Method, apparatus and device for power allocation for integrated network

Abstract

The present disclosure provides a power allocation method, device and equipment for integrated communication and sensing network. The power allocation method comprises: obtaining a first feature of a UAV base station, a second feature of a communication user and a third feature of a sensing target, wherein the first feature comprises a total power constraint of the UAV base station; obtaining target embedding features of nodes in a heterogeneous graph based on the first feature, the second feature and the third feature by using a graph neural network model, and obtaining a first predicted power allocation weight of the communication user associated with the UAV base station and a second predicted power allocation weight of the sensing target associated with the UAV base station based on the target embedding features of the nodes in the heterogeneous graph; determining a first target power value of the communication user associated with the UAV base station according to the total power constraint and the first predicted power allocation weight, and determining a second target power value of the sensing target associated with the UAV base station according to the total power constraint and the second predicted power allocation weight.

Description

Technical Field

[0001] This disclosure relates to the field of communication technology, and more specifically, to a power distribution method, apparatus, and device for a communication and sensing integrated network. Background Technology

[0002] With the rapid development of communication technology, integrated communication and sensing technology has become a core technology for realizing the synergy of communication and sensing functions, and has significant application value in typical scenarios such as UAV emergency communication and vehicle-to-everything (V2X) communication. Especially in post-disaster emergency scenarios, traditional ground communication facilities are easily damaged, and UAV base stations, with their advantages of rapid deployment and three-dimensional mobility, have become a key carrier for restoring communication and sensing capabilities. Summary of the Invention

[0003] A brief overview of this disclosure is given below to provide a basic understanding of some aspects of it. However, it should be understood that this overview is not an exhaustive summary of this disclosure. It is not intended to identify key or essential parts of this disclosure, nor is it intended to limit the scope of this disclosure. Its purpose is merely to present certain concepts of this disclosure in a simplified form as a prelude to the more detailed description that follows.

[0004] The network needs to support various services, including voice communication, video transmission, life detection, and environmental monitoring. Different services have varying Quality of Service (QoS) requirements regarding transmission rate and accuracy. However, the transmission power of drone base stations is limited. Therefore, dynamically and intelligently allocating this limited power to numerous communication and sensing links, while maintaining the overall power of the drone base station, to meet the differentiated needs of multi-dimensional services, has become a core challenge in improving the overall performance and mission reliability of the integrated sensing network.

[0005] In related technologies, heuristic methods such as fixed power allocation ratios are used for power allocation. These methods lack dynamic coordination capabilities across services, resulting in poor flexibility. This can easily lead to tasks with high power requirements (such as life detection and other urgent tasks) performing poorly due to insufficient power allocation. Furthermore, related technologies often separate the power allocation issues of communication links and sensing links for optimization, failing to effectively capture and utilize the inherent interference and coordination relationships between communication and sensing links. This results in the inability to achieve system-level global optimization and efficient utilization under limited power resources, and cannot meet the needs of complex integrated communication and sensing networks.

[0006] To address the aforementioned issues, this disclosure proposes the following solutions.

[0007] According to a first aspect of this disclosure, a power allocation method for a sensing-interactive network is provided, comprising: acquiring a first feature of an unmanned aerial vehicle (UAV) base station, a second feature of a communication user, and a third feature of a sensing target, wherein the first feature includes a total power constraint of the UAV base station; using a graph neural network model, obtaining target embedding features of nodes in a heterogeneous graph based on the first feature, the second feature, and the third feature, and obtaining a first predicted power allocation weight of a communication user associated with the UAV base station and a second predicted power allocation weight of a sensing target associated with the UAV base station based on the target embedding features of nodes in the heterogeneous graph, wherein the heterogeneous graph corresponds to the sensing-interactive network, the UAV base station, the communication user, and the sensing target constitute nodes in the heterogeneous graph, and the edges in the heterogeneous graph include edges between UAV base station nodes and associated communication user nodes, and edges between UAV base station nodes and associated sensing target nodes; determining a first target power value of a communication user associated with the UAV base station based on the total power constraint and the first predicted power allocation weight, and determining a second target power value of a sensing target associated with the UAV base station based on the total power constraint and the second predicted power allocation weight.

[0008] In some embodiments, the first feature further includes at least one of the location of the UAV base station, the current power used, the total backhaul link capacity, and the occupied backhaul link capacity; the second feature includes at least one of the location of the communication user, the communication rate quality of service requirement threshold, and the task priority weight coefficient; and the third feature includes at least one of the location of the sensing target, the sensing rate quality of service requirement threshold, and the task priority weight coefficient.

[0009] In some embodiments, the graph neural network model includes multiple computational layers. Using the graph neural network model, target embedding features of nodes in a heterogeneous graph are obtained based on the first feature, the second feature, and the third feature. Obtaining a first predicted power allocation weight for a communication user associated with a UAV base station and a second predicted power allocation weight for a sensing target associated with a UAV base station based on the target embedding features of nodes in the heterogeneous graph includes: in each computational layer of the multiple computational layers, for each node in the heterogeneous graph, using the embedding features of the node's neighboring nodes obtained in the previous computational layer, an aggregation result for the node is obtained; using the embedding features of the node obtained in the previous computational layer and the aggregation result of the node, the embedding feature of the node in that computational layer is obtained, wherein the embedding feature of the node in the last computational layer of the multiple computational layers is used as the target embedding feature of the node, wherein in the first computational layer of the multiple computational layers, the initial embedding feature of the node is used as the embedding feature obtained by the node in the previous computational layer, the initial embedding feature of the UAV base station node is obtained based on the first feature, the initial embedding feature of the communication user node is obtained based on the second feature, and the initial embedding feature of the sensing target node is obtained based on the third feature.

[0010] In some embodiments, the graph neural network model is a graph attention neural network model. The aggregation result of the node is obtained by utilizing the embedding features of its neighboring nodes obtained in the previous computational layer. This includes: for each neighboring node, obtaining the attention weight coefficient between the node and its neighboring nodes based on the task priority weight vector, the attention weight vector, the embedding features of the node obtained in the previous computational layer, and the embedding features of the neighboring nodes obtained in the previous computational layer; wherein the task priority weight vector is obtained based on the task priority weight coefficient of the node and the task priority weight coefficient of the neighboring nodes; and obtaining the aggregation result of the node based on the embedding features of each neighboring node obtained in the previous computational layer and the attention weight coefficients between the node and each neighboring node.

