Drive test data processing method, device and equipment, and storage medium

Abstract

The present disclosure provides a road test data processing method and device, equipment and storage medium, relating to the technical field of deep learning, comprising: acquiring a road test data set and a first latitude and longitude of each planning point, wherein each road test data includes a second latitude and longitude of a sampling point, a sampling frequency and an interference value; based on each first latitude and longitude, a planning point binary tree is constructed; based on the second latitude and longitude in each road test data, the binary tree is traversed to determine the sampling point set associated with each planning point; based on a preset interference identification model, the road test data in each sampling point set is identified to determine the interference type of each sampling point. Thus, the original test sampling road test data set can be automatically counted, analyzed and quickly evaluated, and the inter-network interference of each planning site location can be automatically output, and whether the inter-network interference will be received can be intelligently judged.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of deep learning, and particularly relates to a drive test data processing method and device, equipment and a storage medium. BACKGROUND

[0002] The uplink rate decreases by 50%-92% and the downlink rate decreases by 9%-23% for every 10dB increase in interference. Inter-network interference of a network affects the overall quality of the network. If the planning points that may be affected by inter-network interference can be identified in advance before the base station is built, the value of deploying the planning site and the user experience will be greatly improved.

[0003] The existing planning station inter-network interference analysis method mainly analyzes the frequency sweep instrument test data manually, which is easy to miss and depends on the experience of employees, and the work efficiency is low. Therefore, how to accurately analyze the interference of the planning station is a problem to be solved at present. SUMMARY

[0004] The present disclosure aims to at least partially solve one of the technical problems in the related art.

[0005] The first aspect of the present disclosure provides a drive test data processing method, comprising:

[0006] obtaining a drive test data set and a first latitude and longitude of each planning point, wherein each drive test data includes a second latitude and longitude of a sampling point, a sampling frequency and an interference value;

[0007] constructing a planning point binary tree based on each first latitude and longitude;

[0008] traversing the binary tree based on the second latitude and longitude in each drive test data to determine a set of sampling points associated with each planning point;

[0009] identifying the drive test data in each set of sampling points based on a preset interference identification model to determine the interference type of each sampling point.

[0010] The second aspect of the present disclosure provides a drive test data processing device, comprising:

[0011] an obtaining module configured to obtain a drive test data set and a first latitude and longitude of each planning point, wherein each drive test data includes a second latitude and longitude of a sampling point, a sampling frequency and an interference value;

[0012] a constructing module configured to construct a planning point binary tree based on each first latitude and longitude;

[0013] a first determining module configured to traverse the binary tree based on the second latitude and longitude in each drive test data to determine a set of sampling points associated with each planning point;

[0014] The second determining module is configured to identify the drive test data in each of the sampling point set based on a preset interference identification model to determine the interference type of each sampling point.

[0015] The third aspect of the present disclosure provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for processing drive test data according to the first aspect of the present disclosure is implemented.

[0016] The fourth aspect of the present disclosure provides a non-transitory computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the method for processing drive test data according to the first aspect of the present disclosure is implemented.

[0017] The method, device, equipment and storage medium for processing drive test data provided by the present disclosure have the following beneficial effects:

[0018] In the embodiments of the present disclosure, first, a set of drive test data and a first latitude and longitude of each planning point are obtained, wherein the second latitude and longitude of a sampling point, the sampling frequency and the interference value are included in each of the drive test data, then a planning point binary tree is constructed based on the first latitude and longitude, then the binary tree is traversed based on the second latitude and longitude in each of the drive test data to determine a set of sampling points associated with each of the planning points, and then the drive test data in each of the sampling point set is identified based on a preset interference identification model to determine the interference type of each sampling point. Thus, the set of drive test data of the original test sampling can be automatically counted, analyzed and quickly evaluated, the inter-network interference at each planning site position can be automatically output, and it can be intelligently determined whether the inter-network interference will occur, thereby solving the problem of low automation and intelligence of the existing method, having strong universality, and being capable of being upgraded to realize analysis and processing of different frequency ranges, greatly reducing the workload of field verification in actual application, realizing automatic and personalized data analysis, and reducing labor costs.

[0019] Additional aspects and advantages of the present disclosure will be described in part in the description that follows, will become apparent from the description, or will be learned by practicing the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0020] The above and / or additional aspects and advantages of the present disclosure will become apparent and more readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings:

[0021] Figure 1 A flowchart of a method for processing drive test data provided by the embodiments of the present disclosure;

[0022] Figure 2A structural block diagram of a road test data processing method provided by an embodiment of the present disclosure is shown in FIG. 1.

[0023] Figure 3 A structural block diagram of a road test data processing method provided by an embodiment of the present disclosure is shown in FIG. 1.

[0024] Figure 4 A structural block diagram of a road test data processing device provided by an embodiment of the present disclosure is shown in FIG. 1.

[0025] Figure 5 A block diagram of an exemplary computer device suitable for implementing embodiments of the present disclosure is shown. DETAILED DESCRIPTION

[0026] Embodiments of the present disclosure are described in detail below with reference to the accompanying drawings, examples of which are shown in the drawings, wherein the same or similar notations represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present disclosure, and cannot be understood as a limitation of the present disclosure.

[0027] A road test data processing method, device, computer device and storage medium of an embodiment of the present disclosure are described below with reference to the accompanying drawings.

