Position prediction method, apparatus and device

Abstract

The application discloses a position prediction method, device and equipment. The position prediction method comprises the following steps: obtaining a measurement report reported by a target terminal; determining a target state of the target terminal based on the measurement report; wherein the target state comprises a motion state or a non-motion state; determining position information of the target terminal based on a prediction model matched with the target state; wherein the prediction model comprises a first model or a second model, the first model is used for predicting position information of a terminal in a motion state, and the second model is used for predicting position information of a terminal in a non-motion state.

Description

Technical Field

[0001] This application relates to the field of wireless communication technology, and in particular to a location prediction method, apparatus, and device. Background Technology

[0002] With the rapid development of wireless communication technology, terminal location prediction has become a core requirement for scenarios such as intelligent network optimization, resource scheduling, and emergency services. Current location prediction schemes mainly rely on a single model to predict terminal location, but they do not fully consider the terminal's state characteristics (such as motion and inert states), resulting in low accuracy in location prediction. Summary of the Invention

[0003] The purpose of this application is to provide a location prediction method, apparatus, and device to solve the problem of low accuracy in location prediction caused by using a single model to predict the location of a terminal.

[0004] In a first aspect, a location prediction method is provided, comprising: acquiring a measurement report reported by a target terminal; determining a target state of the target terminal based on the measurement report; wherein the target state includes a motion state or a non-motion state; determining location information of the target terminal based on a prediction model matching the target state; wherein the prediction model includes a first model or a second model, the first model being used to predict the location information of a terminal in a motion state; and the second model being used to predict the location information of a terminal in a non-motion state.

[0005] Secondly, a location prediction method is provided, comprising: acquiring training data, the training data including first data of a terminal in motion and second data of a terminal in a non-motion state; training a first model based on the first data and training a second model based on the second data; wherein the first model is used to predict the location information of a terminal in motion; and the second model is used to predict the location information of a terminal in a non-motion state.

[0006] In some embodiments, after acquiring the training data, the method further includes: filtering the position bounce data and / or latitude and longitude drift data in the training data based on the motion direction angle difference corresponding to the training data.

[0007] Thirdly, a location prediction device is provided, comprising: an acquisition module for acquiring a measurement report reported by a target terminal; a state determination module for determining a target state of the target terminal based on the measurement report; wherein the target state includes a motion state or a non-motion state; and a location determination module for determining the location information of the target terminal based on a prediction model matching the target state; wherein the prediction model includes a first model or a second model, the first model being used to predict the location information of a terminal in a motion state; and the second model being used to predict the location information of a terminal in a non-motion state.

[0008] Fourthly, a location prediction device is provided, comprising: an acquisition module for acquiring training data, the training data including first data of a terminal in motion and second data of a terminal in a non-motion state; and a model training module for training a first model based on the first data and training a second model based on the second data; wherein the first model is used to predict the location information of a terminal in motion; and the second model is used to predict the location information of a terminal in a non-motion state.

[0009] Fifthly, an electronic device is provided, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the method as described in the first or second aspect.

[0010] In a sixth aspect, 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 as described in the first or second aspect.

[0011] In a seventh aspect, a computer program product is provided, comprising a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of the methods of the first or second aspect.

[0012] The location prediction method provided in this application involves acquiring a measurement report reported by a target terminal; determining the target state of the target terminal based on the measurement report, whereby the target state may be either a moving state or a non-moving state; and determining the location information of the target terminal based on a prediction model matching the target state. The prediction model may include a first model or a second model, where the first model is used to predict the location information of a terminal in a moving state, and the second model is used to predict the location information of a terminal in a non-moving state. This application embodiment performs location prediction based on both the measurement reports of terminals in a moving state and those of terminals in a non-moving state, thereby improving the accuracy of location prediction. Attached Figure Description

[0013] 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: Figure 1 This illustration shows a flowchart of a location prediction method provided in an embodiment of this application; Figure 2 This diagram illustrates the motion trajectory of the target terminal within the slice in an embodiment of this application. Figure 3 This illustration shows a flowchart of a location prediction method provided in an embodiment of this application; Figure 4 This illustration shows a flowchart of a location prediction method provided in an embodiment of this application; Figure 5 This document illustrates the flying points and processing effects in embodiments of this application. Figure 6 This illustration shows a schematic diagram of the distance between two points in an embodiment of this application. Figure 7 This diagram illustrates a bounce at a location in an embodiment of this application. Figure 8 This illustration shows two locations where a bounce occurs in an embodiment of this application. Figure 9 This document illustrates a schematic diagram of latitude and longitude drift in an embodiment of this application. Figure 10 This illustration shows a schematic diagram of the processing of location bounce data or latitude and longitude drift data in an embodiment of this application; Figure 11 This diagram illustrates the direction of motion in an embodiment of this application. Figure 12 This diagram illustrates the angle difference in the direction of motion in an embodiment of this application. Figure 13 This illustration shows a schematic diagram of position jump data processing in an embodiment of this application; Figure 14 This diagram illustrates latitude and longitude drift data processing in an embodiment of this application. Figure 15 This diagram illustrates the difference in the distance traveled between motion and non-motion states in the embodiments of this application. Figure 16 This diagram illustrates the difference in the distance traveled between motion and non-motion states in the embodiments of this application. Figure 17 This document illustrates an example of the variance difference between the motion direction angle differences of data generated in motion and non-motion states in embodiments of this application. Figure 18This illustration shows a feature comparison diagram of data generated in motion and non-motion states in an embodiment of this application; Figure 19 This illustration shows a schematic diagram of the motion state and position prediction steps in an embodiment of this application; Figure 20 This diagram illustrates a multi-task learning framework under the hard-sharing mechanism in an embodiment of this application. Figure 21 This diagram illustrates the structure of the location prediction device provided in an embodiment of this application. Figure 22 This diagram illustrates the structure of the location prediction device provided in an embodiment of this application. Figure 23 A schematic diagram of the hardware structure of an electronic device for implementing the location prediction method provided in the embodiments of this application. Detailed Implementation

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. The drawing numbers in this application are only used to distinguish the various steps in the solution and are not used to limit the execution order of the various steps. The specific execution order is subject to the description in the specification.

