Multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment

Abstract

The invention discloses a multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment, and the system comprises an equipment anchoring unit which is configured to anchor spectrum sensing equipment in an area to be sensed; the real-time feature acquisition unit is configured to acquire real-time spectrum features of the spectrum sensing equipment; the real-time feature input unit is configured to input the real-time spectrum features into a pre-constructed spectrum state recognition model; the state sequence output unit is configured to output a frequency spectrum state time sequence of each frequency spectrum sensing device in the to-be-sensed area according to the frequency spectrum state recognition model; the frequency spectrum equipment management unit is configured to perform unified management on the state data of the plurality of heterogeneous frequency spectrum sensing equipment according to the frequency spectrum state time sequence of each frequency spectrum sensing equipment; according to the method, through combination of geographic coordinate clustering and directional edge building, uniform structure expression of equipment with uneven spatial distribution and inconsistent frequency bands is realized, so that the system has good heterogeneous adaptive capacity and state fusion capacity.

Description

Technical Field

[0001] The present invention relates to a spectrum data management system, in particular to a multi-source heterogeneous data fusion management system based on spectrum sensing and detection equipment. Background Art

[0002] With the rapid development of wireless communication technology and the increasing scarcity of spectrum resources, spectrum sensing technology plays a key role in scenarios such as cognitive radio, interference detection, and spectrum resource scheduling. Traditional spectrum sensing systems primarily rely on individual spectrum sensing devices to independently collect and determine the status of their local frequency bands. However, in actual deployments, due to objective conditions such as device heterogeneity, uneven distribution, and varying frequency band coverage, spectrum status data suffers from spatial fragmentation and incomplete observations, making it difficult to form a unified and continuous spectrum status view. Patent publication number CN115250473B discloses a spectrum sharing method, apparatus, and computing device based on a graph convolutional neural network, which can easily and efficiently obtain optimal spectrum access solutions. However, due to the significant spatial correlation and consistent behavior patterns of spectrum status, accurate inference of the status of other nodes cannot be achieved solely through rule-based methods (such as power thresholding) or a small number of annotated points. In large-scale deployments involving joint monitoring of multiple heterogeneous devices, incomplete spectrum status can severely impact the real-time accuracy of spectrum scheduling strategies. Summary of the Invention

[0003] In view of the shortcomings of the existing technology, the present invention provides a multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment, which solves the technical problems raised in the background technology by constructing a graph structure spectrum state recognition of the area to be sensed.

[0004] To achieve the above objectives, the present invention is implemented through the following technical solutions: A multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment includes: a device anchoring unit, configured to anchor a spectrum sensing device in a to-be-sensed area; a real-time feature acquisition unit, configured to acquire real-time spectrum features of the spectrum sensing device; A real-time feature input unit configured to input the real-time spectrum features into a pre-built spectrum state recognition model; The spectrum state recognition model has fitted the time series mapping between the spectrum characteristics and spectrum state of any spectrum sensing device in the area to be sensed; a state sequence output unit, configured to output a spectrum state time sequence sequence of each spectrum sensing device in the to-be-sensed area according to a spectrum state recognition model; a spectrum device management unit configured to uniformly manage status data of a plurality of heterogeneous spectrum sensing devices according to a spectrum status time sequence of each spectrum sensing device; In some specific embodiments, the step of pre-building the spectrum state recognition model includes: S1. Delineating the area to be sensed on an electronic map; S2. Construct a graph structure sample of N spectrum sensing devices at the current timestamp in the area to be sensed; S3. Sequentially collect graph structure samples of multiple timestamps in a time window with a fixed step size on the time axis to generate a graph structure sample sequence; S4. Input the graph structure sample sequence into a time series model based on a graph neural network for iterative training to fit the time series mapping between the spectrum characteristics and spectrum state of any node in the area to be perceived.