[0028] The existing planning station network interference analysis method mainly analyzes the test data of a frequency sweeper manually. Specifically, a frequency sweeper is used to lock a certain frequency band for field network test, and after the test is completed, the test data is manually analyzed. In the analysis process, not only the overall interference of the test area is analyzed, but also the interference of each station is analyzed manually. The present disclosure provides a road test data processing method and device, which solves the problem of automatic analysis of planning station interference, realizes one-key batch output of the interference of each planning station, and can output the average interference value graph and spectrum analysis graph of each frequency point of a single planning station.

[0029] It should be noted that the execution subject of the road test data processing method in the embodiments of the present disclosure is a road test data processing device, which can be realized by software and / or hardware, and the device can be configured in any electronic device.

[0030] Figure 1 A flowchart of a road test data processing method provided by the first embodiment of the present disclosure is shown in FIG. 1.

[0031] As shown in FIG. 1, the road test data processing method can include the following steps: Figure 1

[0032] In step 101, a road test data set and a first latitude and longitude of each planning point are obtained, wherein the road test data includes a second latitude and longitude of a sampling point, a sampling frequency and an interference value.​

[0033] The road test data set contains a plurality of road test data.

[0034] It should be noted that the sweep frequency test refers to collecting and testing wireless signals during network construction or optimization to collect network site communication data and signal strength information. The sweep frequency test data can be used to evaluate the coverage and quality of the network, and analyze the network capacity and interference. The road test data can include signal strength, signal-to-noise ratio, and bit error rate.

[0035] It should be noted that many planning stations need to be established during network construction, each planning station corresponding to a planning point and corresponding planning point information. Each planning point has a corresponding first latitude and longitude, i.e., the geographical location information of the planning point.

[0036] The road test data can be obtained by using a road test instrument (such as Anritsu, Keysight, etc.) to record the second latitude and longitude of the sampling point, the sampling frequency, and the interference value.

[0037] The second latitude and longitude can be the latitude and longitude corresponding to the sampling point, and the first latitude and longitude can be the latitude and longitude corresponding to the planning point.

[0038] Optionally, the interference values of each sampling point at a plurality of sampling frequencies can be counted based on a preset frequency granularity to obtain statistical data of each sampling point at each frequency granularity, and the accuracy of the second latitude and longitude of each sampling point can be reduced to obtain a third latitude and longitude of each sampling point.

[0039] The frequency granularity can be a frequency statistical unit.

[0040] As an example, in the present embodiment, the preset frequency granularity can be MHz units.

[0041] For example, if the original statistical data of the interference value is KHz units, it can be converted to MHz units for statistics.

[0042] For example, the frequency value of 1.8GHz is counted into the frequency band of 1MHz-2MHz.

[0043] The third latitude and longitude can be the second latitude and longitude after the latitude and longitude accuracy compression.

[0044] It should be noted that by reducing the precision of the second latitude and longitude of each sampling point to obtain the third latitude and longitude, location precision compression can be achieved. Since the statistical values ​​do not need to retain the exact location of the original sampling points, a certain range of deviation (e.g., 15 meters) in the location of a single sampling point is permissible. For example, the second latitude and longitude of the sampling points can be truncated to four decimal places to remove duplicates, thereby reducing the amount of data and allowing the determination of the third latitude and longitude of each sampling point after location compression.

[0045] Optionally, the level values ​​can be statistically analyzed according to preset segment thresholds, facilitating flexible data processing later. For example, the level values ​​can be statistically analyzed in segments such as -110dBm to -100dBm and -100dBm to -90dBm.

[0046] Step 102: Construct a binary tree of planning points based on each first latitude and longitude.

[0047] The planning point binary tree can have multiple nodes, and each node can correspond to a planning station.

[0048] Understandably, since the location information of the planned stations is deterministic two-dimensional data, and the data volume is much smaller than that of drive test data, a binary tree can be constructed based on the longitude and latitude of the planned stations.

[0049] It should be noted that, firstly, the two-dimensional spatial coordinates of each planning point can be calculated based on its corresponding latitude and longitude. This can be done using methods that convert latitude and longitude to planar coordinates. Then, based on these coordinates, a planning point can be selected as the root node of the binary tree. Alternatively, any two planning points can be selected as the starting nodes, and other planning points can be chosen as nodes to construct the binary tree.

[0050] For other planning points, their distances to the root node are calculated sequentially, and the planning points are inserted into the binary tree using the distance as the key. Specifically, starting from the root node, the distance between the current planning point and the node is compared. If the distance is less than the node's distance, the planning point is inserted into the node's left subtree; if the distance is greater than the node's distance, the planning point is inserted into the node's right subtree. If the distances are equal, the planning point can be inserted into either the left or right subtree as needed, until all planning points are inserted into the binary tree.

[0051] Through the above steps, a binary tree with planning points as nodes can be constructed, and the structure of the tree can be divided according to the distance and relationship between the planning points. In actual application, some properties of the binary tree, such as searching and traversal, can be used for searching and processing of the planning points. At the same time, it needs to be noted that in the process of constructing the binary tree, the number and distribution of the planning points need to be considered to ensure the balance and efficiency of the binary tree.