[0015] In related technologies, the main process of terminal location prediction scheme is as follows: cleaning measurement reports (MR) with location information (such as latitude and longitude information) and removing abnormal data, such as removing data that is too far away or has incomplete information; using the above data for feature extraction and model training, and establishing a single prediction model on a cell-by-cell basis; extracting features from measurement reports without location information, and using the above prediction model to predict the location of terminal measurement reports without location information.

[0016] However, when a terminal sends a measurement report, it may be in a stationary state (e.g., stationary) or a moving state. In a stationary state, factors such as changes in the atmospheric ionosphere, obstruction, and multipath reflection can cause satellite positioning results to drift or even fluctuate significantly, leading to a mismatch between the reported location and the actual wireless environment. In a moving state, due to continuous updates and corrections to the satellite signal, the drift is relatively smaller, and the reported location is more consistent with the wireless environment. Therefore, there are significant differences between measurement reports from stationary and moving terminals. It is necessary to establish separate position prediction models for moving and non-moving states based on the measurement reports from terminals in different states, to predict the location information of terminals in both states.

[0017] Figure 1 This diagram illustrates a flow chart of a location prediction method provided in an embodiment of this application. This method can be executed by an electronic device, such as a terminal or server device. The server device includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. Figure 1 As shown, the method may include the following steps.

[0018] S102: Obtain the measurement report reported by the target terminal.

[0019] In some embodiments, the target terminal may periodically report measurement reports to the network-side device, or the target terminal may report measurement reports to the network-side device based on event triggering or signaling triggering. Therefore, this step can obtain the measurement reports reported by the target terminal from the network-side device (such as a base station).

[0020] In some embodiments, the measurement report does not include the location information of the target terminal. The measurement report may include at least one of the following information: the field strength of the serving cell, the timing advance (TA), the angle of arrival (AoA), the field strength of neighboring cells, the timestamp, the identifiers (IDs) of the serving cell and neighboring cells, the identifier of the target terminal, etc.; wherein, the field strength of the serving cell or the field strength of the neighboring cells can be indicated by indicators such as Reference Signal Receiving Power (RSRP) or Reference Signal Receiving Quality (RSRQ).

[0021] S104: Determine the target state of the target terminal based on the measurement report; wherein the target state includes a motion state or a non-motion state.

[0022] In some embodiments, the field strength change rate of the serving cell can be determined based on the measurement report and the measurement report at historical time: ΔRSRP = RSRP(t) - RSRP(t-1). If ΔRSRP exceeds a first threshold and the time interval between two measurement reports is less than a second threshold, the target terminal is determined to be in motion. If ΔRSRP is less than or equal to the first threshold, and / or the time interval between two measurement reports is greater than or equal to the second threshold, the target terminal is determined to be in non-motion state.

[0023] In some embodiments, the target state of the target terminal can be determined based on the target state corresponding to the slice to which the measurement report belongs. That is, if the target state corresponding to the slice is a motion state, then the target terminal is a motion state; if the target state corresponding to the slice is a non-motion state, then the target terminal is a non-motion state.

[0024] In this embodiment, the measurement reports of the target terminal can be sorted by their ID (such as MSISDN, IMSI, etc.) and acquisition time. The latitude and longitude of the occupied cells are obtained by matching information such as occupied cells with engineering parameters. Then, the reports are grouped to obtain slices. Subsequent calculations can be performed using slices as units. The grouping condition can be: the time interval between two data points (i.e., two measurement reports) exceeds m minutes, such as 5 minutes. As shown in Table 1, in the first 9 data points, the time interval between any two adjacent data points does not exceed 5 minutes; therefore, the first 9 data points belong to one slice. The time interval between the 9th and 10th data points exceeds 5 minutes; therefore, the 10th data point belongs to another slice. From the 10th to the 18th data points, the time interval between any two adjacent data points does not exceed 5 minutes; therefore, the 10th to the 18th data points belong to one slice. Generally, each row of data in Table 1 corresponds to one measurement report.

[0025] Table 1

[0026] In some embodiments, after obtaining the slice, data cleaning operations can also be performed. For example, if there are at least two data points at the same time, only the first data point is retained; or if two data points with adjacent times have the same latitude and longitude, only the first occurrence of the data point is retained.

[0027] In some embodiments, the method further includes: determining the target state corresponding to the slice based on at least one of the following features: 1) The number of measurement reports n in the slice is shown in Table 1. The number of measurement reports in slice IDx-1 is 9; the number of measurement reports in slice IDx-2 is 9.

[0028] 2) After removing duplicate measurement reports within the slice, the number m of remaining measurement reports is generally considered to be valid measurement reports. In Table 1, after removing duplicate measurement reports from slice IDx-1, the number m of remaining measurement reports can be less than or equal to 9; after removing duplicate measurement reports from slice IDx-2, the number m of remaining measurement reports can be less than or equal to 9.

