by labeling the target object, its characteristics are highly summarized to facilitate human understanding and computer processing.

[0003] Currently, most methods for judging network quality and evaluating the merits of a cell rely on a single or a few indicators that are statistically analyzed as a whole to create a profile of the cell. This often presents a significant barrier for operations and maintenance personnel to gain an intuitive understanding of the cell's overall situation.

[0004] It is evident that there is currently a lack of effective methods for profiling communities, making it impossible to provide effective guidance for network operation and maintenance. Summary of the Invention

[0005] This application provides a method, apparatus, and electronic device for determining cell profile information, which can be used to create more effective cell profiles to guide network operation and maintenance more effectively.

[0006] The embodiments of this application adopt the following technical solutions:

[0007] Firstly, a method for determining community profile information is provided, applied to the server side of an information storage system, the method comprising:

[0008] A three-dimensional electronic map containing multiple cells is divided into grids to obtain several three-dimensional grids, where one cell covers multiple three-dimensional grids;

[0009] The three-dimensional grid where each user is located in the plurality of cells is determined, and the network performance indicators of the plurality of users in the first grid are processed to obtain the network performance indicators of the first grid, wherein the first grid is any grid in the three-dimensional grid covered by the plurality of cells.

[0010] A multi-relationship network model is constructed based on at least one of the shared coverage relationship and configuration handover relationship of the multiple cells. In the multi-relationship network model, a node represents a cell, and there is an edge between nodes with shared coverage relationship.

[0011] Based on the network performance metrics of the three-dimensional grids covered by the multiple cells and the multi-relationship network model, the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance metrics of the first cell is determined, wherein the first cell is any one of the multiple cells.

[0012] Based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, the network performance indicators of the three-dimensional grids covered by the multiple cells are aggregated to obtain the profile information of the first cell.

[0013] Secondly, a device for determining community profile information is provided, the device comprising:

[0014] The 3D raster division module is used to divide a 3D electronic map containing multiple cells into raster grids, resulting in several 3D rasters, where one cell covers multiple 3D rasters.

[0015] The network performance index backfilling module is used to determine the three-dimensional grid where each user is located in the multiple cells, and process the network performance index of the multiple users in the first grid to obtain the network performance index of the first grid, wherein the first grid is any grid in the three-dimensional grid covered by the multiple cells.

[0016] The network model construction module is used to construct a multi-relationship network model based on at least one of the co-coverage relationship and configuration handover relationship of the multiple cells, wherein in the multi-relationship network model, a node represents a cell, and there is an edge between nodes with co-coverage relationship;

[0017] The contribution determination module is used to determine the Pagerank contribution of the three-dimensional grid covered by the multiple cells to the network performance index of the first cell based on the network performance index of the three-dimensional grid covered by the multiple cells and the multi-relationship network model, wherein the first cell is any one of the multiple cells.

[0018] The profile information determination module is used to aggregate the network performance indicators of the three-dimensional grids covered by the multiple cells based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, so as to obtain the profile information of the first cell.

[0019] Thirdly, an electronic device is provided, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, performs the steps of the method described in the first aspect.

[0020] Fourthly, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the steps of the method described in the first aspect.

[0021] Fifthly, a computer program product is provided, which is stored in a storage medium and is executed by at least one processor to implement the method as described in the first aspect.

[0022] The cell profile information determination method proposed in this application has two advantages. First, it aggregates user network performance indicators at the granularity of a three-dimensional grid, rather than directly fitting single or multiple network performance indicators at the cell level. Second, when determining the cell profile information of a cell, it considers the network topology relationships between cells, rather than creating a profile solely based on the cell's network performance indicators. Therefore, it can provide more intuitive, detailed, and effective cell profile information, which can better guide the network maintenance of the cell. Attached Figure Description

[0023] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0024] Figure 1 This is a flowchart illustrating a method for determining community profile information according to an embodiment of this application.

[0025] Figure 2 This is a schematic diagram of a multi-relationship network model provided in an embodiment of this application.

[0026] Figure 3 This is a schematic diagram of the logic flow of a method for determining community profile information provided in an embodiment of this application.

[0027] Figure 4 This is a schematic diagram of the structure of a community profile information determination device provided in another embodiment of this application.

[0028] Figure 5 This is a schematic diagram of the hardware structure of an electronic device provided in one embodiment of this application. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0030] In related technologies, most network operation and maintenance monitoring and fault identification methods rely heavily on network statistical indicators. These indicators, configured at the cell level, are used to assess network conditions, supporting fault alarms and solution optimization. However, the triggering mechanisms for daily network monitoring and optimization are often the degradation of specific indicators in the real network or the need to optimize a particular indicator. The output of previous solutions and the judgment of site quality often depend on statistical analysis of network-side indicators and manual methods. Most methods for judging network quality and evaluating cell quality rely on fitting single or several indicators from overall cell statistics. This often presents a significant barrier for operations and maintenance personnel to intuitively understand the cell overview. There is no effective means to evaluate the current state of the network.

[0031] Furthermore, many network scoring mechanisms in related technologies focus on statistically analyzing overall network performance metrics. They employ machine learning techniques for expert-based scoring and model training, leading to significant ambiguity in network evaluations for the same cell from different perspectives. This results in varying fault feedback and root causes depending on the monitoring and maintenance methods employed. During optimization, inter-cell coverage and neighbor cell relationships can influence parameter adjustments, lacking analysis of metric fluctuations and user perception caused by the cell's inherent characteristics.

[0032] Furthermore, in related technologies, the application of scoring methods is concentrated on one-to-one problem analysis of data within a certain time period. This leads to low efficiency in identifying corresponding faults and problems, and makes it difficult to reuse current solutions when similar problems occur. At the same time, it is difficult to link corresponding indicators and trace the problems when frequent faults occur in the future. The various types of indicators on the network side and the judgment of signaling also have complex structural relationships, which cause great interference and impact on problem identification.