[0052] In the embodiments of the present disclosure, the binary tree can be a KDTree (k-dimensional tree), which is a tree constructed by recursively cutting a k-dimensional feature space with a median value. Each node is a hyperrectangle, and samples smaller than the node are divided into the left subtree, and samples larger than the node are divided into the right subtree. The binary tree has the highest efficiency when the dimension is less than 20, and is suitable for k-nearest neighbor search when the number of training instances is much larger than the spatial dimension. When the spatial dimension is close to the number of training instances, its efficiency will decrease rapidly, almost approaching linear scanning. Since the location information of the planning station is only 2-dimensional, it is difficult to model using the KDTree algorithm. The input of the KDTree model is the first latitude and longitude data of each planning point.

[0053] In step 103, based on the second latitude and longitude in each road test data, the binary tree is traversed to determine the set of sampling points associated with each planning point.

[0054] It should be noted that by creating a binary tree of the latitude and longitude of the planning station, all test data is filtered to retain sampling point statistical data that meets the distance threshold.

[0055] Among them, the interference analysis is to check whether each planning station exceeds the interference threshold one by one, and finally output the range of interference frequency bands.

[0056] Specifically, based on the second latitude and longitude in each road test data, the binary tree can be traversed, and the latitude and longitude of each road test data is matched with the data space of the planning station KDTree using binary tree search method during traversal, that is, starting from the top of the tree, according to whether the distance between the road test data and the node data meets the filtering threshold, the left / right branch of the next child node is selected. In this way, the purpose of reducing the search node and accelerating the data matching can be achieved. The filtering threshold here can be set to 1km according to the experience value of the coverage radius of the 700M station, that is, only selecting the road test sampling point data within the radius of 1km of the planning station as the sampling point set.

[0057] In step 104, based on a preset interference identification model, the road test data in each sampling point set is identified to determine the interference type of each sampling point.

[0058] The interference type is used to represent interference conditions, such as broadcast interference, non-broadcast interference, interference, and no interference, which are not limited here.

[0059] The interference identification model can be a model using a CART (Classification And Regression Tree) algorithm as an automatic interference type identification algorithm. It should be noted that, compared with other algorithms in the decision tree, the CART algorithm in the present application has the highest model credibility and better model specificity, which can greatly reduce the workload of field verification in practice.

[0060] Specifically, the road test data in each sample point set can be identified by the CART algorithm to determine the interference type of each sample point. For example, the frequency range of the interference frequency range and the downlink signal transmission frequency range of the broadcast transmission tower can be compared to identify whether it is broadcast interference according to the bandwidth overlap.

[0061] In the embodiments of the present disclosure, first, the road test data set and the first latitude and longitude of each planning point are obtained, wherein the second latitude and longitude of the sample point, the sampling frequency and the interference value are included in each road test data. Then, based on the first latitude and longitude, a planning point binary tree is constructed, and then based on the second latitude and longitude in each road test data, the binary tree is traversed to determine the sample point set associated with each planning point. Then, based on the preset interference identification model, the road test data in each sample point set is identified to determine the interference type of each sample point. Thus, the original test sampling road test data set can be automatically counted, analyzed and quickly evaluated, and the inter-network interference of each planning site position can be automatically output. The intelligent judgment whether it will be affected by the inter-network interference solves the problem of low automation and intelligence of the existing method, has strong universality, can be upgraded to realize analysis and processing of different frequency ranges, and can greatly reduce the workload of field verification in practice. It can greatly reduce the workload of field verification in practice, realize automatic and personalized data analysis, and reduce labor costs.

[0062] Figure 2 The flowchart of the road test data processing method provided by the second embodiment of the present disclosure is shown.

[0063] As shown in Figure 2 The road test data processing method can include the following steps:

[0064] Step 201, determine the total amount of data that the device can currently process and the data amount corresponding to each road test data.

[0065] It should be noted that when performing field tests, network tests and saving road test data can be performed continuously. For example, dense urban area traversal test requires 4-5 hours, and road test data can reach more than 20GB.

[0066] Specifically, the total amount of current data can be acquired first, for example, the total amount of original data can be calculated according to the data amount corresponding to each piece of test data. Due to the hardware performance limitation of the device, the total amount of data that can be processed by the device currently is within a certain range, so the total amount of data that can be processed by the device currently can be determined, which is related to the processing capability of the device.

[0067] In step 202, the first quantity of test data to be acquired is determined according to the total amount of data and the data amount corresponding to each piece of test data.

[0068] Optionally, the data amount corresponding to each piece of test data can be added, so that the total amount of collected test data can be determined, for example, 1G. If the total amount of data that can be processed by the device currently is 2G, it means that all the collected test data can be processed, and the first quantity is 1G. If the total amount of data that can be processed by the device currently is 0.5G, it means that the quantity of collected test data that can be processed is 0.5G, and the first quantity is 0.5G.

[0069] In step 203, a test data set containing the first quantity of test data is acquired from the collected test data.

[0070] It should be noted that the test data set containing the first quantity of test data can be acquired from the collected test data each time, that is, the processing can be performed in batches. In this way, the device can simultaneously load data and process data, that is, the next batch of data is loaded and the data that has been loaded currently is processed at the same time.

[0071] For example, if the total amount of data that can be processed by the device is 0.5G, and the collected test data is 2G, the collected test data can be divided into 4 batches, 0.5G of test data is loaded each time, and 0.5G of test data is processed each time, so that the data processing speed can be improved.