[0029] 3) The distance the target terminal moves within the time period corresponding to the slice. , It can be described as the sum of the distances between two adjacent points within a slice (the two adjacent points can be two data points that are adjacent in time), and can be calculated using the following formula:

[0030] The distance between two adjacent points within a slice can be calculated using the following formula:

[0031] in, The coordinates are the Earth's radius, calculated using the Earth's radius data for the region. Lon1 is the longitude of point 1, and Lat1 is the latitude of point 1. Lon2 is the longitude of point 2, and Lat2 is the latitude of point 2. The longitude and latitude in the formula can be the longitude and latitude of the serving cell, not the longitude and latitude of the terminal.

[0032] 4) The maximum distance between the first measurement report in the slice and the other measurement reports in the slice. , This can be described as the maximum distance between the first point in the slice and all other points in the slice, and can be calculated using the following formula:

[0033] It is understood that the embodiments of this application can determine the target state of the slice based on the features constituted by one or more of the combinations of 1) to 4) above. For example, the feature constituted by the combination of 1) and 2) is: the proportion of position change events within the slice = m / n; or the feature constituted by the combination of 3) and 4) is: .

[0034] The embodiments of this application identify the slice state based on multiple features of the slice. Because multiple features of the slice are considered, the accuracy of slice state identification is improved compared with the scheme that uses the terminal speed within the slice to identify the slice state.

[0035] In some embodiments, determining the target state corresponding to the slice includes: determining the target state corresponding to the slice based on the at least one feature and the threshold corresponding to each feature.

[0036] This embodiment can filter measurement reports generated by terminals in motion states using multiple feature combinations based on set thresholds. Slices that meet the threshold features in Table 2 are judged to be in motion states; slices that do not meet the threshold features in Table 2 are judged to be in non-motion states. For slices in motion states, the motion trajectory of the target terminal within the slice can be found in [reference needed]. Figure 2 The star-shaped pattern in the middle.

[0037] Table 2

[0038] The thresholds in Table 2 above are merely exemplary thresholds. The thresholds corresponding to the features may vary depending on the actual application scenario. This specification does not impose any specific limitations on these thresholds in the embodiments.

[0039] The embodiments of this application identify the slice state based on multiple features and thresholds of the slice. Because multiple features of the slice are considered, the accuracy of slice state identification is improved compared with the scheme that uses the terminal speed within the slice to identify the slice state.

[0040] In other embodiments, determining the target state corresponding to the slice includes: inputting the at least one feature into a third model to determine the target state corresponding to the slice; wherein, the third model is used to predict the target state corresponding to the slice, and the third model can be a binary classification model. Due to the advantages of Xgboost in performance and training speed, embodiments of this application can build a binary classification model based on Xgboost to achieve automated classification of the state corresponding to the slice, that is, to determine whether the slice is a moving state or a non-moving state.

[0041] The embodiments of this application can use artificial intelligence algorithms to pre-train a third model based on slices in motion and slices in non-motion states, and determine the target state corresponding to the slice based on the third model, which helps to improve the accuracy of slice state recognition.

[0042] The above embodiments of this application determine the target state corresponding to the slice, and the ultimate goal is to determine the target state corresponding to the measurement report. For example, if the time corresponding to the measurement report is located within a slice of motion, then the target state corresponding to the measurement report is determined to be a motion state; if the time corresponding to the measurement report is located within a slice of non-motion, then the target state corresponding to the measurement report is determined to be a non-motion state; or, for multiple measurement reports, the measurement reports whose time is located within a slice of motion are determined to have a target state of motion, while the remaining measurement reports are determined to be non-motion states.

[0043] S106: Determine the location information of the target terminal based on a prediction model that matches the target state; wherein the prediction model includes a first model or a second model, the first model is used to predict the location information of the terminal in motion; the second model is used to predict the location information of the terminal in non-motion state.

[0044] In this embodiment, when the target state is in motion, the location information of the target terminal is determined based on a first model; when the target state is in a non-motion state, the location information of the target terminal is determined based on a second model. The location information in various embodiments of this application can be latitude and longitude information.

[0045] In some embodiments, determining the location information of the target terminal based on a prediction model matching the target state includes: determining the location information of the target terminal based on a prediction model matching the target state and at least one of the following features: 1) the field strength of the serving cell; 2) the timing advance of the serving cell; 3) the AOA of the serving cell; 4) the field strength of neighboring cells; 5) the difference in field strength between the neighboring cells and the serving cell; 6) the absolute value of the difference in field strength between the neighboring cells and the serving cell.

[0046] This application embodiment targets the measurement report generated by the target terminal in motion. The features obtained are shown in Table 3. The field strength, timing advance, and AOA (if any) of the serving cell can be used as the first type of training features. This application embodiment can also filter out neighboring cells that appear less than 1% of the total number of times. This application embodiment can also take the average field strength of each neighboring cell in each measurement report, sort them from largest to smallest, and take the field strength of the top 30 neighboring cells as the second type of training features. The difference between the field strength of the neighboring cell and the field strength of the serving cell is calculated as the third type of training feature, and the absolute value of the difference between the field strength of the neighboring cell and the serving cell is used as the fourth type of training feature.

[0047] Table 3

[0048] In terms of feature generation for location prediction, this application addresses the issues of differences in wireless capabilities between terminals from different manufacturers and large differences in RSRP at the same location. It introduces two types of features: the field strength difference between neighboring cells and the serving cell, and the absolute value of the field strength difference between neighboring cells and the serving cell. This helps to avoid the problem of reduced location prediction accuracy caused by large differences in RSRP at the same location, and is beneficial to improving location prediction accuracy. Secondly, for the field strength of the serving cell and the field strength of neighboring cells, RSRP can be used as a feature instead of RSRQ, which is affected by cell load, which is beneficial to improving location prediction accuracy.