[0033] To address at least one of the aforementioned security issues, embodiments of this application provide a method and apparatus for determining cell profile information. This method can be executed by an electronic device, such as a terminal device or a server, or it can be executed by software installed in an electronic device.

[0034] This application provides a method for determining cell profile information, aiming to build a cell grid profiling system by combining user perception and network performance, establish a personalized cell grid model, and construct a target management model for grid profiling and cell indicators based on communication scheduling principles and signaling processes by fitting the converged results of the grid profiles. Furthermore, based on network performance indicator backfilling, an integrated parameter optimization model is constructed, providing more intuitive and comprehensive support for network detection and daily optimization by operations and maintenance personnel. Compared to related technologies, the improvements generally include, but are not limited to, at least one of the following:

[0035] 1) Construct a profile model with a three-dimensional grid as the granularity, link user measurement indicators and network statistical indicators to the network model, introduce a multi-objective genetic optimization iterative method to obtain a three-dimensional grid profile model to support the identification and discovery of cell faults.

[0036] 2) Construct a network topology structure from at least one of the two perspectives, including but not limited to shared network coverage neighbor cells and configuration handover neighbor cells, to obtain a multi-relationship network model. Based on this multi-relationship network model, analyze the neighbor cell mapping between cells and the cell profile model.

[0037] 3) A comprehensive assessment of multiple characteristics of nodes-cells, such as user mobility, network indicator differences, and channel indicator tidal changes, is conducted. A PageRank contribution algorithm with a three-dimensional grid as the granularity is built to form a cell profile model. Then, by adjusting the parameters at the grid granularity, the aggregation results of the cells are changed, thereby supporting the output of a cell fault solution based on grid granularity parameter optimization.

[0038] Based on the channel propagation model, this application constructs a three-dimensional grid-level profile of the network coverage sector, combining the dimensions of "network and people". It creates an integrated model for monitoring, identification and optimization of base station sites, site selection, channel environment and network configuration. It uses the overall scoring effect to locate and delimit network problems more intuitively, and maximizes the comprehensiveness and completeness of fault problem and optimization direction identification.

[0039] The following describes a method for determining community profile information proposed in an embodiment of this application.

[0040] like Figure 1 As shown in the embodiment of this application, a method for determining cell profile information may include:

[0041] Step 101: Divide the three-dimensional electronic map containing multiple cells into grids to obtain several three-dimensional grids, wherein one cell covers multiple three-dimensional grids.

[0042] The RAN#69 meeting approved the "Above 6GHz Channel Model Study," which defines a modeling methodology typically applicable to the 0.5-100GHz range. The channel model is suitable for system-level simulations, supporting scenarios including urban micro-cell streets, urban macrocells, indoor offices, rural macrocells, and indoor environments. The corresponding global channel model is created based on deterministic ray tracing on electronic maps (or digital maps) and simulation of certain random components.

[0043] Specifically, step 101 may include: setting up the computing environment and importing a three-dimensional electronic map containing multiple cells that need to be monitored; dividing the three-dimensional electronic map into grids to obtain several three-dimensional grids.

[0044] The 3D electronic map contains the following information: 3D geometric information of each major structure of the building or room; the material and thickness of each wall and its corresponding electromagnetic properties, including dielectric constant and conductivity; random obstacles in certain scenes, etc. In this 3D electronic map, the exterior and interior walls of the building are represented by surfaces and identified by the coordinates of the vertices on each wall.

[0045] Typically, the resulting 3D grid is a cuboid or cube, but other shapes are also possible. For example, on the 3D electronic map, a 5m x 5m grid can be used horizontally, with each vertical grid also based on a 5m base grid. After the 3D grid is divided, each grid is treated as a user terminal (UT), and the base station (BS) identifier and UT number corresponding to the three grids are assigned. The latitude and longitude information of the center position of each 3D grid is used as the coordinates of the current 3D grid, thus completing the grid division within the area.

[0046] Step 102: Determine the three-dimensional grid where each user is located in the multiple cells, and process the network performance indicators of the multiple users in the first grid to obtain the network performance indicators of the first grid, wherein the first grid is any grid in the three-dimensional grid covered by the multiple cells.

[0047] In step 102 above, the purpose of determining the 3D grid where each user is located within the multiple cells is to backfill (or map) the user's network performance indicators to the corresponding 3D grid. There are many ways to determine the 3D grid where each user is located within the multiple cells; three are described below.

[0048] In a first embodiment, determining the three-dimensional grid where each user is located within the plurality of cells may include: determining the three-dimensional grid where each user is located within the plurality of cells based on the location information of each user within the plurality of cells, the size information of a single three-dimensional grid, and the coordinate information of the three-dimensional grids covered by the plurality of cells. It can be understood that if the location coordinates of a user fall exactly within a certain three-dimensional grid, then the user is determined to be within that three-dimensional grid.

[0049] In the second embodiment, determining the three-dimensional grid where each user is located within the plurality of cells may include: obtaining the predicted channel quality information of the three-dimensional grids covered by the plurality of cells within a measurement window through simulation; and determining the three-dimensional grid where the user is located within the plurality of cells based on the predicted channel quality information of the three-dimensional grids covered by the plurality of cells within the measurement window and the actual channel quality information of the users within the plurality of cells within the measurement window.

[0050] In practice, if a user's actual channel quality information within the measurement window is consistent with or close to the predicted channel quality information of a certain three-dimensional grid within the measurement window (the difference is less than a preset threshold), then the user is determined to be within that three-dimensional grid.

[0051] In the third implementation, determining the three-dimensional grid where each user is located within the plurality of cells may include: obtaining the predicted channel quality information of the three-dimensional grids covered by the plurality of cells within a measurement window through simulation; determining the three-dimensional grid where the user is located within the plurality of cells based on the location information of each user within the plurality of cells, the size information of a single three-dimensional grid, the coordinate information of the three-dimensional grids covered by the plurality of cells, the predicted channel quality information of the three-dimensional grids covered by the plurality of cells within the measurement window, and the actual channel quality information of the user within the plurality of cells within the measurement window. It can be seen that the third implementation is a combination of the first and second implementations. Specifically, if a user's location coordinates fall within a certain three-dimensional grid, and the user's actual channel quality information within the measurement window is consistent with or close to the predicted channel quality information of the three-dimensional grid within the measurement window (the difference is less than a preset threshold), then the user is determined to be within that three-dimensional grid.