[0072] Further, after the test data set containing the first quantity of test data is acquired, the test data can be cleaned, and the data with missing longitude and latitude in the test data set can be discarded.

[0073] In step 204, a planning point binary tree is constructed based on each first longitude and latitude.

[0074] It should be noted that the specific implementation of step 204 can refer to the above embodiments, which will not be described here.

[0075] In step 205, the second longitude and latitude of the sampling point are matched with the first longitude and latitude corresponding to the root node in the binary tree, and the first target branch node of the first layer to be matched is determined according to the matching result.

[0076] The root node in the binary tree corresponds to a planning point.

[0077] It should be noted that, in the embodiments of the present disclosure, the process of determining the target branch node to be matched according to the matching result can include:

[0078] First, a first distance between a second latitude and longitude of the sampling point and a first latitude and longitude corresponding to the current first node is determined;

[0079] In the case where the first distance is greater than a distance threshold, a bifurcation dimension corresponding to the first node is determined;

[0080] The values of the second latitude and longitude and the first latitude and longitude corresponding to the first node on the bifurcation dimension are compared to determine the target branch node to be matched.

[0081] It should be noted that, if the second latitude and longitude of the sampling point is matched with the first latitude and longitude corresponding to the root node in the binary tree, the first node is the root node of the binary tree, and the target branch node to be matched is the first target branch node of the first layer to be matched.

[0082] It should be noted that the construction process of the conventional KDTree is to divide the data space by using the Minkowski distance, and in actual application, it is impossible to accurately map the Minkowski distance to the latitude and longitude distance. In the embodiments of the present disclosure, the KDTree algorithm is improved in the binary tree construction process, and the GIS latitude and longitude distance is used to replace the Minkowski distance in the original algorithm, so that the latitude and longitude data set can be more accurately matched.

[0083] Haversin (θ) = sin 2 (θ / 2)

[0084]

[0085] Wherein the second latitude and longitude of the sampling point is (φ1, λ1), the first latitude and longitude corresponding to the first node is (φ2, λ2), φ is the latitude, λ is the longitude, d = the first distance, and R is 6371km.

[0086] Wherein, the bifurcation dimension refers to bifurcation according to the latitude value or bifurcation according to the longitude value, which is not limited here.

[0087] Wherein, the distance threshold can be a threshold value of the distance between the sampling point and the first node, and in the embodiments of the present disclosure, the distance threshold can be 1km, which is not limited here.

[0088] For example, if the bifurcation dimension is bifurcation according to the latitude value, the latitude values of the second latitude and longitude and the first latitude and longitude can be compared, and then the branch to be bifurcated can be determined, and the child node corresponding to the branch is taken as the target branch node.

[0089] Optionally, a binary tree of planning points can be constructed based on the latitude and longitude information of the planning points, and a planning point can be selected as the root node.

[0090] It should be noted that, for a given sampling point, the distance between its second latitude and longitude and the first latitude and longitude corresponding to the root node is calculated, and the nearest planning point is selected as the first target branch node. For the first target branch node, the distance between the first latitude and longitude of all its child nodes and the second latitude and longitude of the sampling point is calculated, and the node whose distance is less than a distance threshold is selected as the second target branch node. This process continues until a target first latitude and longitude node whose distance to the second latitude and longitude of the sampling point is less than the distance threshold is found.

[0091] It is important to note that in practical applications, the number and distribution of planning points, as well as the time efficiency of matching, need to be considered. Optimization measures, such as pruning the binary tree of planning points and narrowing the matching range, can be used to improve matching efficiency and accuracy.

[0092] Step 206: Match the second latitude and longitude of the sampling point with the first latitude and longitude of the first target branch node, and determine the first target branch node of the second layer to be matched based on the matching result, until the target first latitude and longitude is obtained with a distance less than the distance threshold.

[0093] It should be noted that if the second latitude and longitude of the sampling point are matched with the first latitude and longitude of the first target branch node, then the first node is the first target branch node, and the target branch node to be matched is the first target branch node of the second layer to be matched. The specific implementation process can be referred to the above embodiment, and will not be repeated here.

[0094] The target first latitude and longitude can be the first latitude and longitude that is less than the distance threshold between the target and the second latitude and longitude, and the node corresponding to the target first latitude and longitude is taken as the target node.

[0095] Step 207: Determine the sampling point, which is a sampling point associated with the target node corresponding to the first latitude and longitude of the target.

[0096] It should be noted that a sampling point can be associated with the target node corresponding to the first latitude and longitude of the target, and become a sampling point of that node.

[0097] Step 208: Based on the preset interference identification model, identify the road test data in each sampling point set to determine the interference type of each sampling point.

[0098] It should be noted that the specific implementation of step 208 can be referred to the above embodiments, and will not be repeated here.