[0049] The location prediction method provided in this application involves acquiring a measurement report reported by a target terminal; determining the target state of the target terminal based on the measurement report, whereby the target state may be either a moving state or a non-moving state; and determining the location information of the target terminal based on a prediction model matching the target state. The prediction model may include a first model or a second model, where the first model is used to predict the location information of a terminal in a moving state, and the second model is used to predict the location information of a terminal in a non-moving state. This application embodiment performs location prediction based on measurement reports from both moving and non-moving terminals, which helps improve the accuracy of location prediction.

[0050] The location prediction method provided in this application, compared with the method of mixing measurement reports generated by terminals in different states together to establish a location prediction model and make location prediction, proposes to establish location prediction models for terminals in motion and terminals in non-motion states based on measurement reports of terminals in different states, and to make location predictions based on measurement reports of terminals in motion and non-motion states respectively, thereby improving the accuracy of location prediction.

[0051] In some embodiments, before obtaining the measurement report reported by the target terminal, the method further includes: obtaining training data, the training data including first data of a terminal in motion and second data of a terminal in non-motion; training a first model based on the first data and training a second model based on the second data; wherein the first model is used to predict the location information of a terminal in motion; and the second model is used to predict the location information of a terminal in non-motion.

[0052] Based on the measurement reports of terminals in different states, this application establishes position prediction models for terminals in motion and terminals in non-motion states, respectively. This facilitates position prediction based on measurement reports of terminals in motion and terminals in non-motion states, thereby improving the accuracy of position prediction.

[0053] The above combination Figure 1 and Figure 2 The location prediction method according to embodiments of this application is described in detail. The following will combine... Figure 3 A location prediction method according to another embodiment of this application is described in detail.

[0054] Figure 3 This is a schematic diagram illustrating the implementation process of the location prediction method according to an embodiment of this application, which can be applied to electronic devices. For example... Figure 3 As shown, the method 300 includes the following steps.

[0055] S302: Acquire training data, which includes first data of the terminal in motion and second data of the terminal in non-motion state.

[0056] The first and second data in various embodiments of this application may include measurement reports. For details regarding the identification of the terminal's motion and non-motion states, please refer to the descriptions of other embodiments.

[0057] S304: A first model is trained based on the first data and a second model is trained based on the second data; wherein, the first model is used to predict the position information of the terminal in motion; and the second model is used to predict the position information of the terminal in non-motion state.

[0058] The embodiments of this application can train a first model based on at least one of the following features of the first data: the ratio of travel distance, the variance of the difference in the angle of movement direction, and the proportion of effective data; the embodiments of this application can train a second model based on at least one of the following features of the second data: the ratio of travel distance, the variance of the difference in the angle of movement direction, and the proportion of effective data.

[0059] In terms of feature generation, this application's embodiments design features such as the ratio of travel distance, variance of the difference in motion direction angle, and percentage of effective data based on the characteristics of the motion state data source in terms of continuity (time, distance), regularity (direction changes are not too frequent), and speed (within a reasonable range). Compared with the scheme that only uses speed to determine motion state, it can more effectively reflect the difference between data generated in motion and non-motion states, avoid low-speed motion data being judged as non-motion state, thereby obtaining richer motion state data for analysis and improving the prediction accuracy of the obtained model.

[0060] Based on the measurement reports of terminals in different states, this application establishes position prediction models for terminals in motion and terminals in non-motion states, respectively. This facilitates position prediction based on measurement reports of terminals in motion and terminals in non-motion states, thereby improving the accuracy of position prediction.

[0061] In some embodiments, after acquiring the training data, the method further includes: filtering the position bounce data and / or latitude and longitude drift data in the training data based on the motion direction angle difference corresponding to the training data.

[0062] In terms of data cleaning and processing, compared with solutions that only remove invalid and duplicate data, this application embodiment can process flying point data, location bounce data, and latitude and longitude drift data. This can effectively solve the impact of flying point, location bounce, and latitude and longitude drift on the variance of the motion direction angle difference, so that the variance of the motion direction angle difference can better distinguish between moving and non-moving data sources. At the same time, for the stationary state, due to the disordered movement, a large number of records are merged, and thus, in terms of the effective data ratio, the data generated in the moving and non-moving states have obvious differences.

[0063] To illustrate the location prediction method provided in the embodiments of this application in detail, the following will describe it in conjunction with several specific embodiments.

[0064] This application proposes a location prediction method based on terminal state. In the modeling stage, for measurement reports with latitude and longitude information, location prediction models are established for both moving and non-moving measurement reports, respectively. In the prediction stage, for measurement reports without latitude and longitude information, the corresponding models are used for location prediction based on their state. The flowchart of the technical solution of this application embodiment is shown below. Figure 4 As shown.

[0065] The following sections will describe the implementation steps of the modeling phase, combining steps one through five.

[0066] Step 1: Slice generation.

[0067] In this embodiment, measurement reports for the same terminal number are sorted by time according to their ID (such as MSISDN, IMSI, etc.), and then grouped to obtain slices. Subsequent calculations are all based on slices. The grouping condition can be that the time interval between two data points (i.e., two measurement reports) exceeds m minutes, such as 5 minutes. The obtained slices can be shown in Table 4 below.

[0068] Table 4

[0069] Table 4 differs from Table 1. The measurement report in Table 4 generally includes the latitude and longitude information of the terminal, while the measurement report in Table 1 corresponds to the latitude and longitude information of the cell base station.

[0070] Step 2: Data cleaning and processing.