[0005] In some specific embodiments, constructing the graph structure sample of N spectrum sensing devices at the current timestamp in step S2 includes: S2-1. Mark the geographical coordinates of N spectrum sensing devices in the area to be sensed, and define the geographical coordinates as N nodes in a graph structure; S2-2, obtain the spectrum feature vectors of N nodes; S2-3. Based on the spectrum feature vectors and geographic coordinates of N nodes, K cluster core nodes and non-cluster core nodes are anchored in the graph structure; S2-4, define the cluster center node as the direction starting point, and mark the upstream node and downstream node between the node pairs; S2-5. For the upstream node and the downstream node of any node pair, determine the edge weight of the directed edge based on their geographic coordinates and the spectrum feature vector; S2-6. Obtain the spectrum status of K cluster core nodes, and perform feature aggregation on non-cluster core nodes based on the directed edges and edge weights of node pairs until the spectrum status of all nodes is obtained; S2-7. Define the graph structure containing the spectrum states of all nodes as the graph structure sample of the current timestamp.

[0006] In some specific embodiments, obtaining the spectrum feature vectors of N nodes in step S2-2 includes: S2-2-1. Obtain several original spectrum parameters collected by each node in a specified frequency band; S2-2-2. Perform feature normalization on a number of original spectrum parameters to generate a number of spectrum features; S2-2-3. Concatenate several spectral features to form a spectral feature vector for each node.

[0007] In some specific embodiments, the step S2-3 of anchoring K cluster core nodes and non-cluster core nodes in the graph structure includes: S2-3-1. Perform K-value clustering based on the geographic coordinates of the nodes to divide N nodes into K spatial clusters; S2-3-2. Based on the spectrum feature vector of each node, determine the cluster core node and non-cluster core nodes in each spatial cluster.

[0008] In some specific embodiments, dividing the N nodes into K spatial clusters in step S2-2-1 includes: S2-3-1-1. Get the N node coordinates corresponding to N nodes; S2-3-1-2, randomly select K node coordinates from N node coordinates as the initial cluster center; S2-3-1-3. For each node coordinate, calculate its coordinate distance with the K initial cluster centers to obtain K coordinate distances for each node coordinate; S2-3-1-4. Select the minimum coordinate distance from the K coordinate distances of each node coordinate; S2-3-1-5, assign the N node coordinates to the initial cluster center with the minimum coordinate distance to obtain K initial spatial clusters; S2-3-1-6. Calculate the centroid of the node coordinates in each initial spatial cluster and determine K centroids; S2-3-1-7, take K centroids as the next round of cluster centers; S2-3-1-8, calculate the change distance between the next round of cluster centers and the initial cluster centers; S2-3-1-9, if the change distance is greater than the change threshold, repeatedly traverse the N node coordinates and assign the N node coordinates to the next round of cluster centers that are at a distance from their minimum coordinates until the change distance of the cluster center is less than the change threshold; S2-3-1-10. Define the cluster centers whose change distances are less than the change threshold and their assigned node coordinates as the spatial clusters to generate K spatial clusters.

[0009] In some specific embodiments, determining the cluster core node and the non-cluster core node in each spatial cluster in step S2-3-2 includes: S2-3-2-1. Determine the mean vector of the spatial cluster based on the spectral feature vector in each spatial cluster; S2-3-2-2, define the mean vector of the spatial cluster as a virtual node; S2-3-2-3. Calculate the vector distance between each node in the spatial cluster and the virtual node; S2-3-2-4. Define the node corresponding to the minimum vector distance as the cluster center node of the spatial cluster, and the remaining nodes as non-cluster center nodes.

[0010] In some specific embodiments, marking upstream nodes and downstream nodes between node pairs in step S2-4 includes: S2-4-1. Select any node pair in each spatial cluster; S2-4-2. Calculate the vector distance between the two nodes in the node pair and their cluster center node; S2-4-3. Compare the vector distances between two nodes and their cluster center nodes; S2-4-4. Select the node corresponding to the smaller vector distance as the upstream node, and the other node as the downstream node.