[0099] In this embodiment, the road test dataset and the first latitude and longitude of each planning point are first obtained. Each piece of road test data includes the second latitude and longitude of the sampling point, the sampling frequency, and the interference value. Then, a binary tree of planning points is constructed based on each first latitude and longitude. The second latitude and longitude of the sampling point are then matched with the first latitude and longitude corresponding to the root node in the binary tree. Based on the matching result, the first target branch node of the first layer to be matched is determined. The second latitude and longitude of the sampling point are then matched with the first latitude and longitude of the first target branch node. Based on the matching result, the first target branch node of the second layer to be matched is determined. This process continues until a target first latitude and longitude with a distance less than a distance threshold is obtained. The sampling point is then identified as a sampling point associated with the target node corresponding to the target first latitude and longitude. Finally, based on a preset interference identification model, the road test data in each sampling point set is identified to determine the interference type of each sampling point. Thus, the mechanism of batch and segmented reading can balance the contradiction between the hardware performance requirements and processing efficiency of the device, and achieve the purpose of batch processing test data.

[0100] Figure 3 This is a schematic flowchart of the road test data processing method provided in the second embodiment of this disclosure.

[0101] like Figure 3 As shown, the road test data processing method may include the following steps:

[0102] Step 301: Obtain the road test dataset and the first latitude and longitude of each planning point. Each road test dataset includes the second latitude and longitude of the sampling point, the sampling frequency, and the interference value.

[0103] Step 302: Construct a binary tree of planning points based on each first latitude and longitude.

[0104] Step 303: Based on the second latitude and longitude in each road test data, traverse the binary tree to determine the set of sampling points associated with each planning point.

[0105] It should be noted that the specific implementation of steps 301-303 can be referred to the above embodiments, and will not be repeated here.

[0106] Step 304: Obtain the sample dataset and the verification dataset, wherein the sample dataset and the verification dataset respectively include road test data of known interference types and the first attribute information of each road test data.

[0107] As one implementation method, the ratio of the sample dataset to the validation dataset can be 80% / 20%, without limitation. For example, 800 planned sites that have undergone manual analysis of interference can be used as samples, that is, 800 drive test data with known interference types, and then the model can be trained with 80% (training samples) + 20% (validation samples).

[0108] The first attribute information can be the type of device used to collect the road test data and the collection frequency corresponding to the road test data, which is not limited here.

[0109] Step 305: Determine the interference type as the root node, and determine the child nodes of each level of the interference type based on the Gini coefficient corresponding to each first attribute information.

[0110] Optionally, an interference source with a known interference type can be used as the root node.

[0111] The process of generating a decision tree is as follows:

[0112] First, the values ​​of the first attribute information are divided into regions using a heuristic approach. The j-th variable and its first attribute value s are selected as the splitting variable and splitting point, and two regions are defined:

[0113] R1(j,s)={x|x (j) ≤s} and R2(j,s)={x|x (j) >s}

[0114] Secondly, the optimal splitting variables and splitting points are selected using the following formula.

[0115]

[0116] Furthermore, for a fixed input variable j, the optimal split point s can be found:

[0117] and

[0118] Finally, iterate through all input variables to find the optimal splitting variable j, forming a pair (j, s). Divide the input space into two regions sequentially. Then, repeat the above partitioning process for each obtained region until the optimal result is obtained. This completes the construction of the CART decision tree.

[0119] In decision tree hierarchies, the Gini index is primarily used for judgment, and this index also determines the optimal cut point value for that feature. In classification, K represents the number of classes, and P is the probability that a sample point belongs to the k-th class. k Then the Gini index is:

[0120]

[0121] Step 306: Based on the decision tree composed of interference types and child nodes at each level, prune the decision tree according to the interference type identification results of the verification dataset to generate an interference identification model.

[0122] It should be noted that decision tree pruning mainly consists of pre-pruning and post-pruning. CART model pruning primarily employs post-pruning. This involves cutting off subtrees from the bottom up of the generated decision tree, making it smaller and improving its generalization ability, thus achieving accurate predictions of real-world data.

[0123] CART's post-pruning algorithm consists of two steps. First, it continuously prunes the decision tree T0 generated by the generation algorithm from its bottom until it reaches the root node of T0, forming a subtree sequence {T0, T1…T}. n Then, the subtree sequence is tested on an independent validation dataset using cross-validation to select the optimal subtree.

[0124] The pruning is based on the loss function of the decision tree:

[0125] C a (T)=C(T)+a|T|

[0126] In the loss function, T represents any subtree, C(T) is the prediction error on the training data (Gini index), |T| is the number of leaf nodes in the subtree (as a parameter), and C... a (T) represents the overall loss of subtree T when the parameter is a, where parameter a is the tradeoff between the fit of the training data and the model complexity.

[0127] Optionally, the decision tree can be pruned starting from the bottom node based on the attribute information corresponding to each node in the decision tree to obtain multiple first subtree sequences. Then, the interference type identification result of each first subtree sequence for each verification data can be determined. After that, the prediction error of each first subtree sequence can be determined based on the identification result and the known interference type of each verification data. Then, the target subtree sequence corresponding to each interference type can be determined based on the prediction error of each first subtree sequence. Finally, an interference identification model can be generated based on each target subtree sequence.

[0128] For example, if A is the root node, and A's first-level child nodes are B and C, B's child nodes are B1 and B2, and C is further divided into C1 and C2, then there are multiple first-subtree sequences such as B1-BA, B2-BA, C1-CA, and C2-CA. The first sub-sequence after pruning may be B1-A, C2-A, etc., without any restrictions here.

[0129] Optionally, you can first determine the reduction in the loss value of the second subtree column in which each node in the subtree sequence is located after pruning, then determine the target node based on the reduction value corresponding to each node, and then prune the target node to obtain the first subtree sequence.

