Interference prediction method, device, apparatus and computer readable storage medium

Abstract

The embodiment of the present application relates to the technical field of communication, and discloses a kind of interference prediction methods, the method comprises: collection space-time interference data set, the space-time interference data set includes time information, spatial information, network side information, external interference information;The space-time interference data set is input into interference prediction model, and node feature matrix and inter-node correlation matrix are obtained;The node feature matrix is used to represent the characteristics of the spatial information and the network side information;The inter-node correlation matrix is used to represent the interference characteristics of the external interference information on the network side information;According to the node feature matrix and the inter-node correlation matrix, determine the target node and the corresponding target interference source node that exist interference.Through the above mode, the embodiment of the present application realizes the effective prediction of interference.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, specifically to an interference prediction method, apparatus, device, and computer-readable storage medium. Background Technology

[0002] Currently, existing technologies mainly rely on cell lock screen testing, basic data analysis, and manual on-site verification to optimize 4 / 5G interference. During the implementation of this invention, the inventors of this application discovered the following drawbacks of existing technologies: 1. Cell lock screen testing: Only a small range of specific test cells can be manually selected, which is time-consuming, inefficient, and difficult to accurately reflect the overall network situation. 2. Basic data analysis combined with manual on-site verification: Traditional network interference performance index analysis methods do not consider the changes in communication network node characteristics and the influence of the temporal and spatial dimensions of network structure; secondly, traditional data analysis mainly focuses on identifying the root causes of interference based on the interference waveform morphology, resulting in poor accuracy; furthermore, due to the increasingly complex 4 / 5G network structure and the spatiotemporal evolution and dynamism of communication networks, traditional known interference classifications and root cause distributions can no longer meet the needs of the current network. Summary of the Invention

[0003] In view of the above problems, embodiments of the present invention provide an interference prediction method, apparatus, device and computer-readable storage medium to solve the technical problem of poor interference identification and processing efficiency in the prior art.

[0004] According to one aspect of the present invention, an interference prediction method is provided, the method comprising:

[0005] Collect a spatiotemporal interference dataset, which includes time information, spatial information, network-side information, and external interference information;

[0006] The spatiotemporal interference dataset is input into the interference prediction model to obtain the node feature matrix and the inter-node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the inter-node correlation matrix is ​​used to characterize the interference features of the external interference information on the network-side information.

[0007] The target nodes with interference and the corresponding target interference source nodes are determined based on the node feature matrix and the inter-node correlation matrix.

[0008] In one optional approach, before inputting the spatiotemporal interference dataset into the interference prediction model to obtain the node feature matrix and the inter-node correlation matrix, the method includes: acquiring a training dataset; the training dataset includes historically collected time information, spatial information, network-side information, and external interference information; and inputting the training dataset into a preset prediction model for training to obtain the trained interference prediction model.

[0009] In one optional approach, the external interference information includes interference event information and meteorological information; the interference prediction model includes a convolutional neural network, a gating unit, and a node spatiotemporal dynamic change update unit; the step of inputting the training dataset into a preset prediction model for training to obtain the trained interference prediction model includes: inputting historically collected interference event information and meteorological information into the convolutional neural network for learning to obtain an external interference factor vector; inputting the external interference factor vector into the gating unit to obtain an external feature matrix; and inputting spatial information, network-side information, and the external feature matrix into the node spatiotemporal dynamic change update unit for training to obtain the interference prediction model.

[0010] In one optional approach, the spatiotemporal interference dataset is collected, which includes time information, spatial information, network-side information, and external interference information, including: collected engineering parameter data, OMC-R measurement data, MDT data, soft-collected data, XDR data, POI data, meteorological data, and interference event data; the engineering parameter data, the OMC-R measurement data, the MDT data, the soft-collected data, the XDR data, the POI data, the meteorological data, and the interference event data are processed to obtain the collected spatiotemporal interference dataset.

[0011] In one optional approach, after determining the target node with interference and the corresponding target interference source node based on the node feature matrix and the inter-node correlation matrix, the process includes: determining an interference optimization strategy based on the target node with interference and the target interference source node; and executing the interference optimization strategy.

[0012] In one optional approach, the number of target nodes is multiple; determining the interference optimization strategy based on the target nodes and the target interference source nodes includes: determining the priority of the target nodes; and determining the interference optimization strategy based on the priority of the target nodes.