[0052] In the second and third embodiments described above, a user's actual channel quality information within the measurement window comes from the user's measurement report (MR) within the measurement window. As an example, the channel quality information may include at least one of Reference Signal Received Power (RSRP) and [other values].

[0053] The process of obtaining the RSRP of a single 3D grid within the measurement window through simulation is described below.

[0054] First, the 3D positions of the Base Station (BS) and the Under Station (UT) are given, and the LOS AOD (φ) of each BS and UT is calculated in the global coordinate system. LOS,AOD LOS ZOD(θ) LOS,ZOD ), LOS AOA(φ LOS,AOA ), LOS ZOA(θ) LOS,ZOAThe LOS is an abbreviation for line-of-sight (stability limit). It also provides information on the BS radiation area and the UT's orientation relative to the BS, such as azimuth and downtilt, relative to the BS, and the orientation of the BS and UT arrays relative to the Global Coordinate System (GCS). Furthermore, it provides BS operating parameters such as the system center frequency and bandwidth for each BS-UT link. These BS operating parameters may include neighbor cell relationships, broadcast channel power, and access RSRP threshold.

[0055] Secondly, based on virtual ray tracing, path analysis is performed to determine the reachable path from the base station to the three-dimensional grid, thereby conducting link simulation to simulate the end-to-end propagation relationship between a pair of Tx / Rx arrays. Geometric calculations are performed in virtual ray tracing to determine the propagation interaction type of each propagation path, including LOS, reflection, diffraction, penetration, and scattering. Attenuation calculations are then performed on the propagation path based on the determined propagation interaction type (LOS, reflection, diffraction, penetration, and scattering) and f.

[0056] Finally, power calculations are performed on small-scale parameters such as arrival delay for each virtual ray to obtain the corresponding power of all rays within the group (the combination and convergence of multiple virtual rays emitted by the base station arriving at a three-dimensional grid). The arrival angle, azimuth angle, and other information of the virtual rays are simulated to generate corresponding channel coefficients, thus completing the overall channel power assessment. Non-direct-fire virtual rays often require calculations of fading such as multipath interference. Different types of algorithms can be used to optimize the delay assessment and power simulation, which are also key indicators of the overall simulation. Corresponding model calibration and training are performed based on the actual network test data to simulate the delay loss of the real network environment to the greatest extent possible, and to simulate the RSRP of all users within the three-dimensional grid.

[0057] Furthermore, when determining the 3D grid where a user is located, the user's access site information can also be considered. As a concrete example, after completing the 3D grid division based on a 3D electronic map within the area, each user within the area can be accurately located. Combining the collected user measurement data, access site information (such as the access cell ID), and RSRP simulation results of the 3D grid, the measurement data of the corresponding user can be backfilled into the corresponding 3D grid.

[0058] RSRP and SINR can be used to provide feedback on the channel quality of the corresponding grid users; the Synchronization Signal / PBCH Block Index (SSBindex) can be used to determine the user's Sounding Reference Signal (SRS) angle and null RXX matrix; and the Uplink Path Loss (PathLoss) or Downlink Path Loss (PUL / PDL) can be used to determine whether the corresponding user is currently engaging in service activities that consume network resources, thereby completing the backfilling of network performance indicators for grid-level users.

[0059] By combining the RSRP values ​​and SSBindex of the corresponding user measurements in the above figure with AOA and TA, the measurement data at the corresponding measurement time is mapped to the corresponding grid, thereby obtaining the channel quality status of the corresponding user at the current time. This yields RSRP and SINR information that appeared at different times with time stamps in the corresponding grid, as well as the SSBindex information obtained by the user measurements within the grid. Simultaneously, the corresponding user ID and the demodulated DCI1_1 field at the corresponding time are obtained, along with the start and end RB sequences occupied by the user, the PDSCH start symbol l0, and the occupied symbol length under the corresponding measurement data at the corresponding time. Based on the above information, the multi-element array information of the corresponding user is obtained, including: grid ID at the measurement time, access cell CGI information, measurement time, access beam information, RSRP, RSRQ, SINR, whether RB resources are scheduled, occupied resource block (RB) and symbol size, etc., thus completing the MR backfill within the phased measurement time window and completing the preliminary grid simulation preparation work.

[0060] In step 102, there are many ways to process the network performance indicators of multiple users within the first grid to obtain the network performance indicators of the first grid. Two methods are described below.

[0061] In some embodiments, the average network performance index of all users within the first grid can be used as the network performance index of the first grid.

[0062] In other embodiments, considering that different network performance indicators of different users within the first grid have different effects on the network performance indicator of the first grid, the network performance indicators of multiple users within the first grid can be processed based on a multi-objective optimization algorithm to obtain the network performance indicator of the first grid. Here, the first grid is any grid in the three-dimensional grid covered by the multiple cells, and the user parameters of the multiple users come from the MR of the multiple users.

[0063] A regression model for target management is built based on network performance metrics and user signaling and data plane data. Regression analysis and fitting are performed on user measurement data service classification, network statistical load, and interference metrics for different networks, frequency bands, and power levels to establish corresponding target management models. After obtaining relevant user measurement information for all registered users in the current area, the network performance metrics of the three-dimensional grid at the current moment are obtained by fitting the evaluation at the user grid granularity. These user network performance metrics may include, but are not limited to, user dwell time, user scheduling time, average RB occupancy, and probability of handover to other grids. Correspondingly, the cell network performance metrics include, but are not limited to, load metrics (PRB utilization, CCE utilization, RRC link count) and interference metrics (call success rate, handover rate, CQI good rate, RSRP poor rate), etc. This proposal emphasizes the introduction of a multi-objective genetic optimization learning model to determine the regression model for target management by aggregating user metrics at the grid granularity using user-dimensional measurement data.