[0130] Therefore, for each internal node t in T0, calculate the reduction in loss value g(t) for the second subtree column containing t:

[0131]

[0132] Specifically, this formula represents the degree to which the overall loss function is reduced after pruning, subtracting the minimum value of g(t) from T0. t The resulting subtree is designated as T1, and the smallest g(t) is designated as T. a T1 is the optimal subtree of the interval [a1, a2]. The pruned subtree sequence T0, T1…T… n The optimal decision tree T is selected through cross-validation. a .

[0133] Step 307: Based on the second attribute information corresponding to each road test data, traverse each node in the interference identification model in sequence to determine the interference type corresponding to each sampling point. The road test dataset also includes the second attribute information corresponding to each road test data.

[0134] In this embodiment, firstly, a road test dataset and the first latitude and longitude of each planning point are obtained. Each piece of road test data includes the second latitude and longitude of the sampling point, the sampling frequency, and the interference value. Then, a binary tree of planning points is constructed based on each first latitude and longitude. Subsequently, the binary tree is traversed based on the second latitude and longitude of each piece of road test data to determine the set of sampling points associated with each planning point, thereby obtaining a sample dataset and a verification dataset. The sample dataset and the verification dataset respectively include road test data with known interference types and the first attribute information of each piece of road test data. The interference type is determined as the root node, and the child nodes of each level of the interference type are determined based on the Gini coefficient corresponding to each first attribute information. Based on the decision tree composed of the interference type and the child nodes at each level, the decision tree is pruned according to the interference type identification result of the verification dataset to generate an interference identification model. Finally, based on the second attribute information corresponding to each piece of road test data, each node in the interference identification model is traversed sequentially to determine the corresponding interference type of each sampling point. The road test dataset also includes the second attribute information corresponding to each piece of road test data. Therefore, the original test sampling road test dataset can be automatically statistically analyzed and quickly evaluated, and the inter-network interference situation of each planned site location can be automatically output. It can intelligently determine whether it will be affected by inter-network interference, solve the problem of low automation and intelligence of existing methods, and has strong versatility. It can be upgraded to realize analysis and processing of different frequency ranges. In practical applications, it can greatly reduce the workload of on-site verification, realize automated and personalized data analysis, and reduce labor costs.

[0135] To implement the above embodiments, this disclosure also proposes a road test data processing device.

[0136] Figure 4 This is a structural block diagram of the road test data processing device provided in the third embodiment of this disclosure.

[0137] like Figure 4 As shown, the road test data processing device 400 may include:

[0138] The acquisition module 410 is used to acquire the road test dataset and the first latitude and longitude of each planning point, wherein each piece of road test data includes the second latitude and longitude of the sampling point, the sampling frequency, and the interference value;

[0139] Construction module 420 is used to construct a binary tree of planning points based on each of the first latitude and longitude.

[0140] The first determining module 430 is used to traverse the binary tree based on the second latitude and longitude in each of the road test data to determine the set of sampling points associated with each planning point;

[0141] The second determining module 440 is used to identify the road test data in each set of sampling points based on a preset interference identification model, so as to determine the interference type of each sampling point.

[0142] Optionally, the acquisition module is further configured to:

[0143] Based on a preset frequency granularity, the interference values ​​at multiple sampling frequencies corresponding to each sampling point are statistically analyzed to obtain the statistical data of each sampling point at each frequency granularity.

[0144] Reduce the precision of the second latitude and longitude of each sampling point to obtain the third latitude and longitude of each sampling point.

[0145] Optionally, the acquisition module is specifically used for:

[0146] Determine the total amount of data that the device can currently process and the amount of data corresponding to each of the aforementioned drive test data;

[0147] Based on the total amount of data and the amount of data corresponding to each piece of road test data, determine the first quantity of road test data to be acquired;

[0148] Obtain a road test dataset containing the first number of road test data from the collected road test data.

[0149] Optionally, the first determining module is specifically used for:

[0150] The second latitude and longitude of the sampling point are matched with the first latitude and longitude corresponding to the root node in the binary tree, and the first target branch node of the first layer to be matched is determined based on the matching result.

[0151] The second latitude and longitude of the sampling point are matched with the first latitude and longitude of the first target branch node, and the first target branch node of the second layer to be matched is determined according to the matching result, until the target first latitude and longitude is obtained with a distance less than the distance threshold.

[0152] The sampling point is determined to be a sampling point associated with the target node corresponding to the first latitude and longitude of the target.

[0153] Optionally, the first determining module is further configured to:

[0154] Determine the first distance between the second latitude and longitude of the sampling point and the first latitude and longitude corresponding to the current first node;

[0155] If the first distance is greater than the distance threshold, determine the fork dimension corresponding to the first node;

[0156] The values ​​of the second latitude and longitude and the first latitude and longitude corresponding to the first node are compared on the bifurcation dimension to determine the target branch node to be matched.

[0157] Optionally, the second determining module may also include:

[0158] The acquisition unit is used to acquire a sample dataset and a verification dataset, wherein the sample dataset and the verification dataset respectively include road test data of known interference types and first attribute information of each of the road test data;

[0159] The determining unit is configured to determine the interference type as the root node and, based on the Gini coefficient corresponding to each of the first attribute information, determine the child nodes of the interference type at each level.