[0013] According to another aspect of the present invention, an interference prediction device is provided, comprising:

[0014] The acquisition module is used to acquire spatiotemporal interference datasets, which include time information, spatial information, network-side information, and external interference information.

[0015] The prediction module is used to input the spatiotemporal interference dataset into the interference prediction model to obtain a node feature matrix and a node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the node correlation matrix is ​​used to characterize the interference features of the external interference information on the network-side information.

[0016] The determination module is used to determine the target node with interference and the corresponding target interference source node based on the node feature matrix and the inter-node correlation matrix.

[0017] In an optional embodiment, the apparatus further includes: a strategy module for determining an interference optimization strategy based on the target node of the interference and the target interference source node; and an execution module for executing the interference optimization strategy.

[0018] According to another aspect of the present invention, a computing device is provided, including: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other through the communication bus;

[0019] The memory is used to store at least one executable instruction that causes the processor to perform the operation of the interference prediction method.

[0020] According to another aspect of the present invention, a computer-readable storage medium is provided, the storage medium storing at least one executable instruction, which, when executed on a computing device, causes the computing device to perform the operation of the interference prediction method.

[0021] This invention collects a spatiotemporal interference dataset; inputs the spatiotemporal interference dataset into an interference prediction model to obtain a node feature matrix and a node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the node correlation matrix is ​​used to characterize the interference characteristics of the external interference information on the network-side information; based on the node feature matrix and the node correlation matrix, the target nodes with interference and the corresponding target interference source nodes are determined. It can automatically capture the dynamic features and spatiotemporal dependencies of each node in the communication network and adaptively update the graph network structure. In addition, through a gating mechanism, important external influencing factors such as external interference events and meteorological events that may cause interference to the communication network are incorporated into the model, thereby achieving effective prediction and prevention of complex interference in 4 / 5G communication networks.

[0022] The above description is merely an overview of the technical solutions of the embodiments of the present invention. In order to better understand the technical means of the embodiments of the present invention and to implement them in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the embodiments of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0023] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0024] Figure 1 A flowchart illustrating the interference prediction method provided in an embodiment of the present invention is shown;

[0025] Figure 2 A schematic diagram of the interference prediction model provided in an embodiment of the present invention is shown;

[0026] Figure 3 A schematic diagram of the interference prediction device provided in an embodiment of the present invention is shown;

[0027] Figure 4 A schematic diagram of the structure of a computing device provided in an embodiment of the present invention is shown. Detailed Implementation

[0028] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0029] Figure 1 A flowchart of an interference prediction method provided by an embodiment of the present invention is shown. This method is executed by a computing device. The computing device can be a computer, a smart terminal, a network-side device, etc. Figure 1 As shown, the method includes the following steps:

[0030] Step 110: Collect a spatiotemporal interference dataset, which includes time information, spatial information, network-side information, and external interference information.

[0031] External interference information includes interference event information and meteorological information.

[0032] In this embodiment of the invention, source data can be collected, including engineering parameter data, OMC-R measurement data, MDT data, soft-collection data, XDR data, POI data, meteorological data, and interference event data. The engineering parameter data, OMC-R measurement data, MDT data, soft-collection data, XDR data, POI data, meteorological data, and interference event data are then processed to obtain the spatiotemporal interference dataset.

[0033] Specifically, the process of collecting source data is as follows:

[0034] The main fields to be collected in the engineering parameter data include: province, city, region, frequency band, scene type, scene name, base station name, cell name, eNBId, CellId, azimuth, ECGI, antenna height, site longitude, and site latitude.

[0035] Obtain OMC-R measurement data. The main fields to be collected in the OMC-R measurement data include: timestamp, eNBId, CellId, MmeUeS1apId, Earfcn, SubFrameNbr, PRBNbr, PRB granularity eNodeB received interference power, eNodeB received interference power, eNodeB antenna angle of arrival, uplink packet loss rate, downlink packet loss rate, reference signal received power, etc.

[0036] Obtain MDT data, the main fields to be collected in the MDT data include: timestamp, MME UE S1AP ID, IMEI, IMSI, serving cell frequency, serving cell RSRP, serving cell eNodeB antenna angle of arrival, UE longitude, UE latitude, PDCP layer uplink data traffic, PDCP layer downlink data traffic, uplink PDCP packet loss count, downlink PDCP packet loss count, etc.