[0011] In some embodiments, using a graph neural network model, target embedding features of nodes in a heterogeneous graph are obtained based on the first feature, the second feature, and the third feature. The first predicted power allocation weight for a communication user associated with a UAV base station and the second predicted power allocation weight for a sensing target associated with the UAV base station are obtained based on the target embedding features of the nodes in the heterogeneous graph. This includes: obtaining the first predicted power allocation weight based on the target embedding features of the node corresponding to the UAV base station, the target embedding features of the node corresponding to the associated communication user, and the task priority weight coefficient of the communication user; and / or obtaining the second predicted power allocation weight based on the target embedding features of the node corresponding to the UAV base station, the target embedding features of the node corresponding to the associated sensing target, and the task priority weight coefficient of the sensing target.

[0012] In some embodiments, determining a first target power value for a communication user associated with a UAV base station based on the total power constraint and the first predicted power allocation weight, and determining a second target power value for a sensing target associated with a UAV base station based on the total power constraint and the second predicted power allocation weight, includes: for each UAV base station, normalizing the first predicted power allocation weight of each communication user associated with the UAV base station to obtain a corresponding first target power allocation weight, and normalizing the second predicted power allocation weight of each sensing target to obtain a corresponding second target power allocation weight; for each communication user associated with the UAV base station, determining a first target power value for the communication user based on the total power constraint of the UAV base station and the first target power allocation weight of the communication user; and for each sensing target associated with the UAV base station, determining a second target power value for the sensing target based on the total power constraint of the UAV base station and the second target power allocation weight of the sensing target.

[0013] In some embodiments, the power allocation method further includes: before determining a first target power value of a communication user associated with a UAV base station based on the total power constraint and the first predicted power allocation weight, and before determining a second target power value of a sensing target associated with a UAV base station based on the total power constraint and the second predicted power allocation weight, updating the first predicted power allocation weight based on the task priority weight coefficient of the communication user, and using the updated first predicted power allocation weight as the first predicted power allocation weight of the communication user; and updating the second predicted power allocation weight based on the task priority weight coefficient of the sensing target, and using the updated second predicted power allocation weight as the second predicted power allocation weight of the sensing target.

[0014] In some embodiments, the power allocation method further includes: for each UAV base station, controlling the UAV base station to allocate a corresponding first target power value to the communication link of the communication user associated with the UAV base station, and controlling the UAV base station to allocate a corresponding second target power value to the sensing link of the sensing target associated with the UAV base station.

[0015] In some embodiments, the power allocation method further includes: for each UAV base station, obtaining a first initial power value for each communication user associated with the UAV base station and a second initial power value for each sensing target associated with the UAV base station; for each communication user, controlling the UAV base station to allocate a corresponding first initial power value to the communication link of the communication user, and for each sensing target, controlling the UAV base station to allocate a corresponding second initial power value to the sensing link of the sensing target; in response to the determination of the first target power value of each communication user and the second target power value of each sensing target, controlling the UAV base station to adjust the power value of the communication link of each communication user from the corresponding first initial power value to the corresponding first target power value, and controlling the UAV base station to adjust the power value of the sensing link of each sensing target from the corresponding second initial power value to the corresponding second target power value.

[0016] In some embodiments, obtaining a first initial power value for each communication user associated with the UAV base station and a second initial power value for each sensing target associated with the UAV base station includes: obtaining the communication link channel gain of each communication user and the sensing link channel gain of each sensing target; for each communication user, obtaining a first initial allocation coefficient for the communication user based on a first total allocation coefficient and a first sub-allocation coefficient for the communication user, and obtaining a first initial power value for the communication user based on the total power constraint of the UAV base station and the first initial allocation coefficient for the communication user, wherein the first sub-allocation coefficient is obtained based on the communication link channel gain and the task priority weight coefficient of the communication user, and the first total allocation coefficient is obtained by summing the first sub-allocation coefficients of each communication user; for each sensing target, obtaining a second initial allocation coefficient for the sensing target based on a second total allocation coefficient and a second sub-allocation coefficient for the sensing target, and obtaining a second initial power value for the sensing target based on the total power constraint of the UAV base station and the second initial allocation coefficient for the sensing target, wherein the second sub-allocation coefficient is obtained based on the sensing link channel gain and the task priority weight coefficient of the sensing target, and the second total allocation coefficient is obtained by summing the second sub-allocation coefficients of each sensing target.

[0017] In some embodiments, the power allocation method further includes: acquiring training samples, wherein the feature portion of the training samples includes a first feature of the sample UAV base station in the sample sensing integrated network, a second feature of the sample communication user, and a third feature of the sample sensing target, wherein the first feature includes the total power constraint of the sample UAV base station, and the label portion of the training samples includes a first sample power allocation weight of the sample communication user associated with the sample human-machine base station and a second sample power allocation weight of the sample sensing target associated with the sample human-machine base station; and training a graph neural network model based on the training samples.

[0018] In some embodiments, training a graph neural network model based on training samples includes: training the graph neural network model based on training samples with the objective of reducing the value of a target loss function, wherein the target loss function is set based on the mean square error between the first sample power allocation weight of the sample communication user and the first predicted power allocation weight of the sample communication user obtained using the graph neural network model, and the mean square error between the second sample power allocation weight of the sample perception target and the second predicted power allocation weight of the sample perception target obtained using the graph neural network model.

[0019] In some embodiments, the graph neural network model is trained based on a Markov decision process, wherein the reward function in the Markov decision process is based on the state... implement The state at time t is determined by the communication performance reward, perception performance reward, priority reward, backhaul load penalty, and total power penalty. Including the characteristics of each node in the synesthetic network at time t, and the actions at time t. The predicted power allocation weights or target power allocation weights of the nodes associated with the UAV base station are set based on a graph neural network model.

[0020] According to a second aspect of this disclosure, a power allocation device for a sensing-interconnected network is provided, comprising: an acquisition module configured to acquire a first feature of a UAV base station, a second feature of a communication user, and a third feature of a sensing target, wherein the first feature includes a total power constraint of the UAV base station; a weight determination module configured to use a graph neural network model to obtain target embedding features of nodes in a heterogeneous graph based on the first feature, the second feature, and the third feature, and to obtain a first predicted power allocation weight of a communication user associated with the UAV base station and a second predicted power allocation weight of a sensing target associated with the UAV base station based on the target embedding features of nodes in the heterogeneous graph, wherein the heterogeneous graph corresponds to the sensing-interconnected network, the UAV base station, the communication user, and the sensing target constitute nodes in the heterogeneous graph, and the edges in the heterogeneous graph include edges between UAV base station nodes and associated communication user nodes, and edges between UAV base station nodes and associated sensing target nodes; and a power determination module configured to determine a first target power value of a communication user associated with the UAV base station based on the total power constraint and the first predicted power allocation weight, and to determine a second target power value of a sensing target associated with the UAV base station based on the total power constraint and the second predicted power allocation weight.