[0071] Since the measurement reports submitted by the terminal may contain invalid data, duplicate data, data with too close time intervals, and abnormal data, it is necessary to clean and process the invalid and abnormal data within the slice before performing feature calculations to improve the effectiveness of feature calculation results and the accuracy of subsequent terminal status judgments.

[0072] Data cleaning can include removing invalid latitude and longitude data (such as 0 and empty strings). To reduce the impact of excessively short time intervals on speed calculations, the time in the original data can be retained only to the second. For data from the same time, only the first data entry should be retained. For adjacent time data with the same latitude and longitude, only the first occurrence of that data should be retained.

[0073] After data cleaning, "flying points" in the slices can be removed. "Flying points" are points that are far from the central data area within the slice. Figure 5 Within the slice shown, the "flying point" can be point 6.

[0074] This application embodiment can obtain the distance and velocity between each point within a slice and its two adjacent points, and the relationship between these distances and velocity and preset distance and velocity thresholds, to filter out "flying points". For example, points whose distance is greater than a preset distance threshold and / or whose velocity is greater than a preset velocity threshold are identified as "flying points". After deleting "flying points", this application embodiment can also clear data with the same latitude and longitude at adjacent times.

[0075] In this embodiment of the application, the distance between two points within a slice can be calculated using the latitude and longitude of the two points, such as... Figure 6 Illustration. The distance between two points can be determined using the following formula:

[0076] in, Lng1 represents the Earth's radius, calculated using the Earth's radius data for the region; Lng1 represents the longitude of point 1, and Lat1 represents the latitude of point 1; Lng2 represents the longitude of point 2, and Lat2 represents the latitude of point 2.

[0077] In this embodiment of the application, the velocity between two points can be calculated using the distance between the two points and the time difference between them:

[0078] In various embodiments of this application, due to latitude and longitude confusion, position rebound data and / or latitude and longitude drift data may occur. The embodiments of this application filter the position rebound data and / or latitude and longitude drift data based on the angle difference of the movement direction.

[0079] Position rollback refers to a situation where, due to unknown reasons, the latitude and longitude position of a later timestamp is opposite to the direction of movement, such as... Figure 7 , Figure 8 As shown, the arrows indicate the direction of movement, and the points in the time sequence are later ( Figure 7 point 6, Figure 8 Points 6 and 7 are located to the left of point 5, which is earlier in the time sequence. They then resumed their original direction of movement, causing the direction of movement to change by nearly 180 degrees twice, which is inconsistent with the actual state of movement.

[0080] Latitude and longitude drift refers to the instability of location data caused by a stationary or slowly moving object, resulting in a significant difference from the actual motion state. Therefore, latitude and longitude drift data needs to be processed, such as... Figure 9 In the diagram, the two points inside the dashed box are the points where latitude and longitude have shifted.

[0081] This application embodiment can judge the angle difference of the motion direction in chronological order. When a measurement report with an angle greater than 90° appears, that measurement report (or data) is marked, and the motion direction angle difference of subsequent data is judged and identified using the same mark. When data with an angle less than 90° appears continuously at least m times (e.g., ... Figure 10 In the example shown, this is 6 times (up to position T21). The movement is then judged to have entered a more regular state, and the current marking ends. The first ( Figure 10 (T6 in the middle) and the last one ( Figure 10 For data in T15 (where the angle difference in the direction of motion is greater than 90°), the time, longitude, and latitude involved are averaged separately, replacing the data used in the averaging process. After the new data is generated, data with the same longitude and latitude at adjacent times can also be cleared. The data generated during this process is as follows: Figure 10 As shown.

[0082] In various embodiments of this application, the direction of movement can be defined as north as 0°, and the angle between the line connecting point 1 and point 2 and north as the direction of movement, with a value ranging from 0° to 180°. Figure 11 As shown.

[0083] When the calculated angle is greater than or equal to 0° and less than or equal to 180°:

[0084] Angle difference in direction of motion: Sorted by time, adjacent points form a line segment. The angle between two adjacent line segments is the angle difference in direction of motion, ranging from 0 to 180°. Figure 12 As shown, Figure 12 What is explicitly shown is the difference in the angle of motion between the line segment formed by points 2 and 3 and the line segment formed by points 3 and 4.

[0085]

[0086] The embodiments of this application, based on the motion direction angle difference, can process the position rebound data as follows: Figure 13 As shown, the processing results for latitude and longitude drift data can be as follows: Figure 14 As shown.

[0087] Step 3: Feature calculation.

[0088] This step can perform motion feature calculations on slices that have undergone data cleaning and outlier processing.

[0089] Feature 1: Distance traveled by the slice: Sort by time and calculate the sum of the distances between any two adjacent points within the slice.

[0090] Among them, such as Figure 6 As shown, The distance can be calculated using the latitude and longitude of these two points. The formula for calculating the distance is as follows:

[0091] Feature 2, First and Last Point Distance: Points within a slice can be sorted by time, and the distance between the first and last points within the slice can be calculated.

[0092] Feature 3: Maximum distance between the first point and other points in the slice: This can be sorted by time to calculate the maximum distance between the first point and all other points.

[0093] Feature 4: Distance between the first point and the average latitude and longitude of the slice: Points within the slice can be sorted by time, and the distance between the first point and the average latitude and longitude of all other points within the slice can be calculated.

[0094] in, The distance between the starting point and the average latitude and longitude of the slice. This represents the average latitude and longitude of the data for other points within the slice.

[0095] Feature 5, Maximum Distance Between Two Points: Calculates the maximum distance between any two points within a slice.