[0037] Acquire software-collected data. The main fields to be collected in the software-collected data include: timestamp, MME UE S1APID, eNB ID, Sector ID, uplink packet loss rate, downlink packet loss rate, eNB received interference power, angle of arrival (AoA), number of uplink service bytes on the air interface, number of simultaneous online users, etc.

[0038] Acquire user XDR data. The user XDR data mainly collects data streams from the S1-U / S11 / Gn ports. The fields to be collected mainly include: Local City, Procedure Start Time, Cell ID, IMSI, IMEI, Procedure End Time, longitude, latitude, Height, UL Data, DL Data, etc.

[0039] To obtain POI data, the main fields to be collected include: city, tag, POI id, name, query, lat, lng, industry_type, price_section, shop_hours, discount, groupon, etc.

[0040] Meteorological information is obtained from the meteorological bureau website or meteorological monitoring equipment. The main information to be collected in the meteorological information includes: timestamp, station number, longitude, latitude, air pressure, water vapor pressure, temperature, waveguide intensity, waveguide thickness, etc.

[0041] Obtaining interference event information mainly includes: event time, event location information, and event type information. For example, this includes, but is not limited to, interference events such as fake base station information, security equipment information, and jammer information.

[0042] After obtaining the source data, data cleaning is performed to unify the format of fields with the same physical meaning in various information entries, such as standardizing names and timestamps. The data cleaning process includes latitude and longitude coordinate system conversion and removal of outliers and missing values.

[0043] After data cleaning and standardized format processing, standard data was obtained. The OMC-R measurement data, MDT data, and soft-collection data in the standard data were associated according to their timestamp and MME UE S1AP ID fields to obtain intermediate dataset 1. This dataset 1 was then associated with the XDR data using IMSI or IMEI to form intermediate dataset 2. Intermediate dataset 2 was then associated with engineering parameter data, POI data, and meteorological data using timestamps, eNBId, CellId, and latitude / longitude matching to form a spatiotemporal interference dataset including time information, spatial information (latitude / longitude, POI information, etc.), network-side information (site information, interference information, capacity information, service quality information, etc.), and external factor information (meteorological information, interference-related event information, POI information, etc.).

[0044] Step 120: Input the spatiotemporal interference dataset into the interference prediction model to obtain the node feature matrix and the correlation matrix between nodes.

[0045] The node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the inter-node correlation matrix is ​​used to characterize the interference features of the external interference information on the network-side information.

[0046] After obtaining the spatiotemporal interference dataset with correlations, the spatiotemporal interference dataset is input into the interference prediction model to obtain the node feature matrix, the external feature matrix, and the correlation matrix between nodes.

[0047] In this embodiment of the invention, before inputting the spatiotemporal interference dataset into the interference prediction model to obtain the node feature matrix and the inter-node correlation matrix, the interference prediction model needs to be trained in advance. The specific training process is as follows: A training dataset is obtained; the training dataset includes historically collected temporal information, spatial information, network-side information, and external interference information; the training dataset is input into a preset prediction model for training, resulting in the trained interference prediction model.

[0048] Specifically, such as Figure 2The diagram illustrates the structure of an interference prediction model based on a dynamic spatiotemporal graph network in an embodiment of the present invention. The interference prediction model includes a convolutional neural network, a gating unit, and a node spatiotemporal dynamic change update unit. The node spatiotemporal dynamic change update unit is a dynamic spatiotemporal graph network (RGEU), where the graph convolutional neural network can model the spatial dependencies of nodes in a graph using a predefined Laplacian matrix based on node distance. The specific training process of the interference prediction model is as follows: First, a spatiotemporal interference fusion dataset with a relatively long historical time period is acquired. This data is then divided into selectable time granularities as needed. For example, when the collection period is less than the specified time granularity, aggregation can be performed using averaging or extreme values. The dataset after dividing the time granularities is then divided into a training set and a test set according to a selectable ratio. The interference prediction model based on the dynamic spatiotemporal graph network is further trained based on the acquired sample set. For example, the ratio between the training set and the test set can be 8:2. Historically collected interference event information and meteorological information are input into the convolutional neural network for learning, resulting in an external interference factor vector, specifically including a meteorological interference factor vector e. ct and event interference factor vector e pt .