[0021] According to a third aspect of this disclosure, a power distribution device for a sensor-integrated network is provided, including a processor and a memory, wherein the memory stores instructions that, when executed by the processor, implement the steps of the power distribution method as described above.

[0022] According to a fourth aspect of this disclosure, an aerial platform station is provided, comprising: a power distribution device as described above; or a power distribution device as described above.

[0023] According to a fifth aspect of this disclosure, a non-transitory computer-readable storage medium is provided, wherein instructions are stored on the non-transitory computer-readable storage medium, which, when executed by a processor, implement the steps of the power allocation method as described above.

[0024] According to a sixth aspect of this disclosure, a computer program product is provided, including instructions that, when executed by a processor, implement the steps of the power allocation method as described above.

[0025] According to embodiments of this disclosure, UAV base stations, communication users, and sensing targets are treated as nodes in a heterogeneous graph corresponding to the integrated sensing network. By utilizing a graph neural network model, the embedding features of each node are updated to uncover the complex power competition and coordination relationships between various communication links and sensing links from a global perspective. This allows for the discovery of implicit power allocation features in the integrated sensing network, leading to a better power allocation strategy for the integrated sensing network. This achieves multi-dimensional power joint scheduling and globally optimal power allocation across services and links, enabling the effective execution of related tasks and improving their processing efficiency. It also enhances the task completion capability and reliability of the integrated sensing network in emergency scenarios.

[0026] Other features and advantages of this disclosure will become clearer from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0027] The accompanying drawings, which form part of this specification, illustrate embodiments of this disclosure and, together with the specification, serve to explain the principles of this disclosure.

[0028] This disclosure will become clearer with reference to the accompanying drawings and the following detailed description, wherein: Figure 1 A schematic diagram of a synergistic network according to some exemplary embodiments of the present disclosure is shown; Figure 2 A schematic flowchart of a power allocation method for a sensing-integrated network according to some exemplary embodiments of the present disclosure is shown; Figure 3 A schematic diagram of a graph neural network model according to some exemplary embodiments of the present disclosure is shown; Figure 4 A schematic diagram of a process for obtaining the target embedding features of a node using a graph neural network model according to some exemplary embodiments of the present disclosure is shown. Figure 5 A schematic flowchart illustrating the process of obtaining the aggregation result of nodes according to some exemplary embodiments of the present disclosure is shown; Figure 6 A schematic flowchart illustrating the process of obtaining a first target power value and a second target power value according to some exemplary embodiments of the present disclosure is shown. Figure 7 A schematic diagram of a power allocation process based on an initial power value according to some exemplary embodiments of the present disclosure is shown; Figure 8 A flowchart illustrating a training method for a graph neural network model according to some embodiments of the present disclosure is shown. Figure 9A schematic diagram of a power distribution apparatus for a sensing-integrated network according to some exemplary embodiments of the present disclosure is shown; Figure 10 A schematic diagram of a power distribution device for a sensing-integrated network according to some exemplary embodiments of the present disclosure is shown; Figure 11 A schematic block diagram of a computer system on which embodiments of the present disclosure may be implemented is shown.

[0029] Note that in the embodiments described below, the same reference numerals are sometimes used across different figures to denote the same parts or parts having the same function, and repeated descriptions are omitted. In this specification, similar reference numerals and letters are used to denote similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.

[0030] For ease of understanding, the positions, dimensions, and extents of the structures shown in the accompanying drawings and other materials may not represent actual positions, dimensions, and extents. Therefore, the disclosed invention is not limited to the positions, dimensions, and extents disclosed in the accompanying drawings and other materials. Furthermore, the drawings are not necessarily drawn to scale, and some features may be enlarged to show details of specific components. Detailed Implementation

[0031] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the present disclosure.

[0032] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the scope of this disclosure or its application or use. That is, the structures and methods herein are shown in an exemplary manner to illustrate different embodiments of the structures and methods in this disclosure. However, those skilled in the art will understand that they merely illustrate exemplary ways that can be used to implement this disclosure, and not exhaustive ways. Furthermore, the drawings are not necessarily drawn to scale, and some features may be enlarged to show details of specific components.

[0033] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0034] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0035] Figure 1A schematic diagram of a sensory integration network according to some embodiments of the present disclosure is shown, such as... Figure 1 As shown, the integrated sensing network 100 may include a High Altitude Platform Station (HAPS) 110, a drone base station 120, a communication user 130, and a sensing target 140.

[0036] The high-altitude platform station 110, as the central node for data aggregation and command issuance, can establish connections with multiple UAV base stations 120. Here, the high-altitude platform station 110 can establish connections with the corresponding UAV base stations 120 by establishing a backhaul link 150 with the corresponding UAV base station 120.

[0037] As an aerial communication and sensing node, the UAV base station 120 can integrate communication and sensing functions and has high-speed data processing and transmission capabilities. The UAV base station 120 can be configured to collect communication data from its associated communication users 130 and sensing and monitoring data from its associated sensing targets 140, and transmit the collected data to the high-altitude platform station 110 via the backhaul link 150.

[0038] Communication user 130 may include, for example, rescue personnel or disaster victims, and communication user 130 has communication capabilities. In this disclosure, a communication connection with communication user 130 can refer to a communication connection between communication user 130 and a terminal device with communication capabilities. When a communication link 160 is established between drone base station 120 and communication user 130, that is, when drone base station 120 is associated with communication user 130, communication user 130 can transmit data with drone base station 120 and perform functions such as voice communication and video communication through drone base station 120.

[0039] The sensing target 140 may include, for example, a vital signs detection area, an environmental monitoring point, or disaster-stricken people who do not have communication capabilities. When the drone base station 120 establishes a sensing link 170 (indicated by the dashed line in the figure) with the sensing target 140, that is, the drone base station 120 is associated with the sensing target 140, the drone base station 120 can detect the sensing target 140 and collect sensing data through the sensing link 170.

[0040] In some embodiments, the association relationships between the UAV base station 120 and the communication user 130, and between the UAV base station 120 and the sensing target 140 in the integrated sensing network 100 can be obtained in advance. As some implementations, the association relationships between the UAV base station 120 and the communication user 130, and between the UAV base station 120 and the sensing target 140, can be obtained based on the locations of the UAV base station 120, the communication user 130, and the sensing target 140. As some implementations, a preset optimization method can be used to obtain the association relationships between the UAV base station 120 and the communication user 130, and between the UAV base station 120 and the sensing target 140 based on the characteristics of the UAV base station 120, the communication user 130, and the sensing target 140.