[0064] Specifically, the users residing in the community and the user groups activating resources are used as the population. At the same time, the relationship between the community and the users is based on the latitude and longitude distance and azimuth of the connecting line. Given that the performance relationship between the user's location and network registration has been determined, this paper focuses on adopting a multi-objective management learning approach, improving the process of obtaining the best selection from the population each time, and adding a population optimization algorithm (Particle Swarm Optimization, PSO) to perform crossover and mutation from users, thereby completing the calculation of the fitness function and realizing the objective genetic optimization learning model.

[0065] More specifically, before optimizing using a Genetic Algorithm (GA), a fitness function is first defined. Then, considering the target optimization requirements of multiple performance indicators in the current cell, a fitness function is established for each user i:

[0066]

[0067] If we determine all users belonging to the current cell as the entire population, then the probability of each user i being selected is:

[0068] Where i = 1, 2, ... N

[0069] Among them, f i This represents the fitness of user i.

[0070] Given the iteration number t, and two selected users A and B undergoing crossover and mutation operations, let the probabilities of crossover and mutation be Pc and Pm, respectively. Where, for any two individuals... and After crossover and mutation, they become respectively and

[0071]

[0072] Where λ is a random value in the range [0,1].

[0073] Thus, the mean fitness and crossover mutation probability fitness of all users in the community are obtained as follows:

[0074]

[0075] Based on the improved probabilistic optimization iterative objective of PSO, the algorithm combines and integrates the methods of searching for users in each step. The algorithm's methods for user search and compensation are as follows:

[0076] v i (k+1)=w×v i (k)+c1r1(P besti -x i (k))+c2r2(g besti -x i (k))

[0077] x i (k+1)=v i (k+1)+x i (k)

[0078] Where i = 1, 2, ... N is the number of users, and N represents the population size; v i (k) represents the velocity of user i in the k-th dimension; x i (k) represents the position of user i in the k-th dimension; w is the inertia weight; constants c1 and c2 are acceleration coefficients, where C1 represents the user's ability to learn from itself, and C2 represents the user's ability to learn from the entire population; r1 and r2 are random values ​​uniformly distributed in [0,1]; P besti Let g represent the optimal position experienced by user i, and the sum of the terms it belongs to represents the mechanism for learning from itself; besti This represents the position of the best user in the population, and the sum of the terms indicates the mechanism by which the user learns from the entire population.

[0079] By iterating through crossover and mutation multiple times, where 'best' represents the optimal position, the result with the highest adaptability is output. This completes the construction of the fitness function for user and cell metrics.

[0080] Step 103: Construct a multi-relationship network model based on at least one of the co-coverage relationship and configuration handover relationship of the multiple cells, wherein in the multi-relationship network model, a node represents a cell, and there is an edge between nodes with co-coverage relationship.

[0081] Traditional simulations of wireless networks often struggle to accurately represent the mapping relationships between different neighboring cells and different coverage layers. Furthermore, the capacity and load of network nodes are often difficult to express with a linear relationship. This application focuses on studying the interdependent topological relationships of networks through the cascading effect phenomenon of multi-layer coupled networks.

[0082] First, each cell is considered a node in the entire network. Shared coverage relationships are defined as edges between nodes. Shared coverage is defined as the overlapping coverage areas within the coverage radii of two cells. Indicators affecting shared coverage relationships include cell latitude and longitude, and antenna height information, which determine a three-dimensional coordinate system (lat, lo, high) for each cell. The coverage radius L of different cells can be determined using the cell's rated power and reference signal transmission power. The starting angle α and arrival angle β of the cell's coverage sector can be determined using the cell's azimuth angle. This determines the edges between nodes. In this embodiment, the neighbor cell configuration parameters affecting the cell profile represent the relationships on these edges. Ultimately, the edges between any two cells in the multi-layer network are expressed as combinations of already configured neighbor cells between cells, rather than simply combinations of any two adjacent cells. This improves the simulation of neighbor cell networks capable of user migration. Therefore, a multi-relationship network is constructed based on the multi-layer coupled network, represented by the quadruple G = (V, E, R, F).

[0083] 1) V = {v1, v2, v3, ..., vn} represents the cell set, where n is the total number of base station cells in the region.

[0084] 2) E = {<Vm,Vn> Where m and n are cells in V, representing two cells with a shared coverage grid relationship, and the full set E represents the set of all cell pairs configured in the region that have a shared coverage relationship.

[0085] 3) R = {r1, r2, ..., rx} represents the set of neighbor cell configuration switching parameters or shared parameters in V, and ri represents a relationship between the corresponding nodes;

[0086] 4) F:E→R is a mapping of edges and relations.

[0087] Figure 2A schematic diagram of the multi-relationship network model constructed in this embodiment is shown. The coefficient between node v and relationship r is set as k, then the average coefficient corresponding to different relationship networks is represented as q. The relationship strength of each relationship is expressed as the degree of influence of the relationship on the node's connection. The proportion of the relationship coefficient is set as fi. Different switching parameters and coverage parameters cause dynamic changes in the relationship coefficients brought about by the network to which the grid belongs. The forward and reverse switching directions represent the effect. The indicators and relationships of two cells within the area, as well as the overall indicators and relationships, are defined as follows:

[0088]

[0089] Where, r i ∈R, Γ is the set of neighboring nodes of node v with respect to relation r, and α is an adjustable parameter.

[0090] This completes the construction of the multi-relationship network model. After the multi-relationship network model is completed, the grid aggregation set is first used as the cell nodes. Then, the indicator traffic L and relationship ri of each cell node are assigned a reinforcement distribution learning association relationship, thereby completing the construction of the overall cell profile model.

[0091] Step 104: Based on the network performance indicators of the three-dimensional grids covered by the multiple cells and the multi-relationship network model, determine the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, wherein the first cell is any one of the multiple cells.