[0011] In some specific embodiments, determining the edge weight of the directed edge in step S2-5 includes: S2-5-1. Obtain the geographic coordinates of the upstream node and the downstream node in the node pair, and calculate the vector distance of the node pair; S2-5-2. Compare the vector distance of the node pairs with a preset distance threshold, and select node pairs whose vector distance is less than the distance threshold as candidate node pairs; S2-5-3. For candidate node pairs, establish directed edges from upstream nodes to downstream nodes. S2-5-4. Obtain the spectrum feature vectors of the upstream node and the downstream node in each established directed edge; S2-5-5. Calculate the vector similarity of the directed edge based on the spectral feature vectors of the upstream node and the downstream node; The calculation formula of the vector similarity is: ; in, represents the spectrum feature vector of the upstream node, represents the spectrum feature vector of the downstream node, Represents the vector similarity of directed edges; Represents the scale coefficient that controls the similarity attenuation amplitude, Represents the vector Euclidean distance of the spectral feature vectors of the upstream and downstream nodes, Represents the natural exponential function, which is used to map distance values ​​to similarity weights in the interval (0,1); S2-5-6. Obtain the vector similarity of all directed edges; S2-5-7. Normalize the vector similarity of any directed edge with the vector similarity of all directed edges to generate the edge weight of the corresponding directed edge.

[0012] The present invention provides a multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment, effectively solving the problem that traditional rule-based methods cannot achieve state completion and unified management in complex deployment scenarios. Furthermore, by introducing a small number of cluster-core nodes with known spectrum states in a multi-source heterogeneous spectrum sensing environment, the graph structure feature aggregation is used to achieve spectrum state completion for all nodes in the graph, improving the spectrum state coverage capability under conditions of scarce annotations. At the same time, by combining geographic coordinate clustering with directional edge construction, a unified structural expression is achieved for devices with uneven spatial distribution and inconsistent frequency bands, giving the system good heterogeneous adaptability and state fusion capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 This is a structural block diagram of the multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment of the present invention; Figure 2 This is a schematic diagram of the management process of the multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment of the present invention; Figure 3 This is a schematic diagram of the pre-construction process of the spectrum state recognition model of the present invention; Figure 4 The figure is a schematic diagram of the process of generating edge weights of directed edges according to the present invention. DETAILED DESCRIPTION

[0014] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0015] First, the prior art and related concepts involved in the embodiments of the present invention are described: ST-GCN model: Spatio-Temporal Graph Convolutional Network. The ST-GCN model embeds a temporal convolution module in a graph convolutional network to simultaneously model the spatial relationship and temporal dependency of nodes.

[0016] Coordinate distance: used to represent the spatial distance between spectrum sensing devices based on geographic coordinates, using two-dimensional Euclidean distance; Vector distance: used to indicate the similarity between spectral feature vectors in the feature space, using Euclidean distance.

[0017] Example 1: Please refer to Figures 1 to 2The present invention provides a multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment, and the management system includes the following units: a device anchoring unit, configured to anchor a spectrum sensing device in a to-be-sensed area; a real-time feature acquisition unit, configured to acquire real-time spectrum features of the spectrum sensing device; A real-time feature input unit configured to input the real-time spectrum features into a pre-built spectrum state recognition model; The spectrum state recognition model has fitted the time series mapping between the spectrum characteristics and spectrum state of any spectrum sensing device in the area to be sensed; a state sequence output unit, configured to output a spectrum state time sequence sequence of each spectrum sensing device in the to-be-sensed area according to a spectrum state recognition model; Specifically, the spectrum state time series sequence refers to a set of spectrum states identified within a continuous time window for each spectrum sensing device in the area to be sensed, a state sequence composed in timestamp order, which is used to reflect the spectrum state change trend of several spectrum sensing devices in the area to be sensed within the monitored frequency band.