[0041] The high-altitude platform station 110 can be configured to receive power allocation signaling requests and related data from various UAV base stations 120 for centralized processing. Based on the processing results, a power allocation command can be generated and sent to the corresponding UAV base station 120 via the backhaul link 150. Thus, the corresponding UAV base station 120 can allocate a corresponding power value to the communication link 160 of the communication user 130 associated with that UAV base station 120, and / or allocate a corresponding power value to the sensing link 170 of the sensing target associated with that UAV base station 120, based on the received power allocation command.

[0042] Figure 2 A schematic flowchart of a power allocation method for a sensing-integrated network according to some embodiments of the present disclosure is shown. Figure 2 As shown, the power allocation method according to some embodiments of this disclosure may include: Step S210: Obtain the first feature of the UAV base station, the second feature of the communication user, and the third feature of the perceived target.

[0043] Here, the first characteristic of a drone base station may include a total power constraint for the drone base station, indicating the maximum allocatable power value of the drone base station. As a power transmitting node, the drone base station can allocate power to its associated links (e.g., associated communication links and associated sensing links) under its total power constraint.

[0044] In some embodiments, the first feature of the drone base station may further include at least one of the following: the location of the drone base station, the current power used, the total backhaul link capacity, and the occupied backhaul link capacity. Here, the location of the drone base station can be represented by three-dimensional spatial coordinates. The occupied backhaul link capacity can indicate the backhaul link capacity already occupied by the drone base station's communication data and sensing data.

[0045] In some embodiments, the second characteristic of a communication user may include at least one of the following: the user's location, a communication rate quality of service (QoS) requirement threshold, and a task priority weighting coefficient. Here, the user's location can be represented by three-dimensional spatial coordinates. The QoS requirement threshold indicates the critical communication rate value required to maintain or enable the user's communication functions. The task priority weighting coefficient indicates the priority level of the user's communication task; for example, the higher the urgency of the rescue task corresponding to the user, the higher the priority level of the communication task. In some embodiments, the priority level of a communication task may correspond to the magnitude of the user's task priority weighting coefficient. Specifically, the user's task priority weighting coefficient can be selected from multiple first priority weighting coefficients, wherein different magnitudes of the multiple first priority weighting coefficients indicate different communication task priority levels. For example, the priority levels of a communication task may include high, medium, and low priority levels, which can be indicated by three first priority weighting coefficients of different magnitudes, respectively.

[0046] In some embodiments, the third characteristic of the sensing target may include at least one of the following: the location of the sensing target, the sensing rate service quality requirement threshold, and the task priority weighting coefficient. Here, the location of the sensing target can be represented by three-dimensional spatial coordinates. The sensing rate service quality requirement threshold can indicate the critical value of the sensing rate used to maintain or achieve the sensing function of the sensing target. The task priority weighting coefficient of the sensing target can indicate the priority level of the sensing task of the sensing target; for example, the higher the disaster level corresponding to the sensing target, the higher the priority level of the sensing task. In some embodiments, the priority level of the sensing task can correspond to the magnitude of the task priority weighting coefficient of the sensing target. Specifically, the task priority weighting coefficient of the sensing target can be selected from multiple second priority weighting coefficients, wherein the magnitudes of the multiple second priority weighting coefficients are different to indicate different sensing task priority levels. For example, the priority level of the sensing task may include two priority levels, high and low, which can be indicated by two second priority weighting coefficients of different magnitudes, respectively.

[0047] In this disclosure, the first feature, the second feature, and the third feature can be represented by vectors.

[0048] For example, the first feature of drone base station n can be represented as ,in: Represents the three-dimensional spatial coordinates of the UAV base station n, for example ; This represents the total power constraint of drone base station n; This indicates the current power used by drone base station n; This represents the total capacity of the wireless backhaul link between the UAV base station n and the high-altitude platform station. This indicates the backhaul link capacity already occupied by the communication and sensing data of drone base station n.

[0049] For example, communication users The second feature can be represented as .in: Indicates communication user The three-dimensional spatial coordinates; Indicates communication user QoS requirement threshold for communication rate; Indicates communication user Task priority weighting coefficient ,in , and These represent the high, medium, and low task priority weights for communication users, respectively. , .

[0050] For example, perceiving the target The third feature can be represented as .in: Representing the perceived target The three-dimensional spatial coordinates; Representing the perceived target The perceived rate QoS requirement threshold; Representing the perceived target Task priority weight coefficient, ,in These represent the high and low task priority weights of the perceived target, respectively. .

[0051] like Figure 2 As shown, the power allocation method according to some embodiments of this disclosure may further include: Step S220: Using a graph neural network (GNN) model, target embedding features of nodes in a heterogeneous graph are obtained based on the first feature, the second feature, and the third feature. Based on the target embedding features of nodes in the heterogeneous graph, the first predicted power allocation weight of the communication user associated with the UAV base station and the second predicted power allocation weight of the perception target associated with the UAV base station are obtained.

[0052] Here, the heterogeneous graph corresponds to the sensor-integrated network. In some embodiments, graph modeling can be performed based on the sensor-integrated network to obtain the heterogeneous graph. UAV base stations, communication users, and sensing targets can constitute nodes in the heterogeneous graph; that is, the heterogeneous graph can have three different types of nodes. Edges in the heterogeneous graph can be used to represent links requiring power allocation. These edges can be obtained based on the associations between each UAV base station and each communication user, and between each UAV base station and each sensing target in the sensor-integrated network. Specifically, edges in the heterogeneous graph can include edges between nodes corresponding to UAV base stations (also referred to as UAV base station nodes in this document) and nodes corresponding to associated communication users (also referred to as communication user nodes in this document), used to indicate communication links requiring power allocation; edges in the heterogeneous graph can also include edges between UAV base station nodes and nodes corresponding to associated sensing targets (also referred to as sensing target nodes in this document), used to indicate sensing links requiring power allocation. Furthermore, the neighboring nodes of each node in the heterogeneous graph can be obtained based on the edges in the heterogeneous graph.

[0053] In some specific examples, the edge set of a heterogeneous graph It can be represented as ,in, This indicates the relationship between the drone base station n and the communication user. There are edges , Represents the relationship between the UAV base station n and the perceived target. There are edges For each edge in a heterogeneous graph, the corresponding edge features can be used to determine the attributes and state of the corresponding link. Specifically, for the edges corresponding to communication links... Edge features It can be represented as ,in This can represent the relationship between the drone base station n and the communication user. Communication link channel gain between This can represent the relationship between the drone base station n and the communication user. The Euclidean distance between them; for the edge corresponding to the sensing link Edge features It can be represented as ,in, This can represent the relationship between the UAV base station n and the perceived target. Sensing link channel gain between This can represent the relationship between the UAV base station n and the perceived target. The Euclidean distance between them.