[0096] Feature 6: Slice Usage Time: Points within a slice can be sorted by time, and the time difference between the first and last points can be calculated.

[0097] Feature 7, Intra-slice velocity: Calculates the velocity of the slice.

[0098] Feature 8, Travel Ratio 1: Distance between the first and last points divided by the distance the slice travels.

[0099] Feature 9, Travel Ratio 2: The maximum distance between the first point and other points in the slice divided by the distance the slice travels.

[0100] Feature 10, Travel Ratio 3: The distance between the initial point and the average latitude and longitude of the slice divided by the distance the slice has traveled.

[0101] Feature 11, Travel Ratio 4: The maximum distance between two points divided by the distance traversed by the slice.

[0102] Feature 12, Variance of the angle difference of the motion direction: When there are M data in the slice, M-1 line segments can be formed and M-1 motion directions can be calculated. The angle difference of the motion direction can be calculated for adjacent line segments. The variance of the angle difference of the motion direction is the variance of the angle difference of all motion directions in the slice.

[0103] Angle difference in direction of motion: Sorted by time, adjacent points form a line segment. The angle between two adjacent line segments is the angle difference in direction of motion, ranging from 0 to 180°. Figure 12 As shown.

[0104]

[0105] Variance of the angle difference in motion directions: Calculate the variance of the angle differences in all motion directions within the slice.

[0106] in, Let X be the variance of the angular difference in the direction of motion, and let X be the angular difference in the direction of motion between two adjacent line segments within the slice. is the average of the angle differences in the motion directions of two adjacent line segments within the slice, and N is the number of angle differences in the motion directions within the slice.

[0107] Feature 13, Number of valid data entries within a slice: The number of data entries within a slice after all steps in step two are completed.

[0108] Feature 14: Percentage of valid data in a slice: The number of valid data entries in a slice divided by the total number of data entries in the slice.

[0109] Step 4: Status recognition.

[0110] The characteristics of data in motion and non-motion states, apart from speed, show significant differences in the ratio of distance traveled, the variance of the difference in motion direction angle, and the proportion of effective data in the slice.

[0111] like Figure 15 and Figure 16 As shown, the data generated in motion travels a larger distance, while the data generated in non-motion states travels a smaller distance due to the irregularity of motion.

[0112] like Figure 17 As shown, the variance of the angle difference of motion direction of data generated in motion state is small, while the variance of data source in non-motion state is large because the motion is irregular and the direction changes drastically.

[0113] In this embodiment of the application, the feature comparison of data generated in motion and non-motion states is as follows: Figure 18 As shown.

[0114] After obtaining the 14-dimensional features for each slice, there are two ways to determine whether the slice was generated in a moving or non-moving state. The first method is to set a reasonable threshold and use multiple feature combinations to distinguish between data generated in moving and non-moving states. Table 5 below shows the reference threshold values ​​for distinguishing between the two motion states: Table 5

[0115] The reference thresholds in Table 5 above are merely exemplary thresholds. The thresholds corresponding to the features may vary depending on the actual application scenario. This specification does not impose specific limitations on these thresholds in the embodiments.

[0116] The second method is based on deep learning. Training data is constructed using the 14-dimensional features of the slices and slices in known states, classifying the slices into two categories: moving and non-moving states. Due to the performance and training speed advantages of Xgboost, this application builds a binary classification model based on Xgboost to achieve automated classification of slice states. Specifically, the model uses the L1 loss function, sets the maximum tree depth to 6, the initial learning rate to 0.001, and conducts 500 training epochs. In the training data, the ratio of moving to non-moving state data is 5:5, with approximately 6000 samples.

[0117] Step 5: Training the position prediction model for both moving and non-moving states.

[0118] The above four steps separate the measurement reports generated under motion conditions from the measurement reports containing latitude and longitude information, and the measurement reports generated under non-motion conditions. This application uses the measurement reports containing latitude and longitude information as training data, and uses the field strength, timing advance, and field strength of neighboring cells of each measurement report as source data. For each cell, a motion state location prediction model and a non-motion state location prediction model based on multi-task learning and LightGBM are established to predict the latitude and longitude of MR data without latitude and longitude information, such as... Figure 19 As shown.

[0119] To maintain the effectiveness of neighboring cells and ensure the consistency of the neighboring cell order in each measurement report within each cell, this application first uses the field strength, timing advance, and AOA (if any) of the serving cell as the first type of training features for selecting neighboring cells. Then, neighboring cells that appear less than 1% of the total number of times are filtered out from the training data. Finally, the average field strength of each neighboring cell in each measurement report is taken, and the field strengths of the top 30 neighboring cells are selected from largest to smallest as the second type of training features. The difference between the field strength of the neighboring cell and the field strength of the serving cell is calculated as the third type of training feature, and the absolute value of the field strength difference is used as the fourth type of training feature. The detailed data structure of the training features is shown in Table 6.

[0120] Table 6

[0121] The core of the location prediction model is to predict the longitude and latitude of the MR measurement report based on information such as the field strength of the serving cell and neighboring cells. Since longitude and latitude jointly express the accurate positioning of the measurement report, these two predicted values ​​are strongly correlated and should be predicted within a unified framework. This application uses a multi-task learning architecture to establish separate location prediction models for moving and non-moving states, achieving simultaneous prediction of longitude and latitude for both moving and non-moving state measurement reports. Multi-task learning refers to tasks sharing certain structures and semantic relationships, achieving transfer and coordinated learning by sharing some model parameters. Specifically, when sharing parameters, this application adopts a hard-sharing mechanism for hidden layer parameters, sharing the hidden layer for longitude and latitude prediction while retaining one output layer for both tasks to achieve prediction. This sharing method reduces the risk of overfitting during training, specifically as follows: Figure 20 As shown.