[0018] a spectrum device management unit configured to uniformly manage status data of a plurality of heterogeneous spectrum sensing devices according to a spectrum status time sequence of each spectrum sensing device; Among them, unified management can be to implement classification management, interference detection, dynamic scheduling or resource allocation of spectrum usage status based on preset rules to achieve intelligent fusion management of multi-source heterogeneous spectrum data.

[0019] In this embodiment, by providing a device anchoring unit, a real-time feature acquisition unit, a real-time feature input unit, a state sequence output unit, and a spectrum device management unit, the collection of spectrum state time series information for any spectrum sensing device within the sensing area is achieved, establishing a unified management path for multi-source heterogeneous spectrum data. This embodiment completes spectrum state identification and time series expression based on a spectrum state recognition model, and manages the spectrum state of multiple devices through a unified spectrum device management unit, improving centralized spectrum state management.

[0020] Example 2: See Figures 2 to 4 The technical solution of this embodiment 2 differs from that of embodiment 1 in that it discloses a pre-construction step of the spectrum state identification model described in the embodiment, and the pre-construction step includes: S1. Delineating the area to be sensed on an electronic map; The electronic map refers to a geographic information system (GIS) with spatial coordinates, used to mark the geographic location of spectrum sensing equipment. The area to be sensed refers to the spectrum observation target area selected on the electronic map. It is usually a two-dimensional area within the communication coverage area, a local urban grid, or a designated scene boundary, where spectrum sensing equipment is deployed.

[0021] S2. Construct a graph structure sample of N spectrum sensing devices at the current timestamp in the area to be sensed; S3. Sequentially collect graph structure samples of multiple timestamps in a time window with a fixed step size on the time axis to generate a graph structure sample sequence; S4. Input the graph structure sample sequence into a time series model based on a graph neural network for iterative training to fit the time series mapping between the spectrum characteristics and spectrum state of any node in the area to be perceived.

[0022] The temporal mapping refers to taking the spectral features of nodes in the graph structure sample sequence and the relationship between adjacent nodes in the graph structure as input to learn the prediction of the spectral state of any node at the current timestamp; wherein, any node representation model has the spatial generalization ability of the spectral state of all nodes in the perception area.

[0023] It should be noted that the "graph neural network-based time series model" in this embodiment refers to a deep learning model for simultaneously modeling graph structure adjacency relationships and time series dependencies. It can use node features, edge connection relationships, and edge weights in graph structure sample sequences to dynamically predict or identify node states in the time dimension. The ST-GCN model can be used as the model.

[0024] In this embodiment, by constructing a graph structure sample containing the spatial distribution and adjacency relationship of spectrum sensing devices, and generating a graph structure sample sequence on the time axis, which is used as input to train the graph neural network model, the spectrum state recognition model is implemented to fit the time series mapping of the spectrum state of any node at any timestamp, thereby establishing a spectrum state time series prediction model that can cover all spectrum sensing devices in the area to be sensed.

[0025] Exemplarily, step S2 in this embodiment further includes: S2-1. Mark the geographical coordinates of N spectrum sensing devices in the area to be sensed, and define the geographical coordinates as N nodes in a graph structure; S2-2, obtain the spectrum feature vectors of N nodes; S2-3. Based on the spectrum feature vectors and geographic coordinates of N nodes, K cluster core nodes and non-cluster core nodes are anchored in the graph structure; S2-4, define the cluster center node as the direction starting point, and mark the upstream node and downstream node between the node pairs; S2-5. For the upstream node and the downstream node of any node pair, determine the edge weight of the directed edge based on their geographic coordinates and the spectrum feature vector; S2-6. Obtain the spectrum status of K cluster core nodes, and perform feature aggregation on non-cluster core nodes based on the directed edges and edge weights of node pairs until the spectrum status of all nodes is obtained; Among them, the spectrum state refers to the discrete signal usage state of the spectrum sensing device in a specified frequency band, which usually includes but is not limited to "idle state", "occupied state", "interference state" or "unknown state", and is used to indicate whether the device has signal activity and its signal type characteristics in the frequency band. It is the classification target label of the spectrum state recognition model.