[0054] In this disclosure, the edge weights between the UAV base station node and the associated communication user node can be used as the corresponding first predicted power allocation weights to indicate the importance of the UAV base station allocating power to the communication link of the corresponding communication user; the edge weights between the UAV base station node and the associated sensing target node can be used as the corresponding second predicted power allocation weights to indicate the importance of the UAV base station allocating power to the sensing link of the corresponding sensing target.

[0055] Figure 3 A schematic diagram of a graph neural network model according to some embodiments of the present disclosure is illustrated by way of example. Figure 3 As shown, in some embodiments, the graph neural network model 300 may include multiple computation layers 310 to perform layer-by-layer updates and computations using the features of nodes in the heterogeneous graph and the features of the node's neighboring nodes. Here, the processing of each computation layer 310 may include an aggregation step and an update step, wherein the aggregation step may be used to collect information about neighboring nodes, and the update step may be used to use the collected information and the computation results of the current node in the previous computation layer 310 to calculate the computation result of the current node in the current computation layer 310.

[0056] Specifically, such as Figure 3 and Figure 4 As shown, using the graph neural network model 300, target embedding features of nodes in the heterogeneous graph are obtained based on the first feature, the second feature, and the third feature. The first predicted power allocation weight for communication users associated with the UAV base station and the second predicted power allocation weight for sensing targets associated with the UAV base station are obtained based on the target embedding features of nodes in the heterogeneous graph. These may include: Step S410: In each of the multi-layer computation layers 310, for each node in the heterogeneous graph, the aggregation result of the node is obtained by using the embedding features of the node's neighboring nodes obtained in the previous computation layer 310. Step S420: Using the embedding features obtained by the node in the previous computational layer 310 and the aggregation result of the node, the embedding features of the node in the current computational layer 310 are obtained. The embedding features of the node in the last computational layer 310 of the multi-layer computational layers are used as the target embedding features of the node.

[0057] Here, in the first computational layer 310 of the multi-layer computational layer 310, the initial embedding features of a node are the embedding features obtained by the node in the previous computational layer 310. The initial embedding features of the node corresponding to the UAV base station are obtained based on the first feature, the initial embedding features of the node corresponding to the communication user are obtained based on the second feature, and the initial embedding features of the node corresponding to the sensing target are obtained based on the third feature. In some embodiments, a multi-layer perceptron (MLP) can be used to obtain the initial embedding features of the UAV base station node based on the first feature of the UAV base station. In some embodiments, a multi-layer perceptron can be used to obtain the initial embedding features of the communication user node based on the second feature of the communication user. In some embodiments, a multi-layer perceptron can be used to obtain the initial embedding features of the sensing target node based on the third feature of the sensing target. Thus, the first feature, the second feature, and the third feature can be input into the multi-layer perceptron for feature encoding and dimension alignment to obtain the initial embedding features of the corresponding nodes.

[0058] In the above embodiments, a graph neural network model is used to effectively mine the implicit power allocation features in the sensory integration network through the neighbor node aggregation and update mechanism and message passing. This allows for the subsequent acquisition of better first and second predicted power allocation weights, and ultimately the optimal power allocation strategy of the sensory integration network.

[0059] In this disclosure, embedded features can be represented by vectors.

[0060] In some embodiments, the aggregation result of node i can be obtained based on the following formula: ;here, This represents the aggregation result obtained based on the embedding features of the neighboring nodes of node i in the computation layer 310 of the previous layer (i.e., the (l-1)th layer); ( ) represents a nonlinear activation function; Represents the set of neighboring nodes of node i; This represents the learnable weight matrix of the current l-th computational layer 310; This represents the embedding feature of neighbor node j obtained in the previous layer (i.e., the (l-1)th layer) computation layer 310.

[0061] Here, the value of l can be from 1 to L, where L can represent the total number of layers in the calculation layer 310. L can represent the initial embedding features of node j. In some embodiments, L can be 2, 3, 4 or other positive integers.

[0062] In some embodiments, the graph neural network model 300 can be a graph attention neural network model. In the calculation of each computation layer 310, the aggregation result of the node can be obtained based on the attention weight coefficients between the node and each of its neighboring nodes, so that message aggregation can be performed more accurately to obtain better embedding features of the node.

[0063] Specifically, in some embodiments, such as Figure 5 As shown, by utilizing the embedding features of a node's neighboring nodes obtained in the previous computational layer 310, the aggregation result for that node can include: Step S510: For each neighboring node of the node, based on the task priority weight vector, the attention weight vector, the embedding features obtained by the node in the previous computation layer 310, and the embedding features obtained by the neighboring node in the previous computation layer 310, the attention weight coefficient between the node and the neighboring node is obtained. The task priority weight vector is obtained based on the task priority weight coefficient of the node and the task priority weight coefficient of the neighboring node. Step S520: Based on the embedding features of each neighboring node obtained in the previous computation layer 310 and the attention weight coefficients between the node and each neighboring node, the aggregation result of the node is obtained.

[0064] Here, the attention weight vector is a trainable model parameter in the graph attention neural network model. By training the graph attention neural network model, the corresponding attention weight vectors in each computational layer of the graph attention neural network model can be obtained.

[0065] In some embodiments, the task priority weight coefficient of the current node and the task priority weight coefficients of its neighboring nodes can be concatenated and mapped to obtain a task priority weight vector. For example, a multilayer perceptron can be used to obtain the task priority weight vector based on the task priority weight coefficients of the current node and its neighboring nodes. As some implementations, when the current node does not have a task priority weight coefficient, for example, when the current node is a drone base station node, the task priority weight coefficients of its neighboring nodes can be mapped to obtain the task priority weight vector. As some implementations, when the neighboring nodes of the current node do not have task priority weight coefficients, for example, when the neighboring nodes of the current node are drone base station nodes, the task priority weight coefficients of the current node can be mapped to obtain the task priority weight vector. As some implementations, when neither the current node nor its neighboring nodes have task priority weight coefficients, for example, when both the current node and its neighboring nodes are drone base station nodes, the task priority weight vector can be a zero vector.