[0049] Then, the external interference factor vector is input into the gating unit to obtain the external feature matrix, specifically the meteorological interference factor vector ect and the event interference factor vector e. pt Input the gated unit to obtain the representation of the external feature matrix:

[0050] G t =σ[W ec ·e ct +W ep ·e pt ]

[0051] Among them, e ct This represents the vector of meteorological interference factors at time t, which is the output of the CNN network after the meteorological information is input and learned; e pt W represents the vector of event interference factors at time t, which is the output of the CNN network after the interference event information is input and learned. ec and W ep σ represents the parameter matrix of the CNN network; σ(·) represents the sigmoid function.

[0052] Finally, the spatial information, network-side information, and the external feature matrix are input into the node spatiotemporal dynamic change update unit for training to obtain the interference prediction model. The node spatiotemporal dynamic change update unit can be represented as follows:

[0053]

[0054]

[0055]

[0056]

[0057]

[0058] Where t represents time information (such as timestamps in the data of each dimension in the spatiotemporal interference dataset), l represents the l-th layer graph convolution; the node spatiotemporal dynamic change update unit can be a graph convolutional neural network; Represents the node feature matrix output by the convolution of the l-th layer graph at time t; I∈R N×N Represents an identity matrix; N represents the variable number of network nodes, which can be base stations, grids, etc.; E∈R N×C To initialize a randomly learnable C-dimensional network node embedding matrix, C < N; ReLU(·) is the activation function, and softmax(·) is the normalized exponential function; W hz W wz W hr W wr W hw W ww B z B r B w This is the parameter matrix for the science department.

[0059] In this embodiment of the invention, the interference characteristics of the external interference information on the network-side information are characterized by the inter-node correlation matrix. This represents the dynamic correlation between nodes for external disturbance information at time t.

[0060] In this embodiment of the invention, historical interference event information and meteorological information are input into the CNN network for training. At the same time, historical spatial information, network-side information, and the external feature matrix are input into the node spatiotemporal dynamic change update unit for training. This results in the trained CNN network and the parameter matrices of the node spatiotemporal dynamic change update unit, ultimately leading to the trained interference prediction model.

[0061] In this embodiment of the invention, after performing interference prediction based on the trained model, the network interference prediction result at time t is obtained. This interference prediction result includes the node feature matrix H. t Node correlation matrix External feature G t Etc. predict the output.

[0062] Step 130: Determine the target node with interference and the corresponding target interference source node based on the node feature matrix and the inter-node correlation matrix.

[0063] Among them, the node feature matrix H t The threshold eigenvector is β, β∈R 1×F β can be set, H t Each row represents the F-dimensional feature of the corresponding node; the correlation matrix between nodes. The threshold feature vector is α, where the value of α can be set according to the specific scenario. Therefore, after obtaining the node feature matrix H... t Node correlation matrix Then, the following steps are used to determine the target node with interference and the corresponding target interference source node:

[0064] Step 1301: When the predicted output node feature matrix H t If the network-side information contained in the h-th row reaches the threshold feature vector β value, then it is determined that there is interference in the prediction of node h at time t;

[0065] Step 1302: Analyze the inter-node correlation matrix of the predicted output. When the elements of the correlation matrix When the value is greater than the threshold α, the corresponding node i is considered to be the main source of interference for the corresponding node j.

[0066] Step 1303: Combining steps 3031 and 3032, identify the target nodes that are interfering and the corresponding target interference source nodes, and output the corresponding node features H. t .

[0067] Step 1304: Determine an interference optimization strategy based on the target node and the target interference source node, and execute the interference optimization strategy. Specifically, for each target node h (h∈[1,N]) with interference, calculate its interference elimination priority based on the POI information and network-side information contained in the feature representation of the target node h, select the target node h with the higher interference elimination priority, formulate an interference optimization strategy and execute it, and start the timer.

[0068] Step 1305: When the timer is completed, end the interference optimization strategy and re-enter the interference prediction step based on the dynamic spatiotemporal graph network.

[0069] This invention collects a spatiotemporal interference dataset; inputs the spatiotemporal interference dataset into an interference prediction model to obtain a node feature matrix and a node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the node correlation matrix is ​​used to characterize the interference characteristics of the external interference information on the network-side information; based on the node feature matrix and the node correlation matrix, the target nodes with interference and the corresponding target interference source nodes are determined. It can automatically capture the dynamic features and spatiotemporal dependencies of each node in the communication network and adaptively update the graph network structure. In addition, through a gating mechanism, important external influencing factors such as external interference events and meteorological events that may cause interference to the communication network are incorporated into the model, thereby achieving effective prediction and prevention of complex interference in 4 / 5G communication networks.