[0026] The spectrum status may be extracted from historical annotated data, or may be set according to collected spectrum power thresholds, occupancy rates, interference intensity, and other rule conditions.

[0027] It should be noted that feature aggregation of non-cluster-center nodes based on directed edges and edge weights involves collecting the spectral states of all upstream adjacent nodes from the non-cluster-center node as the target node. Combined with the edge weights of the corresponding directed edges, the feature aggregation (message passing) mechanism within graph neural networks is then used to weight the state features of the upstream nodes and aggregate them to the downstream nodes, thus completing state propagation. This feature aggregation process reflects the conditional dependencies of node states within the graph structure. A larger edge weight indicates a more significant state influence, ultimately achieving the gradual propagation of state features from the cluster-center node to the entire graph structure.

[0028] S2-7. Define the graph structure containing the spectrum states of all nodes as the graph structure sample of the current timestamp.

[0029] In this embodiment, the geographic coordinates of the spectrum sensing devices are constructed as nodes of a graph structure, and directed edges and their edge weights are constructed based on the relationship between the spatial positions and spectrum characteristics between the devices, thereby realizing graph structure modeling of the spectrum status at the current timestamp. By calibrating the spectrum status of the cluster center nodes and performing feature aggregation of message transmission based on edge weights, the spectrum status results of all nodes at the current timestamp are generated, thereby completing the propagation of the node spectrum status within the graph structure.

[0030] Illustratively, step S2-2 in this embodiment further includes: S2-2-1. Obtain several original spectrum parameters collected by each node in a specified frequency band; Several spectrum parameters include but are not limited to quantifiable spectrum statistical characteristics such as signal power, bandwidth occupancy, energy spectrum density, burst behavior indicators, spectrum stability, and spectrum usage duration.

[0031] S2-2-2. Perform feature normalization on a number of original spectrum parameters to generate a number of spectrum features; It should be noted that normalization involves adapting different processing methods based on the parameter type of each spectral parameter. Specifically, for continuously varying energy parameters (such as average power and peak power), Z-score normalization can be used to preserve their relative deviation and directionality. For proportional parameters such as occupancy ratio and bandwidth utilization, normalization methods such as minimum-maximum normalization or scaling are recommended. For example, for each of these different types of spectral parameters, an adapted normalization method is ultimately required to unify the spectral parameters to the same dimensional scale to meet input consistency requirements.

[0032] S2-2-3. Concatenate several spectral features to form a spectral feature vector for each node.

[0033] Among them, feature splicing refers to arranging several standardized spectral features corresponding to each node in a preset order to form a one-dimensional vector structure of fixed dimension as the input feature representation of the node in the graph structure.

[0034] Illustratively, step S2-3 in this embodiment further includes: S2-3-1. Perform K-value clustering based on the geographic coordinates of the nodes to divide N nodes into K spatial clusters; S2-3-2. Based on the spectrum feature vector of each node, determine the cluster core node and non-cluster core nodes in each spatial cluster.

[0035] Exemplarily, in this embodiment, dividing the N nodes into K spatial clusters further includes: S2-3-1-1. Get the N node coordinates corresponding to N nodes; S2-3-1-2, randomly select K node coordinates from N node coordinates as the initial cluster center; S2-3-1-3. For each node coordinate, calculate its coordinate distance with the K initial cluster centers to obtain K coordinate distances for each node coordinate; S2-3-1-4. Select the minimum coordinate distance from the K coordinate distances of each node coordinate; S2-3-1-5, assign the N node coordinates to the initial cluster center with the minimum coordinate distance to obtain K initial spatial clusters; S2-3-1-6. Calculate the centroid of the node coordinates in each initial spatial cluster and determine K centroids; It should be noted that the centroid calculation process means: for each spatial cluster, the geographic coordinates of all nodes in the cluster are counted, and the arithmetic mean of the values ​​of all nodes in each dimension of coordinates is taken. The obtained average coordinate is the centroid coordinate of the spatial cluster.