[0066] In some embodiments, the attention weight coefficients between node i and its neighbor j in the l-th computational layer can be obtained based on the following formula. : Here, exp( LeakyReLU represents the exponential function. () represents the linear rectified activation function with leakage. Represents the attention weight vector. express The transpose of , This represents the learnable weight matrix of the l-th computational layer. This represents the embedding features of node i obtained in the previous computational layer (i.e., the (l-1)th layer). This represents the embedding features of neighbor node j obtained in the previous computation layer (i.e., the (l-1)th layer). Let i represent the set of neighboring nodes. This represents the task priority weight vector obtained by mapping the task priority weight coefficient of node i to the task priority weight coefficient of neighboring node j, and || represents the concatenation operation.

[0067] Thus, an attention mechanism can be introduced to obtain the aggregate weights (i.e., attention weight coefficients) of each neighbor node, so as to perform feature weighting based on the corresponding aggregate weights.

[0068] As one implementation method, among the current node's neighboring nodes, those with higher task priority weight coefficients have higher attention weight coefficients. This allows for more accurate message aggregation, resulting in better embedding features for the node.

[0069] In some embodiments, the aggregation result of node i can be obtained based on the following formula: ;here, This represents the aggregation result obtained in the current l-th computation layer; This represents the attention weight coefficient between node i and its neighbor node j in the current layer l computation; ( ) represents a nonlinear activation function; Represents the set of neighboring nodes of node i; This represents the learnable weight matrix of the current l-th computational layer; This represents the embedding feature of neighbor node j obtained in the previous layer (i.e., the (l-1)th layer).

[0070] In some embodiments, obtaining the embedding features of a node in the current computational layer 310 by utilizing the embedding features of the node obtained in the previous computational layer 310 and the aggregation result of the node may include obtaining the embedding features of node i in the current l-th computational layer 310 based on the following formula. : ;here, This represents the embedding feature of node i in the previous layer (i.e., the (l-1)th layer) of computation layer 310. This represents the aggregation result obtained by node i in the current l-th computation layer 310. ( () indicates an update operation. In some embodiments, an update operation includes... After splicing, the data is input into a multilayer perceptron to obtain... .

[0071] Thus, after computation by the Lth computation layer 310, the embedding features of node i in the last (i.e., the Lth) computation layer 310 can be obtained. Thus, the target embedding features of node i can be obtained, i.e. .

[0072] After obtaining the target embedding features of nodes in the heterogeneous graph, a graph neural network model can be used to obtain the first and second predicted power allocation weights based on these features. Specifically, the graph neural network model can be used to obtain the edge weights between the UAV base station node and the communication user node based on the target embedding features of the UAV base station node and the associated communication user node, thus obtaining the first predicted power allocation weight for the communication user. Similarly, the graph neural network model can be used to obtain the edge weights between the UAV base station node and the associated sensing target node based on the target embedding features of the UAV base station node and the associated sensing target node, thus obtaining the second predicted power allocation weight for the sensing target.

[0073] In some embodiments, for each edge in the heterogeneous graph, the edge weight can be obtained based on the target embedding features of the node corresponding to that edge. Thus, for each UAV base station, a first predicted power allocation weight for each communication user associated with it, and a second power allocation weight for each sensing target associated with it, can be obtained.

[0074] like Figure 3As shown, in some embodiments, the graph neural network model 300 may further include a decoding layer 320 disposed after the multi-layer computation layer 310. The decoding layer 320 may be used to obtain the edge weights between the UAV base station node and the communication user node based on the target embedding features of the UAV base station node and the target embedding features of the communication user node to obtain a first predicted power allocation weight. The decoding layer 320 may also be used to obtain the edge weights between the UAV base station node and the sensing target node based on the target embedding features of the UAV base station node and the target embedding features of the sensing target node to obtain a second predicted power allocation weight. In some embodiments, the decoding layer 320 may include a multilayer perceptron.

[0075] In some embodiments, the corresponding predicted power allocation weight can also be obtained based on the task priority weight coefficient of the node.

[0076] Specifically, in some embodiments, using a graph neural network model, target embedding features of nodes in a heterogeneous graph are obtained based on a first feature, a second feature, and a third feature. The first predicted power allocation weight of the communication user associated with the UAV base station and the second predicted power allocation weight of the perception target associated with the UAV base station are obtained based on the target embedding features of the nodes in the heterogeneous graph. This may include obtaining the first predicted power allocation weight based on the target embedding features of the nodes corresponding to the UAV base station, the target embedding features of the nodes corresponding to the communication user, and the task priority weight coefficient of the communication user.

[0077] As some implementation methods, the first predicted power allocation weight can be obtained through the decoding layer 320 based on the target embedding features of the UAV base station node, the target embedding features of the communication user node, and the task priority weight coefficient of the communication user.

[0078] Here, as one implementation method, for each edge corresponding to an established communication link... It can embed target features and communication users of UAV base station n. Target embedding features, communication users Task priority weight coefficient After concatenation, the data is input to the decoding layer 320 to pass through the multilayer perceptron and activation function layer in the decoding layer 320 to obtain the first predicted power allocation weight.

[0079] For example, communication users associated with drone base station n First predicted power allocation weight It can be obtained based on the following formula: ,here, ( This indicates that the Softplus activation function ensures that the output value is positive. This represents the target embedding features of drone base station n. Indicates communication user Target embedding features, Indicates communication user The task priority weight coefficient, || represents the concatenation operation. Indicates will , , The result is obtained by concatenating and inputting it into the MLP.

[0080] In the above embodiments, by using the task priority weight coefficient of the communication user as the input feature of the decoding layer 320, the model can directly perceive the urgency of the corresponding communication task when predicting the corresponding allocation weight, so that the communication links of high-priority communication users can obtain higher allocation weights.

[0081] In some embodiments, using a graph neural network model, the target embedding features of nodes in a heterogeneous graph are obtained based on the first feature, the second feature, and the third feature. The first predicted power allocation weight of the communication user associated with the UAV base station and the second predicted power allocation weight of the sensing target associated with the UAV base station are obtained based on the target embedding features of the nodes in the heterogeneous graph. This may include obtaining the second predicted power allocation weight based on the target embedding features of the node corresponding to the UAV base station, the target embedding features of the node corresponding to the sensing target, and the task priority weight coefficient of the sensing target.

[0082] As some implementation methods, the second predicted power allocation weight can be obtained through the decoding layer 320 based on the target embedding features of the UAV base station node, the target embedding features of the sensing target node, and the task priority weight coefficient of the sensing target.

[0083] Here, as one implementation method, for each edge corresponding to an established sensing link... It can embed target features and perceive targets from UAV base station n. Target embedding features, perceived target Task priority weight coefficient After concatenation, the data is input to the decoding layer 320 to pass through the multilayer perceptron and activation function layer in the decoding layer 320 to obtain the second predicted power allocation weight.