[0160] The generation unit is used to prune the decision tree based on the interference type identification results of the verification dataset, which is composed of the interference type and the child nodes at each level, so as to generate the interference identification model.

[0161] Optionally, the generation unit includes:

[0162] The sub-unit is used to prune the decision tree starting from the bottom node based on the attribute information corresponding to each node in the decision tree, so as to obtain multiple first subtree sequences corresponding to the decision tree.

[0163] The first determining subunit is used to determine the interference type identification result of each of the first subtree sequences for each verification data;

[0164] The second determining subunit is used to determine the prediction error of each first subtree sequence based on the identification result and the known interference type of each of the verification data;

[0165] The third determining subunit is used to determine the target subtree sequence corresponding to each interference type based on the prediction error of each first subtree sequence;

[0166] A sub-unit is generated to generate the interference identification model based on each target subtree sequence.

[0167] Optionally, the acquisition subunit is specifically used for:

[0168] Determine the reduction in the loss value of the second subtree column to which each node in the subtree sequence belongs after it has been pruned;

[0169] The target node is determined based on the reduction value corresponding to each of the nodes;

[0170] The target node is pruned to obtain the first subtree sequence.

[0171] Optional, the second determining module is specifically used for:

[0172] Based on the second attribute information corresponding to each of the road test data, each node in the interference identification model is traversed sequentially to determine the interference type corresponding to each sampling point. The road test dataset also includes the second attribute information corresponding to each of the road test data.

[0173] In this embodiment, a road test dataset and the first latitude and longitude of each planning point are first obtained. Each piece of road test data includes the second latitude and longitude of the sampling point, the sampling frequency, and the interference value. Then, a binary tree of planning points is constructed based on each of the first latitude and longitude. Next, the binary tree is traversed based on the second latitude and longitude in each piece of road test data to determine the set of sampling points associated with each planning point. Then, based on a preset interference identification model, the road test data in each set of sampling points is identified to determine the interference type of each sampling point. Therefore, the original test sampled road test dataset can be automatically statistically analyzed and quickly evaluated, automatically outputting the inter-network interference situation of each planning station location, and intelligently determining whether it will be subject to inter-network interference. This solves the problem of low automation and intelligence in existing methods, has strong versatility, and can be upgraded to analyze and process different frequency ranges. In practical applications, it can greatly reduce the workload of on-site verification, realize automated and personalized data analysis, and reduce labor costs.

[0174] To implement the above embodiments, this disclosure also proposes a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the road test data processing method proposed in the foregoing embodiments of this disclosure.

[0175] To implement the above embodiments, this disclosure also proposes a non-transitory computer-readable storage medium storing a computer program, which, when executed by a processor, implements the road test data processing method proposed in the foregoing embodiments of this disclosure.

[0176] To implement the above embodiments, this disclosure also proposes a computer program product that, when the instruction processor in the computer program product is executed, performs the road test data processing method as proposed in the foregoing embodiments of this disclosure.

[0177] Figure 4 A block diagram of an exemplary computer device suitable for implementing embodiments of the present disclosure is shown. Figure 4 The computer device 12 shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments disclosed herein.

[0178] like Figure 4As shown, the computer device 12 is represented in the form of a general-purpose computing device. The components of the computer device 12 may include, but are not limited to: one or more processors or processing units 16, system memory 28, and a bus 18 connecting different system components (including system memory 28 and processing unit 16).

[0179] Bus 18 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. Examples of these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0180] Computer device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by computer device 12, including volatile and non-volatile media, removable and non-removable media.

[0181] Memory 28 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM) 30 and / or cache memory 32. Computer device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be used to read and write non-removable, non-volatile magnetic media (…). Figure 4 Not shown; usually referred to as a "hard drive"). Although Figure 4Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk") and an optical disc drive for reading and writing to a removable non-volatile optical disc (e.g., a compact disc read-only memory (CD-ROM), a digital video disc read-only memory (DVD-ROM), or other optical media) may be provided. In these cases, each drive may be connected to bus 18 via one or more data media interfaces. Memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of this disclosure.

[0182] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 42 typically perform the functions and / or methods described in the embodiments of this disclosure.

[0183] Computer device 12 can also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), and with one or more devices that enable a user to interact with computer device 12, and / or with any device that enables computer device 12 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed via input / output (I / O) interface 22. Furthermore, computer device 12 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 20. As shown, network adapter 20 communicates with other modules of computer device 12 via bus 18. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with computer device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0184] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, such as implementing the methods mentioned in the foregoing embodiments.

[0185] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0186] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this disclosure, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0187] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of this disclosure includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of this disclosure pertain.

[0188] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0189] It should be understood that various parts of this disclosure can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0190] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0191] Furthermore, the functional units in the various embodiments of this disclosure can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0192] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present disclosure have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present disclosure. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present disclosure.