[0070] Figure 3 A schematic diagram of the interference prediction device provided in an embodiment of the present invention is shown. Figure 3 As shown, the device 200 includes: a data acquisition module 210, a prediction module 220, and a determination module 230.

[0071] The acquisition module 210 is used to acquire a spatiotemporal interference dataset, which includes time information, spatial information, network-side information, and external interference information.

[0072] The prediction module 220 is used to input the spatiotemporal interference dataset into the interference prediction model to obtain a node feature matrix and a node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the node correlation matrix is ​​used to characterize the interference features of the external interference information on the network-side information.

[0073] The determination module 230 is used to determine the target node with interference and the corresponding target interference source node based on the node feature matrix and the inter-node correlation matrix.

[0074] In one optional approach, before inputting the spatiotemporal interference dataset into the interference prediction model to obtain the node feature matrix and the inter-node correlation matrix, the method includes: acquiring a training dataset; the training dataset includes historically collected time information, spatial information, network-side information, and external interference information; and inputting the training dataset into a preset prediction model for training to obtain the trained interference prediction model.

[0075] In one optional approach, the external interference information includes interference event information and meteorological information; the interference prediction model includes a convolutional neural network, a gating unit, and a node spatiotemporal dynamic change update unit; the step of inputting the training dataset into a preset prediction model for training to obtain the trained interference prediction model includes: inputting historically collected interference event information and meteorological information into the convolutional neural network for learning to obtain an external interference factor vector; inputting the external interference factor vector into the gating unit to obtain an external feature matrix; and inputting spatial information, network-side information, and the external feature matrix into the node spatiotemporal dynamic change update unit for training to obtain the interference prediction model.

[0076] In one optional approach, the spatiotemporal interference dataset is collected, which includes time information, spatial information, network-side information, and external interference information, including: collected engineering parameter data, OMC-R measurement data, MDT data, soft-collected data, XDR data, POI data, meteorological data, and interference event data; the engineering parameter data, the OMC-R measurement data, the MDT data, the soft-collected data, the XDR data, the POI data, the meteorological data, and the interference event data are processed to obtain the collected spatiotemporal interference dataset.

[0077] In one optional approach, after determining the target node with interference and the corresponding target interference source node based on the node feature matrix and the inter-node correlation matrix, the process includes: determining an interference optimization strategy based on the target node with interference and the target interference source node; and executing the interference optimization strategy.

[0078] In one optional approach, the number of target nodes is multiple; determining the interference optimization strategy based on the target nodes and the target interference source nodes includes: determining the priority of the target nodes; and determining the interference optimization strategy based on the priority of the target nodes.

[0079] The specific working process of the interference prediction device in this embodiment of the invention is largely the same as the specific method steps in the above method embodiment, and will not be repeated here.

[0080] This invention collects a spatiotemporal interference dataset; inputs the spatiotemporal interference dataset into an interference prediction model to obtain a node feature matrix and a node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the node correlation matrix is ​​used to characterize the interference characteristics of the external interference information on the network-side information; based on the node feature matrix and the node correlation matrix, the target nodes with interference and the corresponding target interference source nodes are determined. It can automatically capture the dynamic features and spatiotemporal dependencies of each node in the communication network and adaptively update the graph network structure. In addition, through a gating mechanism, important external influencing factors such as external interference events and meteorological events that may cause interference to the communication network are incorporated into the model, thereby achieving effective prediction and prevention of complex interference in 4 / 5G communication networks.

[0081] Figure 4 The diagram shows a structural schematic of a computing device provided in an embodiment of the present invention. The specific embodiments of the present invention do not limit the specific implementation of the computing device.

[0082] like Figure 4 As shown, the computing device may include: a processor 302, a communications interface 304, a memory 306, and a communications bus 308.

[0083] The processor 302, communication interface 304, and memory 406 communicate with each other via communication bus 308. Communication interface 304 is used to communicate with other network elements, such as clients or other servers. The processor 302 executes program 310, specifically performing the relevant steps described above in the interference prediction method embodiment.