[0122] Because many measurement reports have few neighboring regions, there are many null values ​​in the features. LightGBM's optimization for sparse data makes it perform well in processing datasets with a large number of missing values. LightGBM also supports multiple optimization objectives and custom loss functions to meet the needs of different scenarios. In this application, LightGBM is used as the base regressor in a multi-task learning architecture. Minimization of Drive Tests (MDT) data and a small amount of MR data with latitude and longitude are used as training data to predict the latitude and longitude of MR data without location information.

[0123] Regarding the model hyperparameter settings, the motion state prediction model and the non-motion state prediction model of this application use the same hyperparameters, setting the tree model complexity to 31, the maximum tree depth to 6, the learning rate to 0.001, and the number of training epochs to 1000.

[0124] Regarding the training sample size, there are more non-motion state measurement reports than motion state measurement reports in each cell. Therefore, for each cell, an average of 50,000 motion measurement reports can be extracted as training data for the motion state location prediction model, and an average of 80,000 non-motion measurement reports can be extracted as training data for the non-motion state location prediction model.

[0125] Regarding the loss function, since the data may contain a small number of outliers, this application uses HuberLoss as the model's loss function. This loss function combines mean squared error and mean absolute error, aiming to overcome the shortcomings of both. Mean squared error is used for small errors, while mean absolute error is used for large errors, thus better representing the spatial relationships between measurement reports and reducing the impact of outliers on latitude and longitude predictions. The specific formula for this loss function is as follows:

[0126] In the formula, Represents the Huber Loss function. Represents the true value. Indicates the predicted value. These are hyperparameters used to control the limits of error magnitude; when the error is less than... When using mean square error, when the error is greater than The mean absolute error is used. Since even small differences between latitude and longitude can lead to large-area offsets in measurement reports, this application... The value is 0.00001.

[0127] Figure 21The diagram shows a structural schematic of a location prediction device 2100 provided in an embodiment of this application. The device 2100 includes the following modules.

[0128] The acquisition module 2102 is used to acquire the measurement report reported by the target terminal.

[0129] The state determination module 2104 is used to determine the target state of the target terminal based on the measurement report; wherein the target state includes a motion state or a non-motion state.

[0130] The location determination module 2106 is used to determine the location information of the target terminal based on a prediction model that matches the target state; wherein the prediction model includes a first model or a second model, the first model is used to predict the location information of the terminal in motion, and the second model is used to predict the location information of the terminal in non-motion state.

[0131] The embodiments of this application perform location prediction based on measurement reports from terminals in motion and terminals in non-motion states, which helps to improve the accuracy of location prediction.

[0132] In some embodiments, the state determination module 2104 is configured to determine the target state of the target terminal based on the target state corresponding to the slice to which the measurement report belongs; the state determination module 2104 is further configured to determine the target state corresponding to the slice based on at least one of the following features: 1) the number n of measurement reports in the slice; 2) the number m of measurement reports remaining in the slice after removing measurement reports with duplicate locations; 3) the distance the target terminal moves within the time period corresponding to the slice. 4) The maximum distance between the first measurement report in the slice and the other measurement reports in the slice. .

[0133] In some embodiments, the state determination module 2104 is used to determine the target state corresponding to the slice based on the at least one feature and the threshold corresponding to each feature; or, input the at least one feature into a third model to determine the target state corresponding to the slice; wherein, the third model is used to predict the target state corresponding to the slice.

[0134] In some embodiments, the location determination module 2106 is used to determine the location information of the target terminal based on a prediction model matching the target state and at least one of the following features: 1) the field strength of the serving cell; 2) the time advance of the serving cell; 3) the AOA of the serving cell; 4) the field strength of the neighboring cells; 5) the difference in field strength between the neighboring cells and the serving cell; 6) the absolute value of the difference in field strength between the neighboring cells and the serving cell.

[0135] In some embodiments, the acquisition module 2102 is further configured to acquire training data, the training data including first data of a terminal in motion and second data of a terminal in non-motion state; the device further includes a model training module, configured to train a first model based on the first data and train a second model based on the second data; wherein, the first model is used to predict the position information of a terminal in motion state; and the second model is used to predict the position information of a terminal in non-motion state.

[0136] Figure 22 The diagram shows a structural schematic of a location prediction device 2200 provided in an embodiment of this application. The device 2200 includes the following modules.

[0137] The acquisition module 2202 is used to acquire training data, which includes first data of the terminal in motion and second data of the terminal in non-motion state.

[0138] The model training module 2204 is used to train a first model based on the first data and a second model based on the second data; wherein, the first model is used to predict the position information of the terminal in motion; and the second model is used to predict the position information of the terminal in non-motion state.

[0139] Based on the measurement reports of terminals in different states, this application establishes position prediction models for terminals in motion and terminals in non-motion states, respectively. This facilitates position prediction based on measurement reports of terminals in motion and terminals in non-motion states, thereby improving the accuracy of position prediction.

[0140] In some embodiments, after acquiring the training data, the model training module 2204 is further configured to filter the position bounce data and / or latitude and longitude drift data in the training data based on the motion direction angle difference corresponding to the training data.

[0141] The device provided in this application embodiment can execute any of the embodiments described in the foregoing method embodiments and achieve the functions and beneficial effects of any of the embodiments described in the foregoing method embodiments, which will not be repeated here.