[0036] S2-3-1-7, take K centroids as the next round of cluster centers; S2-3-1-8, calculate the change distance between the next round of cluster centers and the initial cluster centers; S2-3-1-9, if the change distance is greater than the change threshold, repeatedly traverse the N node coordinates and assign the N node coordinates to the next round of cluster centers that are at a distance from their minimum coordinates until the change distance of the cluster center is less than the change threshold; S2-3-1-10. Define the cluster centers whose change distances are less than the change threshold and their assigned node coordinates as the spatial clusters to generate K spatial clusters.

[0037] In this embodiment, by performing K-value clustering based on the geographic coordinates of spectrum sensing devices, adopting cluster center iterative update and distance convergence judgment mechanism, N nodes are adaptively divided into K spatial clusters, thereby realizing structured clustering of spectrum sensing devices in geographic space.

[0038] Exemplarily, in this embodiment, determining the cluster core node and the non-cluster core node in each spatial cluster in step S2-3-2 includes: S2-3-2-1. Determine the mean vector of the spatial cluster based on the spectral feature vector in each spatial cluster; S2-3-2-2, define the mean vector of the spatial cluster as a virtual node; S2-3-2-3. Calculate the vector distance between each node in the spatial cluster and the virtual node; S2-3-2-4. Define the node corresponding to the minimum vector distance as the cluster center node of the spatial cluster, and the remaining nodes as non-cluster center nodes.

[0039] In this embodiment, by calculating the mean vector of the spectral feature vectors of all nodes in each spatial cluster, constructing a virtual node with reference to the mean vector, and then selecting the node with the smallest distance as the cluster center node, the cluster center node that is most representative in terms of spectral behavior is selected, providing a characteristic aggregation representative for the propagation starting point of the spectral state.

[0040] Exemplarily, the step S2-4 further includes: S2-4-1. Select any node pair in each spatial cluster; S2-4-2. Calculate the vector distance between the two nodes in the node pair and their cluster center node; S2-4-3. Compare the vector distances between two nodes and their cluster center nodes; S2-4-4. Select the node corresponding to the smaller vector distance as the upstream node, and the other node as the downstream node.

[0041] In this embodiment, by comparing the vector distances between two nodes in any node pair and the cluster center node to which they belong, the one with a closer distance is determined as the upstream node, and the one with a farther distance is determined as the downstream node, a regular division of directed edge directions based on spatial geometric positions is implemented, thereby constructing edges with propagation directionality, ensuring that the spectrum state is transmitted in an orderly manner along a distance-increasing path.

[0042] Illustratively, step S2-5 in this embodiment further includes: S2-5-1. Obtain the geographic coordinates of the upstream node and the downstream node in the node pair, and calculate the vector distance of the node pair; S2-5-2. Compare the vector distance of the node pairs with a preset distance threshold, and select node pairs whose vector distance is less than the distance threshold as candidate node pairs; S2-5-3. For candidate node pairs, establish directed edges from upstream nodes to downstream nodes. S2-5-4. Obtain the spectrum feature vectors of the upstream node and the downstream node in each established directed edge; S2-5-5. Calculate the vector similarity of the directed edge based on the spectral feature vectors of the upstream node and the downstream node; The calculation formula of the vector similarity is: ; in, represents the spectrum feature vector of the upstream node, represents the spectrum feature vector of the downstream node, Represents the vector similarity of directed edges; Represents the scale coefficient that controls the similarity attenuation amplitude, The vector Euclidean distance (Euclidean distance) representing the spectral feature vectors of the upstream and downstream nodes, Represents the natural exponential function, which is used to map distance values ​​to similarity weights in the interval (0,1); Specifically, by negatively compressing the vector distance through the natural exponential function, the goal of smaller distance, greater similarity, and stronger edge weight can be achieved, which is in line with the logic of "stronger information coupling between similar nodes" in the dependency relationship.