[0084] Specifically, program 310 may include program code, which includes computer-executable instructions.

[0085] Processor 302 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The computing device may include one or more processors of the same type, such as one or more CPUs; or it may include processors of different types, such as one or more CPUs and one or more ASICs.

[0086] Memory 306 is used to store program 310. Memory 306 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0087] Specifically, program 310 can be called by processor 302 to cause the computing device to perform the following operations:

[0088] Collect a spatiotemporal interference dataset, which includes time information, spatial information, network-side information, and external interference information;

[0089] The spatiotemporal interference dataset is input into the interference prediction model to obtain the node feature matrix and the inter-node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the inter-node correlation matrix is ​​used to characterize the interference features of the external interference information on the network-side information.

[0090] The target nodes with interference and the corresponding target interference source nodes are determined based on the node feature matrix and the inter-node correlation matrix.

[0091] In one optional approach, before inputting the spatiotemporal interference dataset into the interference prediction model to obtain the node feature matrix and the inter-node correlation matrix, the method includes: acquiring a training dataset; the training dataset includes historically collected time information, spatial information, network-side information, and external interference information; and inputting the training dataset into a preset prediction model for training to obtain the trained interference prediction model.

[0092] In one optional approach, the external interference information includes interference event information and meteorological information; the interference prediction model includes a convolutional neural network, a gating unit, and a node spatiotemporal dynamic change update unit; the step of inputting the training dataset into a preset prediction model for training to obtain the trained interference prediction model includes: inputting historically collected interference event information and meteorological information into the convolutional neural network for learning to obtain an external interference factor vector; inputting the external interference factor vector into the gating unit to obtain an external feature matrix; and inputting spatial information, network-side information, and the external feature matrix into the node spatiotemporal dynamic change update unit for training to obtain the interference prediction model.

[0093] In one optional approach, the spatiotemporal interference dataset is collected, which includes time information, spatial information, network-side information, and external interference information, including: collected engineering parameter data, OMC-R measurement data, MDT data, soft-collected data, XDR data, POI data, meteorological data, and interference event data; the engineering parameter data, the OMC-R measurement data, the MDT data, the soft-collected data, the XDR data, the POI data, the meteorological data, and the interference event data are processed to obtain the collected spatiotemporal interference dataset.

[0094] In one optional approach, after determining the target node with interference and the corresponding target interference source node based on the node feature matrix and the inter-node correlation matrix, the process includes: determining an interference optimization strategy based on the target node with interference and the target interference source node; and executing the interference optimization strategy.

[0095] In one optional approach, the number of target nodes is multiple; determining the interference optimization strategy based on the target nodes and the target interference source nodes includes: determining the priority of the target nodes; and determining the interference optimization strategy based on the priority of the target nodes.

[0096] The specific working process of the computing device in this embodiment of the invention is largely the same as the specific method steps in the above method embodiments, and will not be repeated here.

[0097] This invention collects a spatiotemporal interference dataset; inputs the spatiotemporal interference dataset into an interference prediction model to obtain a node feature matrix and a node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the node correlation matrix is ​​used to characterize the interference characteristics of the external interference information on the network-side information; based on the node feature matrix and the node correlation matrix, the target nodes with interference and the corresponding target interference source nodes are determined. It can automatically capture the dynamic features and spatiotemporal dependencies of each node in the communication network and adaptively update the graph network structure. In addition, through a gating mechanism, important external influencing factors such as external interference events and meteorological events that may cause interference to the communication network are incorporated into the model, thereby achieving effective prediction and prevention of complex interference in 4 / 5G communication networks.

[0098] This invention provides a computer-readable storage medium storing at least one executable instruction that, when executed on a computing device, causes the computing device to perform the interference prediction method in any of the above method embodiments.

[0099] Executable instructions can be used to cause a computing device to perform the following operations:

[0100] Collect a spatiotemporal interference dataset, which includes time information, spatial information, network-side information, and external interference information;

[0101] The spatiotemporal interference dataset is input into the interference prediction model to obtain the node feature matrix and the inter-node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the inter-node correlation matrix is ​​used to characterize the interference features of the external interference information on the network-side information.

[0102] The target nodes with interference and the corresponding target interference source nodes are determined based on the node feature matrix and the inter-node correlation matrix.