[0092] In some embodiments, step 104 may specifically include: determining the neighboring cells of the first cell based on the multi-relationship network model; and determining the Pagerank contribution of the three-dimensional grid covered by the first cell and its neighboring cells to the network performance indicators of the first cell based on Pagerank.

[0093] By tracing back the contribution of users in different scenarios at different times within a 3D grid to the cell to which the 3D grid belongs, the contribution of the 3D grid to the network performance metrics of the cell and the impact on neighboring cells can be effectively obtained, which is divided into two cases of positive and reverse relationships. Each 3D grid has multiple metrics, including: basic longitude and latitude feature information, user residence duration, user scheduling duration, and average metrics of users occupying RBs, etc. The cell also includes statistical metrics: PRB utilization rate, traffic, and handover rate, etc. The metrics related to the neighboring cells are those related to the handover rate and interference rate. It can be seen that the network performance metrics of the 3D grid can be statistically extended to the network performance metrics of the entire cell. Among them, the PRB utilization rate is the ratio of the occupied RBs to the frequency band, the handover rate is the ratio of the change in the accessed site CGI when the user does not move at the 3D grid position, and the disconnection rate is the ratio after the user remains resident in the 3D grid and reconnects to the 3D grid. Taking the current feature as the basic feature of the vector, it can be statistically obtained that the cell covers a total of n grids, and among them, m grids overlap with k neighboring cells, which will cause partial interference.

[0094] The general definition of PageRank is as follows: Given an arbitrary directed graph with n nodes, a general random walk model is defined on the directed graph, that is, a first-order Markov chain. The transition matrix of this model is composed of the linear combination of two parts: one part is the basic transition matrix M of the directed graph, indicating that the transition probability from one node to all the outgoing nodes is equal; the other part is a completely random transition matrix, indicating that the transition probability from any node to other nodes is 1 / n, and the linear combination coefficient is the damping factor d (0 < d < 1). This general random walk Markov chain has a stationary distribution, denoted as R. Defining the stationary distribution vector R as the general PageRank of the directed graph is as follows:

[0095]

[0096] Among them, 1 is a vector with n dimensions and 1 column, and each component has a size of 1, and t is the number of iterations of the algorithm.

[0097] Furthermore, adding the BERT multi-head attention redness as the overall pooling layer, the feature identifier of the unit node is X, α is the attention weight, taking the function sim as the evaluation of measuring the input similarity, after calculating the similarity of every two 3D grids, the similarity ratio is statistically obtained, where, Expressing the similarity between every two 3D grids, for the total N 3D grids, replacing the transition matrix in the pagerank first-order chain with W (that is, replacing the above matrix M with W), the deformation formula of the vector R can be obtained as:

[0098]

[0099] After performing PageRank iterations until the difference in PageRank values ​​is less than the preset conclusion, the contribution coefficients of each 3D grid to the cell features are calculated. For results that tend to stabilize, the independent weight coefficients of the current 3D grid to the main cell and neighboring cells are fitted. Based on the correlation relationships of all grids in the multi-layer coupled network, contribution evaluation indicators are established respectively, thus obtaining:

[0100] The network performance index of the first cell = ∑ contribution of the grid covered by the first cell * network performance index of the grid covered by the first cell + ∑ contribution of the grid covered by neighboring cells * network performance index of the grid covered by neighboring cells.

[0101] Step 105: Based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, the network performance indicators of the three-dimensional grids covered by the multiple cells are aggregated to obtain the profile information of the first cell.

[0102] It is understandable that the profile information of the first community obtained through step 105 is actually a profile model.

[0103] In some embodiments, step 105 may specifically include: based on the Pagerank contribution of the three-dimensional grids covered by the plurality of cells to the network performance indicators of the first cell, aggregating the network performance indicators of the three-dimensional grids covered by the plurality of cells to obtain the profile information of the first cell.

[0104] In other embodiments, Figure 1 The method further includes: determining the edge coefficients of each edge in the multi-relationship network based on the mapping relationship between edges and relations and the relation strength. Accordingly, step 105 may specifically include: based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, aggregating the network performance indicators of the three-dimensional grids covered by the multiple cells to obtain the first profile information of the first cell; and determining the second profile information of the first cell based on the first profile information and the edge coefficients of the first cell in the multi-relationship network model.

[0105] Specifically, the second profile information of the first cell can be obtained by multiplying the first profile information by the edge coefficient of the first cell in the multi-relationship network model.

[0106] It's understandable that, based on the cell's coverage radius and the corresponding grid situation, a profile model of the current cell is obtained through aggregation. This profile monitoring model is then completed through multi-dimensional indicator evaluation. Adjusting the user's grid position affects the performance indicators of the cell to which the grid belongs. A complete link function is used to construct a full-link indicator system fitting encompassing user behavior, user measurement, grid access, grid coverage, cell indicators, neighboring cell network, and cell profile. Therefore, each time cell parameters are adjusted to assess user behavior, changes in cell indicators can be obtained based on the cell profile system. Simultaneously, grid-level backtracking is performed to locate indicator anomalies caused by changes in user behavior, improving operational efficiency while providing a more explicit and visual display and usability of the cell profile system.

[0107] The cell profile information determination method proposed in this application has two advantages. First, it aggregates user network performance indicators at the granularity of a three-dimensional grid, rather than directly fitting single or multiple network performance indicators at the cell level. Second, when determining the cell profile information of a cell, it considers the network topology relationships between cells, rather than creating a profile solely based on the cell's network performance indicators. Therefore, it can provide more intuitive, detailed, and effective cell profile information, which can better guide the network maintenance of the cell.