[0084] For example, the perceived target associated with the drone base station n Second predicted power allocation weight It can be obtained based on the following formula: ,here, ( This indicates that the Softplus activation function ensures that the output value is positive. This represents the target embedding features of drone base station n. Representing the perceived target Target embedding features, Representing the perceived target The task priority weight coefficient, || represents the concatenation operation. Indicates will , , The result is obtained by concatenating and inputting it into the MLP.

[0085] In the above embodiments, by using the task priority weight coefficient of the perceived target as the input feature of the decoding layer 320, the model can directly perceive the urgency of the corresponding perceived task when predicting the corresponding allocation weight, so that the perception link of the perceived target with a high priority level can obtain a higher allocation weight.

[0086] After obtaining the first and second predicted power allocation weights, corresponding target power values ​​can be allocated to the corresponding communication users and the corresponding sensing targets under the total power constraint of the UAV base station. Specifically, as follows: Figure 2 As shown, the power allocation method according to some embodiments of this disclosure may further include: Step S230: Determine the first target power value of the communication user associated with the UAV base station based on the total power constraint and the first predicted power allocation weight, and determine the second target power value of the sensing target associated with the UAV base station based on the total power constraint and the second predicted power allocation weight.

[0087] In this way, under the total power constraint of the UAV base station, the first target power value of the corresponding communication user can be determined by using the first predicted power allocation weight obtained by the graph neural network model, and the second target power value of the corresponding sensing target can be determined by using the second predicted power allocation weight obtained by the graph neural network model.

[0088] Considering that the graph neural network model is based on the global perspective of the corresponding integrated sensing network, the first predicted power allocation weight and the second predicted power allocation are obtained. In order to make the power allocation of each UAV base station for each link better meet the corresponding total power constraints and thus improve the accuracy of power allocation, in some embodiments, the first predicted power allocation weight of each communication user associated with the UAV base station and the second predicted power allocation weight of each sensing target associated with the UAV base station can be normalized for each UAV base station.

[0089] Specifically, such as Figure 6As shown, determining the first target power value of the communication user associated with the UAV base station based on the total power constraint and the first predicted power allocation weight, and determining the second target power value of the sensing target associated with the UAV base station based on the total power constraint and the second predicted power allocation weight may include: Step S610: For each UAV base station, based on the first predicted power allocation weight of each communication user associated with the UAV base station and the second predicted power allocation weight of each sensing target associated with the UAV base station, the first predicted power allocation weight of each communication user is normalized to obtain the corresponding first target power allocation weight, and the second predicted power allocation weight of each sensing target is normalized to obtain the corresponding second target power allocation weight. Step S620: For each communication user associated with the UAV base station, determine the first target power value of the communication user based on the total power constraint of the UAV base station and the first target power allocation weight of the communication user. Step S630: For each sensing target associated with the UAV base station, determine the second target power value of the sensing target based on the total power constraint of the UAV base station and the second target power allocation weight of the sensing target.

[0090] As one implementation method, for each UAV base station, a summation operation can be performed on the first predicted power allocation weights of each communication user associated with the UAV base station and the second predicted power allocation weights of each sensing target associated with the UAV base station to obtain a total allocation weight. Then, for each communication user, a first target power allocation weight is obtained based on the total allocation weight and the first predicted power allocation weight of that communication user; and for each sensing target, a second target power allocation weight is obtained based on the total allocation weight and the second predicted power allocation weight of that sensing target.

[0091] For example, the target power allocation weight of node j (the sensing target or communication user) associated with UAV base station n. It can be obtained based on the following formula: ,here, This represents the set of nodes associated with drone base station n, i.e., the communication users associated with drone base station n. and the sensed targets associated with the drone base station n. The set, Let represent the predicted power allocation weight of node j associated with UAV base station n, where j is a communication user. This represents the corresponding first predicted power allocation weight, where j is the sensing target. This represents the corresponding weight for the second predicted power allocation.

[0092] As one implementation method, for a drone base station n, the target power value is assigned to the associated node j. It can be obtained based on the following formula: ;here, Assign weights to the target power of node j associated with UAV base station n. Where j is a communication user, This represents the corresponding first target power allocation weight, where j is the sensing target. This indicates the corresponding weight for the second target power allocation.

[0093] In some embodiments, before determining the first target power value of the communication user associated with the UAV base station based on the total power constraint and the first predicted power allocation weight, and before determining the second target power value of the sensing target associated with the UAV base station based on the total power constraint and the second predicted power allocation weight, the first predicted power allocation weight may be updated. Specifically, after obtaining the first predicted power allocation weight and the second predicted power allocation weight using a graph neural network model, and before normalizing the first predicted power allocation weight and the second predicted power allocation weight, the first predicted power allocation weight may be updated based on the task priority weight coefficient of the communication user, and the updated first predicted power allocation weight may be used as the first predicted power allocation weight of the communication user; similarly, the second predicted power allocation weight may be updated based on the task priority weight coefficient of the sensing target, and the updated second predicted power allocation weight may be used as the second predicted power allocation weight of the sensing target.

[0094] As one implementation method, the corresponding predicted power allocation weights can be updated based on the following formula: ,here, This represents the predicted power allocation weight of node j associated with UAV base station n; This represents an adjustable exponential factor, where, It can be greater than or equal to 1; This represents the task priority weight coefficient of node j; This represents the updated predicted power allocation weights. The updated predicted power allocation weights are obtained after... Afterwards, it can be made = The updated first predicted power allocation weight will be used as the first predicted power allocation weight for the communication user, and the updated second predicted power allocation weight will be used as the second predicted power allocation weight for the sensing target.

[0095] In the above embodiments, updating the predicted power allocation weight based on the task priority weight coefficient can further improve the allocation weight of nodes with higher priority levels, thereby better meeting the business needs of nodes with higher priority levels.

[0096] After obtaining the first target power value and the second target power value, the drone base station can be controlled to allocate power to the corresponding link.

[0097] Specifically, in some embodiments, the power allocation method according to some embodiments of this disclosure may further include: for each UAV base station, controlling the UAV base station to allocate a corresponding first target power value to the communication link of the communication user associated with the UAV base station, and controlling the UAV base station to allocate a corresponding second target power value to the sensing link of the sensing target associated with the UAV base station.

[0098] In this way, based on the power allocation optimization scheme determined by the graph neural network model, each UAV base station can be controlled to allocate the corresponding target power value to the corresponding link.