Claims

1. A method for processing road test data, characterized in that, include: Obtain the road test dataset and the first latitude and longitude of each planning point. Each road test dataset includes the second latitude and longitude of the sampling point, the sampling frequency, and the interference value. Construct a binary tree of planning points based on each of the first latitude and longitude coordinates; Based on the second latitude and longitude in each of the road test data, the binary tree is traversed to determine the set of sampling points associated with each planning point; Based on a preset interference identification model, the road test data in each set of sampling points are identified to determine the type of interference for each sampling point; The step of traversing the binary tree based on the second latitude and longitude of each of the road test data to determine the set of sampling points associated with each planning point includes: The second latitude and longitude of the sampling point are matched with the first latitude and longitude corresponding to the root node in the binary tree, and the first target branch node of the first layer to be matched is determined based on the matching result. The second latitude and longitude of the sampling point are matched with the first latitude and longitude of the first target branch node, and the first target branch node of the second layer to be matched is determined according to the matching result, until the target first latitude and longitude is obtained with a distance less than the distance threshold. The sampling point is determined to be a sampling point associated with the target node corresponding to the first latitude and longitude of the target; The process of determining the target branch node to be matched based on the matching results includes: Determine the first distance between the second latitude and longitude of the sampling point and the first latitude and longitude corresponding to the current first node; If the first distance is greater than the distance threshold, determine the fork dimension corresponding to the first node; The values ​​of the second latitude and longitude and the first latitude and longitude corresponding to the first node are compared on the bifurcation dimension to determine the target branch node to be matched.

2. The method as described in claim 1, characterized in that, After obtaining the road test dataset, the following is also included: Based on a preset frequency granularity, the interference values ​​at multiple sampling frequencies corresponding to each sampling point are statistically analyzed to obtain the statistical data of each sampling point at each frequency granularity. Reduce the precision of the second latitude and longitude of each sampling point to obtain the third latitude and longitude of each sampling point.

3. The method as described in claim 1, characterized in that, The acquisition of the road test dataset includes: Determine the total amount of data that the device can currently process and the amount of data corresponding to each of the aforementioned drive test data; Based on the total amount of data and the amount of data corresponding to each piece of road test data, determine the first quantity of road test data to be acquired; Obtain a road test dataset containing the first number of road test data from the collected road test data.

4. The method as described in claim 1, characterized in that, Before identifying the road test data in each set of sampling points based on the preset interference identification model, the method further includes: Obtain a sample dataset and a verification dataset, wherein the sample dataset and the verification dataset respectively include road test data of known interference types and first attribute information of each of the road test data; The interference type is determined as the root node, and the child nodes of each level of the interference type are determined based on the Gini coefficient corresponding to each of the first attribute information. Based on the decision tree composed of the interference types and child nodes at each level, the decision tree is pruned according to the interference type identification results of the verification dataset to generate the interference identification model.

5. The method as described in claim 4, characterized in that, The step of pruning the decision tree based on the interference type identification results of the verification dataset, which is composed of the interference type and the child nodes at each level, includes: Based on the attribute information corresponding to each node in the decision tree, the decision tree is pruned starting from the bottom node to obtain multiple first subtree sequences corresponding to the decision tree. Determine the interference type identification result for each of the first subtree sequences for each verification data; Based on the identification results and the known interference type of each of the verification data, the prediction error of each first subtree sequence is determined; Based on the prediction error of each first subtree sequence, determine the target subtree sequence corresponding to each interference type; The interference identification model is generated based on the sequence of each target subtree.

6. The method as described in claim 5, characterized in that, The step of pruning the decision tree includes: Determine the reduction in the loss value of the second subtree column to which each node in the subtree sequence belongs after it has been pruned; The target node is determined based on the reduction value corresponding to each of the nodes; The target node is pruned to obtain the first subtree sequence.

7. The method as described in claim 6, characterized in that, The step of identifying the drive test data in each of the sampling point sets to determine the interference type of each sampling point includes: Based on the second attribute information corresponding to each of the road test data, each node in the interference identification model is traversed sequentially to determine the interference type corresponding to each sampling point. The road test dataset also includes the second attribute information corresponding to each of the road test data.

8. A road test data processing device, characterized in that, include: The acquisition module is used to acquire the road test dataset and the first latitude and longitude of each planning point. Each road test data includes the second latitude and longitude of the sampling point, the sampling frequency, and the interference value. A construction module is used to construct a binary tree of planning points based on each of the first latitude and longitude coordinates; The first determining module is used to traverse the binary tree based on the second latitude and longitude in each of the road test data to determine the set of sampling points associated with each planning point; The second determining module is used to identify the road test data in each set of sampling points based on a preset interference identification model, so as to determine the interference type of each sampling point; The first determining module is specifically used for: The second latitude and longitude of the sampling point are matched with the first latitude and longitude corresponding to the root node in the binary tree, and the first target branch node of the first layer to be matched is determined based on the matching result. The second latitude and longitude of the sampling point are matched with the first latitude and longitude of the first target branch node, and the first target branch node of the second layer to be matched is determined according to the matching result, until the target first latitude and longitude is obtained with a distance less than the distance threshold. The sampling point is determined to be a sampling point associated with the target node corresponding to the first latitude and longitude of the target; The first determining module is further configured to: Determine the first distance between the second latitude and longitude of the sampling point and the first latitude and longitude corresponding to the current first node; If the first distance is greater than the distance threshold, determine the fork dimension corresponding to the first node; The values ​​of the second latitude and longitude and the first latitude and longitude corresponding to the first node are compared on the bifurcation dimension to determine the target branch node to be matched.

9. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the drive test data processing method according to any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the road test data processing method as described in any one of claims 1-7.

11. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the road test data processing method as described in any one of claims 1-7.

Images