[0108] Figure 3 This illustration shows a logical flow diagram of a method for determining cell profile information according to an embodiment of this application. Figure 3 As shown in the embodiments of this application, a method for determining cell profile information includes the following: three-dimensional grid division; backfilling user measurement reports into the three-dimensional grid; optimizing the user measurement reports and network indicators backfilled into the three-dimensional grid based on a multi-objective genetic optimization algorithm to obtain the network performance indicators of the three-dimensional grid; determining the PageRank contribution of the three-dimensional grids covered by a main cell and its neighboring cells to the network performance indicators of the main cell based on the PageRank algorithm; based on the PageRank contribution of the three-dimensional grids covered by a main cell and its neighboring cells to the network performance indicators of the main cell, aggregating the network performance indicators of the three-dimensional grids covered by the main cell and its neighboring cells (multi-layer coupled composite network) to obtain the network performance indicators of the main cell—that is, the profile information of the main cell; then, iteratively optimizing the network performance indicators of the main cell based on the profile information of the main cell, the multi-layer coupled composite network, and the multi-relationship network model; and re-measuring the user's MR based on the neighboring cell relationship parameters to establish communication network link association.

[0109] according to Figure 3 It can be seen that the method for determining community profile information proposed in this application has at least one of the following technical advantages:

[0110] 1) Nodes are constructed by combining multiple types of neighboring networks with shared coverage and shared configuration, as well as natural geographical features, through topological network structure relationships;

[0111] 2) Combining machine learning algorithms provides a new direction for scoring in two dimensions: cell and grid. It combines multiple types of indicators, considering causal and logical relationships while also taking into account the confounding effects of environmental variables.

[0112] 3) Combine the PageRank algorithm to complete the convergence relationship and association model of grids and cells, and build a cell profile system for network evaluation and daily monitoring.

[0113] The above continues the description of a method for determining community profile information proposed in the embodiments of this application. Correspondingly, the embodiments of this application also propose a device for determining community profile information, which will be described below.

[0114] like Figure 4 As shown, an embodiment of this application proposes a cell profile information determination device 400, which includes: a three-dimensional grid division module 401, a network performance index backfilling module 402, a network model construction module 403, a contribution determination module 404, and a profile information determination module 405.

[0115] The three-dimensional grid division module 401 is used to divide a three-dimensional electronic map containing multiple cells into grids to obtain several three-dimensional grids, wherein one cell covers multiple three-dimensional grids.

[0116] Specifically, the three-dimensional grid division module 401 can be used to: set up the computing environment and import a three-dimensional electronic map containing multiple cells that need to be monitored; and perform grid division on the three-dimensional electronic map to obtain several three-dimensional grids.

[0117] The 3D electronic map contains the following information: 3D geometric information of each major structure of the building or room; the material and thickness of each wall and its corresponding electromagnetic properties, including dielectric constant and conductivity; random obstacles in certain scenes, etc. In this 3D electronic map, the exterior and interior walls of the building are represented by surfaces and identified by the coordinates of the vertices on each wall.

[0118] The network performance index backfilling module 402 is used to determine the three-dimensional grid where each user is located in the multiple cells, and process the network performance index of the multiple users in the first grid to obtain the network performance index of the first grid, wherein the first grid is any grid in the three-dimensional grid covered by the multiple cells.

[0119] In the network performance metric backfilling module 402, the purpose of determining the three-dimensional grid where each user is located within the multiple cells is to backfill (or map) the user's network performance metrics to the corresponding three-dimensional grid. There are many ways to determine the three-dimensional grid where each user is located within the multiple cells; three are described below.

[0120] In the first embodiment, the network performance index backfilling module 402 can be used to: determine the three-dimensional grid where each user is located in the multiple cells based on the location information of each user in the multiple cells, the size information of a single three-dimensional grid, and the coordinate information of the three-dimensional grids covered by the multiple cells. It can be understood that if the location coordinates of a user fall exactly within a certain three-dimensional grid, then the user is determined to be within that three-dimensional grid.

[0121] In the second embodiment, the network performance index backfilling module 402 can be used to: obtain the predicted channel quality information of the three-dimensional grids covered by the multiple cells within the measurement window through simulation; and determine the three-dimensional grids where the users in the multiple cells are located based on the predicted channel quality information of the three-dimensional grids covered by the multiple cells within the measurement window and the actual channel quality information of the users in the multiple cells within the measurement window.

[0122] In practice, if a user's actual channel quality information within the measurement window is consistent with or close to the predicted channel quality information of a certain three-dimensional grid within the measurement window (the difference is less than a preset threshold), then the user is determined to be within that three-dimensional grid.

[0123] In the third implementation, the network performance index backfilling module 402 can be used to: obtain the predicted channel quality information of the three-dimensional grids covered by the multiple cells within the measurement window through simulation; and determine the three-dimensional grid where the user in the multiple cells is located based on the location information of each user in the multiple cells, the size information of a single three-dimensional grid, the coordinate information of the three-dimensional grids covered by the multiple cells, the predicted channel quality information of the three-dimensional grids covered by the multiple cells within the measurement window, and the actual channel quality information of the user in the multiple cells within the measurement window. It can be seen that the third implementation is a combination of the first and second implementations. Specifically, if the location coordinates of a user fall exactly within a certain three-dimensional grid, and the user's actual channel quality information within the measurement window is consistent with or close to the predicted channel quality information of the three-dimensional grid within the measurement window (the difference is less than a preset threshold), then it is determined that the user is within that three-dimensional grid.

[0124] In the second and third embodiments described above, a user's actual channel quality information within the measurement window comes from the user's measurement report (MR) within the measurement window. As an example, the channel quality information may include at least one of Reference Signal Received Power (RSRP) and [other values].

[0125] In the network performance index backfilling module 402, there are many ways to process the network performance indexes of multiple users in the first grid to obtain the network performance index of the first grid. Two of them are introduced below.

[0126] In some embodiments, the average network performance index of all users within the first grid can be used as the network performance index of the first grid.

[0127] In other embodiments, considering that different network performance indicators of different users within the first grid have different effects on the network performance indicator of the first grid, the network performance indicators of multiple users within the first grid can be processed based on a multi-objective optimization algorithm to obtain the network performance indicator of the first grid. Here, the first grid is any grid in the three-dimensional grid covered by the multiple cells, and the user parameters of the multiple users come from the MR of the multiple users.