[0103] In one optional approach, before inputting the spatiotemporal interference dataset into the interference prediction model to obtain the node feature matrix and the inter-node correlation matrix, the method includes: acquiring a training dataset; the training dataset includes historically collected time information, spatial information, network-side information, and external interference information; and inputting the training dataset into a preset prediction model for training to obtain the trained interference prediction model.

[0104] In one optional approach, the external interference information includes interference event information and meteorological information; the interference prediction model includes a convolutional neural network, a gating unit, and a node spatiotemporal dynamic change update unit; the step of inputting the training dataset into a preset prediction model for training to obtain the trained interference prediction model includes: inputting historically collected interference event information and meteorological information into the convolutional neural network for learning to obtain an external interference factor vector; inputting the external interference factor vector into the gating unit to obtain an external feature matrix; and inputting spatial information, network-side information, and the external feature matrix into the node spatiotemporal dynamic change update unit for training to obtain the interference prediction model.

[0105] In one optional approach, the spatiotemporal interference dataset is collected, which includes time information, spatial information, network-side information, and external interference information, including: collected engineering parameter data, OMC-R measurement data, MDT data, soft-collected data, XDR data, POI data, meteorological data, and interference event data; the engineering parameter data, the OMC-R measurement data, the MDT data, the soft-collected data, the XDR data, the POI data, the meteorological data, and the interference event data are processed to obtain the collected spatiotemporal interference dataset.

[0106] In one optional approach, after determining the target node with interference and the corresponding target interference source node based on the node feature matrix and the inter-node correlation matrix, the process includes: determining an interference optimization strategy based on the target node with interference and the target interference source node; and executing the interference optimization strategy.

[0107] In one optional approach, the number of target nodes is multiple; determining the interference optimization strategy based on the target nodes and the target interference source nodes includes: determining the priority of the target nodes; and determining the interference optimization strategy based on the priority of the target nodes.

[0108] The specific working process of the computing device in this embodiment of the invention is largely the same as the specific method steps in the above method embodiments, and will not be repeated here.

[0109] This invention collects a spatiotemporal interference dataset; inputs the spatiotemporal interference dataset into an interference prediction model to obtain a node feature matrix and a node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the node correlation matrix is ​​used to characterize the interference characteristics of the external interference information on the network-side information; based on the node feature matrix and the node correlation matrix, the target nodes with interference and the corresponding target interference source nodes are determined. It can automatically capture the dynamic features and spatiotemporal dependencies of each node in the communication network and adaptively update the graph network structure. In addition, through a gating mechanism, important external influencing factors such as external interference events and meteorological events that may cause interference to the communication network are incorporated into the model, thereby achieving effective prediction and prevention of complex interference in 4 / 5G communication networks.

[0110] This invention provides an interference prediction device for performing the above-described interference prediction method.

[0111] This invention provides a computer program that can be called by a processor to cause a computing device to execute the interference prediction method in any of the above method embodiments.

[0112] This invention provides a computer program product, which includes a computer program stored on a computer-readable storage medium. The computer program includes program instructions that, when executed on a computer, cause the computer to perform the interference prediction method in any of the above method embodiments.

[0113] The algorithms or displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, the embodiments of the present invention are not directed to any particular programming language. It should be understood that the content of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the invention.

[0114] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0115] Similarly, it should be understood that, in order to streamline the invention and aid in understanding one or more of the various aspects of the invention, features of the embodiments of the invention are sometimes grouped together in a single embodiment, figure, or description thereof in the above description of exemplary embodiments of the invention. However, this disclosure should not be construed as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim.

[0116] Those skilled in the art will understand that modules in the device of the embodiments can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiments can be combined into a single module, unit, or component, and can be divided into multiple sub-modules, sub-units, or sub-components. Except where at least some of such features and / or processes or units are mutually exclusive, any combination can be used to combine all features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all processes or units of any method or device so disclosed. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that serves the same, equivalent, or similar purpose.

[0117] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be construed as limiting the order of execution.