[0043] S2-5-6. Obtain the vector similarity of all directed edges; S2-5-7: Normalize the vector similarity of any directed edge with the vector similarity of all directed edges to generate an edge weight for the corresponding directed edge. The edge weight is used to represent the strength of state dependency between nodes.

[0044] In this embodiment, the similarity between the spectrum feature vectors of node pairs whose vector distance is less than a threshold is calculated, and the negative exponential function of the vector distance is used to map it to a continuous value in the interval (0,1]. All similarities are then normalized to obtain edge weights for directed edges in the graph structure. This implements state dependency modeling based on behavioral similarity within the graph structure, quantifies the transmission strength of the spectrum state between nodes, and enables each node to reasonably aggregate the spectrum state transmitted upstream.

[0045] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the process or function described in the embodiments of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired method (e.g., infrared, wireless, microwave, etc.).

[0046] The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., DVD ), or semiconductor media. The semiconductor media can be a solid-state drive.

[0047] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not implemented. In addition, the coupling or direct coupling or communication connection shown or discussed can be through some interface, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0048] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment, including: A device anchoring unit, configured to anchor a spectrum sensing device in a to-be-sensed area; a real-time feature acquisition unit, configured to acquire real-time spectrum features of the spectrum sensing device; A real-time feature input unit configured to input real-time spectrum features into a pre-built spectrum state recognition model; The spectrum state recognition model has fitted the time series mapping between the spectrum characteristics and spectrum state of any spectrum sensing device in the area to be sensed; a state sequence output unit configured to output a spectrum state time sequence sequence of each spectrum sensing device in the to-be-sensed area according to a spectrum state recognition model; The spectrum device management unit is configured to uniformly manage status data of a plurality of heterogeneous spectrum sensing devices according to a spectrum status time sequence of each spectrum sensing device.

2. The multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment according to claim 1 is characterized in that: The pre-construction step of the spectrum state recognition model includes: S1. Delineating the area to be sensed on an electronic map; S2. Construct a graph structure sample of N spectrum sensing devices at the current timestamp in the area to be sensed; S3. Sequentially collect graph structure samples of multiple timestamps in a time window with a fixed step size on the time axis to generate a graph structure sample sequence; S4. Input the graph structure sample sequence into a time series model based on a graph neural network for iterative training to fit the time series mapping between the spectrum characteristics and spectrum state of any node in the area to be perceived.

3. The multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment according to claim 2 is characterized in that: The step S2 of constructing the graph structure sample of N spectrum sensing devices at the current timestamp includes: S2-1. Mark the geographical coordinates of N spectrum sensing devices in the area to be sensed, and define the geographical coordinates as N nodes in a graph structure; S2-2, obtain the spectrum feature vectors of N nodes; S2-3. Based on the spectrum feature vectors and geographic coordinates of N nodes, K cluster core nodes and non-cluster core nodes are anchored in the graph structure; S2-4, define the cluster center node as the direction starting point, and mark the upstream node and downstream node between the node pairs; S2-5. For the upstream node and the downstream node of any node pair, determine the edge weight of the directed edge based on their geographic coordinates and the spectrum feature vector; S2-6. Obtain the spectrum status of K cluster core nodes, and perform feature aggregation on non-cluster core nodes according to the directed edges and edge weights corresponding to the nodes, until the spectrum status of all nodes is obtained; S2-7. Define the graph structure containing the spectrum states of all nodes as the graph structure sample of the current timestamp.

4. The multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment according to claim 3 is characterized in that: The step S2-2 of obtaining the spectrum feature vectors of N nodes includes: S2-2-1. Obtain several original spectrum parameters collected by each node in a specified frequency band; S2-2-2. Perform feature normalization on a number of original spectrum parameters to generate a number of spectrum features; S2-2-3. Concatenate several spectral features to form a spectral feature vector for each node.