[0128] The network model construction module 403 is used to construct a multi-relationship network model based on at least one of the co-coverage relationship and configuration handover relationship of the multiple cells, wherein in the multi-relationship network model, a node represents a cell, and there is an edge between nodes with co-coverage relationship.

[0129] The contribution determination module 404 is used to determine the Pagerank contribution of the three-dimensional grid covered by the multiple cells to the network performance index of the first cell based on the network performance index of the three-dimensional grid covered by the multiple cells and the multi-relationship network model, wherein the first cell is any one of the multiple cells.

[0130] In some embodiments, the contribution determination module 404 may be used to: determine the neighboring cells of the first cell based on the multi-relationship network model; and determine the Pagerank contribution of the three-dimensional grid covered by the first cell and its neighboring cells to the network performance indicators of the first cell based on Pagerank.

[0131] Wherein, the network performance index of the first cell = ∑ contribution of the grid covered by the first cell * network performance index of the grid covered by the first cell + ∑ contribution of the grid covered by neighboring cells * network performance index of the grid covered by neighboring cells.

[0132] The profile information determination module 405 is used to aggregate the network performance indicators of the three-dimensional grids covered by the multiple cells based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, so as to obtain the profile information of the first cell.

[0133] In some embodiments, the profile information determination module 405 can be specifically used to: aggregate the network performance indicators of the three-dimensional grids covered by the multiple cells based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, and obtain the profile information of the first cell.

[0134] In other embodiments, Figure 4 The apparatus further includes an edge coefficient determination module, used to determine the edge coefficients of each edge in the multi-relationship network based on the mapping relationship between edges and relations and the relation strength. Correspondingly, the profile information determination module 405 can specifically be used to: aggregate the network performance indicators of the three-dimensional grids covered by the multiple cells based on the Pagerank contribution of the three-dimensional grids to the network performance indicators of the first cell, to obtain first profile information of the first cell; and determine second profile information of the first cell based on the first profile information and the edge coefficients of the first cell in the multi-relationship network model.

[0135] Specifically, the second profile information of the first cell can be obtained by multiplying the first profile information by the edge coefficient of the first cell in the multi-relationship network model.

[0136] The cell profile information determination device proposed in this application, on the one hand, aggregates user network performance indicators at the granularity of a three-dimensional grid, rather than directly fitting single or multiple network performance indicators at the cell dimension; on the other hand, when determining the cell profile information of a cell, it considers the network topology relationship between cells, rather than profiling based solely on the network performance indicators of that cell. Therefore, it can provide more intuitive, detailed, and effective cell profile information, which can better guide the network maintenance of the cell.

[0137] It should be noted that, Figure 4 The shown community profile information determination device 400 can achieve Figure 1 The method described in the embodiment achieves the same technical effect, and can be specifically referred to in the above description. Figure 1The method for determining cell profile information in the illustrated embodiment will not be described in detail here.

[0138] Figure 5 This is a schematic diagram of the structure of an electronic device provided in one embodiment of this application. Please refer to it. Figure 5 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and memory. The memory may include main memory, such as high-speed random-access memory (RAM), or non-volatile memory, such as at least one disk drive. Of course, the electronic device may also include other hardware required for other business operations.

[0139] The processor, network interface, and memory can be interconnected via an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 5 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0140] Memory is used to store programs. Specifically, programs may include program code, which includes computer operation instructions. Memory may include main memory and non-volatile memory, and provides instructions and data to the processor.

[0141] The processor reads the corresponding computer program from non-volatile memory into main memory and then executes it, forming a cell profile information determination device at the logical level. The processor executes the program stored in memory and specifically performs the following operations:

[0142] A three-dimensional electronic map containing multiple cells is divided into grids to obtain several three-dimensional grids, where one cell covers multiple three-dimensional grids;

[0143] The three-dimensional grid where each user is located in the plurality of cells is determined, and the network performance indicators of the plurality of users in the first grid are processed to obtain the network performance indicators of the first grid, wherein the first grid is any grid in the three-dimensional grid covered by the plurality of cells.

[0144] A multi-relationship network model is constructed based on at least one of the shared coverage relationship and configuration handover relationship of the multiple cells. In the multi-relationship network model, a node represents a cell, and there is an edge between nodes with shared coverage relationship.

[0145] Based on the network performance metrics of the three-dimensional grids covered by the multiple cells and the multi-relationship network model, the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance metrics of the first cell is determined, wherein the first cell is any one of the multiple cells.

[0146] Based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, the network performance indicators of the three-dimensional grids covered by the multiple cells are aggregated to obtain the profile information of the first cell.

[0147] The above is as stated in this application. Figure 1 The cell profile information determination method disclosed in the embodiments described above can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by integrated logic circuits in the processor's hardware or by instructions in software form. The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in one or more embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in one or more embodiments of this application can be directly implemented by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method.

[0148] The electronic device can also perform Figure 1The method for determining the community profile information is not described in detail here.

[0149] This application also proposes a computer-readable storage medium that stores one or more programs, the programs including instructions that, when executed by a portable electronic device including multiple applications, enable the portable electronic device to perform... Figure 1 The method of the embodiment shown is specifically used to perform the following operations:

[0150] A three-dimensional electronic map containing multiple cells is divided into grids to obtain several three-dimensional grids, where one cell covers multiple three-dimensional grids;

[0151] The three-dimensional grid where each user is located in the plurality of cells is determined, and the network performance indicators of the plurality of users in the first grid are processed to obtain the network performance indicators of the first grid, wherein the first grid is any grid in the three-dimensional grid covered by the plurality of cells.

[0152] A multi-relationship network model is constructed based on at least one of the shared coverage relationship and configuration handover relationship of the multiple cells. In the multi-relationship network model, a node represents a cell, and there is an edge between nodes with shared coverage relationship.