Claims

1. An interference prediction method, characterized in that, The method includes: A spatiotemporal interference dataset is collected, which includes time information, spatial information, network-side information, and external interference information; the external interference information includes interference event information and meteorological information. The spatiotemporal interference dataset is input into the interference prediction model to obtain the node feature matrix and the inter-node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the inter-node correlation matrix is ​​used to characterize the interference features of the external interference information on the network-side information. The target nodes exhibiting interference and their corresponding source nodes are determined based on the node feature matrix and the inter-node correlation matrix; wherein, the node feature matrix... The threshold feature vector is , , The element can be set. Each row represents the F-dimensional feature of the corresponding node; the correlation matrix between nodes. The threshold feature vector is , The value is set according to the scenario; the step of determining the target node with interference and the corresponding target interference source node based on the node feature matrix and the inter-node correlation matrix further includes: when the predicted output node feature matrix H t If the network-side information contained in the h-th row reaches the threshold eigenvector β value, then it is determined that the prediction of node h at time t is interfered with; the inter-node correlation matrix of the prediction output is... When the elements of the correlation matrix Greater than the threshold At that time, node i is considered to be the main source of interference for node j. Identify the target nodes that are interfering and their corresponding source nodes, and output the node feature matrix. ; An interference optimization strategy is determined based on the target node and the target interference source node, and then the interference optimization strategy is executed.

2. The method according to claim 1, characterized in that, Before inputting the spatiotemporal interference dataset into the interference prediction model to obtain the node feature matrix and the inter-node correlation matrix, the following steps are included: Obtain the training dataset; the training dataset includes historically collected time information, spatial information, network-side information, and external interference information; The training dataset is input into a preset prediction model for training to obtain the trained interference prediction model.

3. The method according to claim 2, characterized in that, The interference prediction model includes a convolutional neural network, a gating unit, and a node spatiotemporal dynamic change update unit; the step of inputting the training dataset into the preset prediction model for training to obtain the trained interference prediction model includes: The historically collected information on interference events and meteorological information are input into the convolutional neural network for learning, thereby obtaining a vector of external interference factors. The external interference factor vector is input into the gating unit to obtain the external feature matrix; Spatial information, network-side information, and the external feature matrix are input into the node spatiotemporal dynamic change update unit for training to obtain the interference prediction model.

4. The method according to any one of claims 1-3, characterized in that, The collected spatiotemporal interference dataset includes time information, spatial information, network-side information, and external interference information, including: Collect engineering parameter data, OMC-R measurement data, MDT data, soft data collection data, XDR data, POI data, meteorological data, and interference event data; The engineering parameter data, the OMC-R measurement data, the MDT data, the soft acquisition data, the XDR data, the POI data, the meteorological data, and the interference event data are processed to obtain the spatiotemporal interference dataset.

5. The method according to claim 1, characterized in that, The number of target nodes is multiple; the step of determining the interference optimization strategy based on the target nodes and the target interference source nodes includes: Determine the priority of the target node; The interference optimization strategy is determined based on the priority of the target node.

6. An interference prediction device, characterized in that, The device includes: The acquisition module is used to acquire spatiotemporal interference datasets, which include time information, spatial information, network-side information, and external interference information; the external interference information includes interference event information and meteorological information. The prediction module is used to input the spatiotemporal interference dataset into the interference prediction model to obtain a node feature matrix and a node correlation matrix; the node feature matrix is ​​used to characterize the features of the spatial information and the network-side information; the node correlation matrix is ​​used to characterize the interference features of the external interference information on the network-side information. The determination module is used to determine the target nodes with interference and the corresponding target interference source nodes based on the node feature matrix and the inter-node correlation matrix; wherein, the node feature matrix The threshold feature vector is , , The element can be set. Each row represents the F-dimensional feature of the corresponding node; the correlation matrix between nodes. The threshold feature vector is , The value is set according to the scenario; the step of determining the target node with interference and the corresponding target interference source node based on the node feature matrix and the inter-node correlation matrix further includes: when the predicted output node feature matrix H t If the network-side information contained in the h-th row reaches the threshold eigenvector β value, then it is determined that the prediction of node h at time t is interfered with; the inter-node correlation matrix of the prediction output is... When the elements of the correlation matrix Greater than the threshold At that time, node i is considered to be the main source of interference for node j. Identify the target nodes that are interfering and their corresponding source nodes, and output the node feature matrix. ; The strategy module is used to determine the interference optimization strategy based on the target node of the interference and the target interference source node; An execution module is used to execute the interference optimization strategy.

7. A computing device, characterized in that, include: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction that causes the processor to perform the operation of the interference prediction method as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The storage medium stores at least one executable instruction, which, when executed on a computing device, causes the computing device to perform the operation of the interference prediction method as described in any one of claims 1-5.

Images