5. The multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment according to claim 3 is characterized in that: In step S2-3, anchoring K cluster core nodes and non-cluster core nodes in the graph structure includes: S2-3-1. Perform K-value clustering based on the geographic coordinates of the nodes to divide N nodes into K spatial clusters; S2-3-2. Based on the spectrum feature vector of each node, determine the cluster core node and non-cluster core nodes in each spatial cluster.

6. The multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment according to claim 5, characterized in that: The step S2-2-1 of dividing N nodes into K spatial clusters includes: S2-3-1-1. Get the N node coordinates corresponding to N nodes; S2-3-1-2, randomly select K node coordinates from N node coordinates as the initial cluster center; S2-3-1-3. For each node coordinate, calculate its coordinate distance with the K initial cluster centers to obtain K coordinate distances for each node coordinate; S2-3-1-4. Select the minimum coordinate distance from the K coordinate distances of each node coordinate; S2-3-1-5, assign the N node coordinates to the initial cluster center with the minimum coordinate distance to obtain K initial spatial clusters; S2-3-1-6. Calculate the centroid of the node coordinates in each initial spatial cluster and determine K centroids; S2-3-1-7, take K centroids as the next round of cluster centers; S2-3-1-8, calculate the change distance between the next round of cluster centers and the initial cluster centers; S2-3-1-9, if the change distance is greater than the change threshold, repeatedly traverse the N node coordinates and assign the N node coordinates to the next round of cluster centers that are at a distance from their minimum coordinates until the change distance of the cluster center is less than the change threshold; S2-3-1-10. Define the cluster centers whose change distances are less than the change threshold and their assigned node coordinates as the spatial clusters to generate K spatial clusters.

7. The multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment according to claim 5, characterized in that: Determining the cluster core nodes and non-cluster core nodes in each spatial cluster in step S2-3-2 includes: S2-3-2-1. Determine the mean vector of the spatial cluster based on the spectral feature vector in each spatial cluster; S2-3-2-2, define the mean vector of the spatial cluster as a virtual node; S2-3-2-3. Calculate the vector distance between each node in the spatial cluster and the virtual node; S2-3-2-4. Define the node corresponding to the minimum vector distance as the cluster center node of the spatial cluster, and the remaining nodes as non-cluster center nodes.

8. The multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment according to claim 3, characterized in that: The step S2-4 of marking the upstream node and the downstream node between the node pairs includes: S2-4-1. Select any node pair in each spatial cluster; S2-4-2. Calculate the vector distance between the two nodes in the node pair and their cluster center node; S2-4-3. Compare the vector distances between two nodes and their cluster center nodes; S2-4-4. Select the node corresponding to the smaller vector distance as the upstream node, and the other node as the downstream node.

9. The multi-source heterogeneous data fusion management system based on spectrum sensing detection equipment according to claim 3, characterized in that: Determining the edge weight of the directed edge in step S2-5 includes: S2-5-1. Obtain the geographic coordinates of the upstream node and the downstream node in the node pair, and calculate the vector distance of the node pair; S2-5-2. Compare the vector distance of the node pairs with a preset distance threshold, and select node pairs whose vector distance is less than the distance threshold as candidate node pairs; S2-5-3. For candidate node pairs, establish directed edges from upstream nodes to downstream nodes. S2-5-4. Obtain the spectrum feature vectors of the upstream node and the downstream node in each established directed edge; S2-5-5. Calculate the vector similarity of the directed edge based on the spectral feature vectors of the upstream node and the downstream node; The calculation formula of the vector similarity is: ; in, represents the spectrum feature vector of the upstream node, represents the spectrum feature vector of the downstream node, Represents the vector similarity of directed edges; Represents the scale coefficient that controls the similarity attenuation amplitude, The vector Euclidean distance between the spectral feature vectors of the upstream and downstream nodes, Represents the natural exponential function, which is used to map distance values ​​to similarity weights in the interval (0,1); S2-5-6. Obtain the vector similarity of all directed edges; S2-5-7. Normalize the vector similarity of any directed edge with the vector similarity of all directed edges to generate the edge weight of the corresponding directed edge.

Images