[0153] Based on the network performance metrics of the three-dimensional grids covered by the multiple cells and the multi-relationship network model, the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance metrics of the first cell is determined, wherein the first cell is any one of the multiple cells.

[0154] Based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, the network performance indicators of the three-dimensional grids covered by the multiple cells are aggregated to obtain the profile information of the first cell.

[0155] Of course, in addition to software implementation, the electronic device of this application does not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. In other words, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0156] This application also proposes a computer program product, which is stored in a storage medium and executed by at least one processor to implement the cell profile information determination method provided in this application.

[0157] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0158] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0159] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0160] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0161] It should be noted that all embodiments in this application are described in a related manner, and the same or similar parts between the embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for the device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiments.

[0162] It should also be noted that 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 process, method, article, or apparatus. Unless otherwise specified, 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 that element.

[0163] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for determining community profile information, characterized in that, The method includes: A three-dimensional electronic map containing multiple cells is divided into grids to obtain several three-dimensional grids, where one cell covers multiple three-dimensional grids; The three-dimensional grid where each user is located in the plurality of cells is determined, and the network performance indicators of the plurality of users in the first grid are processed to obtain the network performance indicators of the first grid, wherein the first grid is any grid in the three-dimensional grid covered by the plurality of cells. A multi-relationship network model is constructed based on at least one of the shared coverage relationship and configuration handover relationship of the multiple cells. In the multi-relationship network model, a node represents a cell, and there is an edge between nodes with shared coverage relationship. Based on the network performance metrics of the three-dimensional grids covered by the multiple cells and the multi-relationship network model, the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance metrics of the first cell is determined, wherein the first cell is any one of the multiple cells. Based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, the network performance indicators of the three-dimensional grids covered by the multiple cells are aggregated to obtain the profile information of the first cell.

2. The method according to claim 1, characterized in that, Determining the three-dimensional grid where each user is located within the plurality of cells includes: The predicted channel quality information of the three-dimensional grids covered by the multiple cells within the measurement time window was obtained through simulation. Based on the predicted channel quality information of the three-dimensional grids covered by the multiple cells within the measurement window, and the actual channel quality information of the users within the multiple cells within the measurement window, the three-dimensional grids where the users within the multiple cells are located are determined. The actual channel quality information of a user within the measurement window comes from the measurement report (MR) of that user within the measurement window.

3. The method according to claim 2, characterized in that, The channel quality information includes the Reference Signal Received Power (RSRP).

4. The method according to claim 1, characterized in that, The process of processing the network performance metrics of multiple users within the first grid to obtain the network performance metrics of the first grid includes: The network performance metrics of multiple users within the first grid are processed using a multi-objective optimization algorithm to obtain the network performance metrics of the first grid. The first grid is any grid in the three-dimensional grid covered by the multiple cells, and the user parameters of the multiple users are derived from the MR of the multiple users.

5. The method according to claim 1, characterized in that, In the multi-relationship network model, there are multiple relationships between nodes, and each relationship corresponds to a relationship strength, which represents the degree of influence of the relationship on the edges between nodes. The method further includes: Based on the mapping relationship between edges and relations and the strength of relations, the edge coefficients of each edge in the multi-relation network are determined; The process of aggregating the network performance indicators of the three-dimensional grids covered by the multiple cells to obtain the profile information of the first cell, based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, includes: Based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, the network performance indicators of the three-dimensional grids covered by the multiple cells are aggregated to obtain the first profile information of the first cell. Based on the first profile information and the edge coefficients of the first cell in the multi-relationship network model, the second profile information of the first cell is determined.

6. The method according to claim 5, characterized in that, The determination of the PageRank contribution of the three-dimensional grid covered by the multiple cells to the network performance index of the first cell, based on the network performance index of the three-dimensional grid covered by the multiple cells and the multi-relationship network model, includes: The neighboring cells of the first cell are determined based on the multi-relationship network model; The PageRank contribution of the three-dimensional grid covered by the first cell and its neighboring cells to the network performance metrics of the first cell is determined based on PageRank.

7. The method according to claim 6, characterized in that, The method involves aggregating the network performance indicators of the three-dimensional grids covered by the multiple cells to obtain the first profile information of the first cell, based on the Pagerank contribution of the grids to the network performance indicators of the first cell. This includes: Based on the Pagerank contribution of the three-dimensional grids covered by the first cell and its neighboring cells to the network performance indicators of the first cell, the network performance indicators of the three-dimensional grids covered by the first cell and its neighboring cells are aggregated to obtain the first profile information of the first cell.

8. A device for determining community profile information, characterized in that, The device includes: The 3D raster division module is used to divide a 3D electronic map containing multiple cells into raster grids, resulting in several 3D rasters, where one cell covers multiple 3D rasters. The network performance index backfilling module is used to determine the three-dimensional grid where each user is located in the multiple cells, and process the network performance index of the multiple users in the first grid to obtain the network performance index of the first grid, wherein the first grid is any grid in the three-dimensional grid covered by the multiple cells. The network model construction module is used to construct a multi-relationship network model based on at least one of the co-coverage relationship and configuration handover relationship of the multiple cells, wherein in the multi-relationship network model, a node represents a cell, and there is an edge between nodes with co-coverage relationship; The contribution determination module is used to determine the Pagerank contribution of the three-dimensional grid covered by the multiple cells to the network performance index of the first cell based on the network performance index of the three-dimensional grid covered by the multiple cells and the multi-relationship network model, wherein the first cell is any one of the multiple cells. The profile information determination module is used to aggregate the network performance indicators of the three-dimensional grids covered by the multiple cells based on the Pagerank contribution of the three-dimensional grids covered by the multiple cells to the network performance indicators of the first cell, so as to obtain the profile information of the first cell.

9. An electronic device, comprising: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the method as described in any one of claims 1-7.

10. A computer-readable storage medium, wherein instructions in the storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the method as described in any one of claims 1-7.

11. A computer program product, characterized in that, The program product is stored in a storage medium and is executed by at least one processor to implement the method as described in any one of claims 1-7.

Images