[0142] In this application, the modules in the apparatus provided can also implement the method steps provided in the method embodiments. Alternatively, the apparatus provided in this application may include other modules besides those described above to implement the method steps provided in the method embodiments. Furthermore, the apparatus provided in this application can achieve the technical effects achievable by the method embodiments.

[0143] This application also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the various processes of the above-described location prediction method embodiments and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0144] Figure 23 The diagram illustrates the hardware structure of an electronic device implementing the embodiments of this application. Referring to the diagram, at the hardware level, the electronic device includes a processor, and may also include an internal bus, a network interface, and a memory. The memory may include RAM, such as high-speed random-access memory (RAM), and may also include non-volatile memory, such as at least one disk storage device. Of course, the electronic device may also include other hardware required for other services.

[0145] The processor, network interface, and memory can be interconnected via an internal bus, which can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be categorized as an address bus, data bus, control bus, etc. For ease of illustration, only a single bidirectional arrow is used in this diagram, but this does not imply that there is only one bus or one type of bus.

[0146] 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.

[0147] The processor reads the corresponding computer program from non-volatile memory into main memory and then executes it, forming a device for locating the target terminal at the logical level. The processor executes the program stored in memory and specifically performs the following: Figure 1-20 The methods disclosed in the embodiments shown achieve the functions and beneficial effects of the methods described in the preceding method embodiments, and will not be repeated here.

[0148] The above is as stated in this application. Figure 1-20The methods disclosed in the illustrated embodiments can be applied to or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above methods 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 the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed 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.

[0149] The electronic device can also execute any of the embodiments described in the foregoing method embodiments and achieve the functions and beneficial effects of any of the embodiments described in the foregoing method embodiments, which will not be repeated here.

[0150] 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.

[0151] This application also provides a computer-readable storage medium storing a computer program. When executed by a processor, this computer program implements the various processes of the above-described location prediction method embodiments and achieves the same technical effects. To avoid repetition, it will not be described again here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0152] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program. The computer program is operable to cause a computer to perform some or all of the steps of the above-described location prediction method embodiments and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0153] 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.

[0154] 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.

[0155] 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.

[0156] 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.

[0157] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0158] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0159] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0160] 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 a process, method, article, or apparatus. Without further limitation, 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.

[0161] 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.

[0162] The above description is merely an embodiment of this application and is 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 location prediction method, characterized in that, include: Obtain the measurement report reported by the target terminal; The target state of the target terminal is determined based on the measurement report; wherein, the target state includes a moving state or a non-moving state; Based on a prediction model that matches the target state, the location information of the target terminal is determined; wherein, the prediction model includes a first model or a second model, the first model is used to predict the location information of the terminal in motion; the second model is used to predict the location information of the terminal in non-motion state.

2. The method according to claim 1, characterized in that, Determining the target state of the target terminal based on the measurement report includes: determining the target state of the target terminal based on the target state corresponding to the slice to which the measurement report belongs; the method further includes: determining the target state corresponding to the slice based on at least one of the following features: The number of measurement reports within the slice, n; After removing duplicate measurement reports within the slice, the number of remaining measurement reports is m. The distance the target terminal moves within the time period corresponding to the slice; The maximum distance between the first measurement report in the slice and the other measurement reports in the slice. .

3. The method according to claim 2, characterized in that, Determining the target state corresponding to the slice includes: Based on the at least one feature and the threshold corresponding to each feature, determine the target state corresponding to the slice; or, The at least one feature is input into a third model to determine the target state corresponding to the slice; wherein the third model is used to predict the target state corresponding to the slice.

4. The method according to claim 1, characterized in that, Determining the location information of the target terminal based on a prediction model matching the target state includes: determining the location information of the target terminal based on a prediction model matching the target state and at least one of the following features: The field strength serving the community; Lead time for serving the community; The angle of arrival (AOA) of the service area; Field strength in neighboring cells; The difference in field strength between neighboring cells and the service cell; The absolute value of the difference in field strength between the neighboring cell and the serving cell.

5. The method according to any one of claims 1 to 4, characterized in that, Before obtaining the measurement report reported by the target terminal, the method further includes: Acquire training data, which includes first data of the terminal in motion and second data of the terminal in non-motion state; A first model is trained based on the first data, and a second model is trained based on the second data; wherein, the first model is used to predict the location information of a terminal in motion, and the second model is used to predict the location information of a terminal in non-motion state.

6. A location prediction method, characterized in that, include: Acquire training data, which includes first data of the terminal in motion and second data of the terminal in non-motion state; A first model is trained based on the first data, and a second model is trained based on the second data; wherein, the first model is used to predict the position information of the terminal in motion state; The second model is used to predict the location information of the terminal in a non-motion state.

7. A location prediction device, characterized in that, include: The acquisition module is used to acquire the measurement report reported by the target terminal; A state determination module is used to determine the target state of the target terminal based on the measurement report; wherein the target state includes a motion state or a non-motion state; A location determination module is used to determine the location information of the target terminal based on a prediction model that matches the target state; wherein the prediction model includes a first model or a second model, the first model is used to predict the location information of the terminal in motion, and the second model is used to predict the location information of the terminal in non-motion state.

8. A location prediction device, characterized in that, include: The acquisition module is used to acquire training data, which includes first data of the terminal in motion and second data of the terminal in non-motion state. The model training module is used to train a first model based on the first data and a second model based on the second data; wherein, the first model is used to predict the position information of the terminal in motion state; The second model is used to predict the location information of the terminal in a non-motion state.

9. An electronic device, characterized in that, include: 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, implements the steps of the method as described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the method as described in any one of claims 1 to 6.

11. A computer program product, characterized in that, The computer program product includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of the method as described in any one of claims 1 to 6.

Images