[0099] In some embodiments, where power values ​​have already been allocated to the corresponding links of each UAV base station, the power values ​​of the communication links between each UAV base station and each communication user, as well as the power values ​​of the sensing links between each UAV base station and each sensing target, can be updated based on the current target power value obtained using the graph neural network model. As some implementations, network status (e.g., actual link speeds, UAV base station power consumption) can be continuously monitored to periodically or based on events (e.g., QoS non-compliance, node location updates, sudden changes in backhaul load) trigger the re-run of the power allocation method, thereby achieving intelligent dynamic power allocation in the integrated sensing network.

[0100] Considering that the calculations using graph neural network models take a certain amount of time, in order to reduce the possibility of service interruption due to computation time, in some embodiments, an initial power allocation can be established to allocate initial power values ​​to the corresponding links in order to execute the corresponding services, and then the power values ​​of each link can be readjusted and allocated after obtaining the first target power value and the second target power value by using the graph neural network model.

[0101] Specifically, such as Figure 7 As shown, the power allocation method according to some embodiments of this disclosure may further include: Step S710: For each UAV base station, obtain the first initial power value of each communication user associated with the UAV base station and the second initial power value of each sensing target associated with the UAV base station. Step S720: For each communication user, control the UAV base station to allocate a corresponding first initial power value to the communication link of the communication user; for each sensing target, control the UAV base station to allocate a corresponding second initial power value to the sensing link of the sensing target. In step S730, in response to the completion of determining the first target power value of each communication user and the second target power value of each sensing target, the UAV base station is controlled to adjust the power value of the communication link of each communication user from the corresponding first initial power value to the corresponding first target power value, and the UAV base station is controlled to adjust the power value of the sensing link of each sensing target from the corresponding second initial power value to the corresponding second target power value.

[0102] Thus, before using a graph neural network model to obtain the target embedding features of nodes in the heterogeneous graph based on the first, second, and third features, and before obtaining the first predicted power allocation weights for communication users associated with the UAV base station and the second predicted power allocation weights for sensing targets associated with the UAV base station based on the target embedding features of nodes in the heterogeneous graph, corresponding initial power values ​​can be allocated to the relevant links of each UAV base station based on the initial power allocation scheme. Then, after obtaining the first and second target power values, the power values ​​of the corresponding links are adjusted.

[0103] In some embodiments, for each UAV base station, equal power allocation can be performed on each communication user associated with the UAV base station and each sensing target associated with the UAV base station based on the total power constraint of the UAV base station, so as to obtain a first initial power value for each communication user and a second initial power value for each sensing target.

[0104] In other embodiments, initial power allocation can be based on link channel state and corresponding task priority weights, so that the initially allocated power value is more likely to meet the needs of each communication task and / or each sensing task.

[0105] Specifically, obtaining the first initial power value of each communication user associated with the UAV base station and the second initial power value of each sensing target associated with the UAV base station includes: obtaining the communication link channel gain of each communication user and the sensing link channel gain of each sensing target; for each communication user, obtaining the first initial allocation coefficient of the communication user based on the first total allocation coefficient and the first sub-allocation coefficient of the communication user, and obtaining the first initial power value of the communication user based on the total power constraint of the UAV base station and the first initial allocation coefficient of the communication user, wherein the first sub-allocation coefficient is obtained based on the communication link channel gain and the task priority weight coefficient of the communication user, and the first total allocation coefficient is obtained by summing the first sub-allocation coefficients of each communication user; for each sensing target, obtaining the second initial allocation coefficient of the sensing target based on the second total allocation coefficient and the second sub-allocation coefficient of the sensing target, and obtaining the second initial power value of the sensing target based on the total power constraint of the UAV base station and the second initial allocation coefficient of the sensing target, wherein the second sub-allocation coefficient is obtained based on the sensing link channel gain and the task priority weight coefficient of the sensing target, and the second total allocation coefficient is obtained by summing the second sub-allocation coefficients of each sensing target.

[0106] As one implementation method, the communication user associated with the drone base station n First initial power value It can be obtained based on the following formula: ;here, Indicates communication user Task priority weight coefficient, This indicates the relationship between the drone base station n and the communication user. Communication link channel gain between This represents the total power constraint of drone base station n. Let represent the set of edges corresponding to the communication links of UAV base station n, where This is the first total allocation coefficient. That is, communication user The first sub-allocation coefficient, This is the first initial allocation coefficient.

[0107] As one implementation method, the sensing target associated with the drone base station n Second initial power value It can be obtained based on the following formula ;here, Representing the perceived target Task priority weight coefficient, Represents the relationship between the UAV base station n and the perceived target. Sensing link channel gain between This represents the total power constraint of drone base station n. Let represent the set of edges corresponding to the sensing links of UAV base station n, where This is the second total allocation coefficient. That is, the target of perception The second sub-allocation coefficient, This is the second initial allocation coefficient.

[0108] In this way, initial power allocation can be performed based on the channel gain of the corresponding link and the corresponding task priority weight.

[0109] In the above embodiments, considering that the computation using the graph neural network model takes time, a fast and low-complexity initial power allocation strategy is adopted to allocate corresponding initial power to all established links. This achieves "allocation first, optimization later," meaning that initial power is allocated to each link in the shortest possible time to enable the execution of corresponding tasks, and then the power is redistributed to each link using an optimized power allocation strategy based on the graph neural network model. This is especially useful in post-disaster scenarios where rapid execution of communication or sensing services is often required. This approach enables rapid response and execution of services, reduces the possibility of prolonged service interruptions, and improves the efficiency of communication and sensing service execution.

[0110] According to another aspect of this disclosure, a method for training a graph neural network model is also provided. It should be understood that the graph neural network model training method according to some embodiments of this disclosure can be used as a method applied in any of the above embodiments of the power allocation method to determine a first target power value of a communication user associated with a UAV base station and a second target power value of a sensing target associated with a UAV base station. That is, any of the above embodiments of the power allocation method may include the graph neural network model training method according to some embodiments of this disclosure; or, the graph neural network model training method according to some embodiments of this disclosure can also be used as a training method independent of any of the above embodiments of the power allocation method.

[0111] like Figure 8 As shown, a method for training a graph neural network model according to some embodiments of this disclosure may include: Step S810: Obtain training samples. The feature part of the training samples includes the first feature of the sample UAV base station in the sample sensory integration network, the second feature of the sample communication user, and the third feature of the sample perception target. The first feature of the sample includes the total power constraint of the sample UAV base station. The label part of the training samples includes the first sample power allocation weight of the sample communication user associated with the sample human-machine base station and the second sample power allocation weight of the sample perception target associated with the sample human-machine base station. Step S820: Train the graph neural network model based on the training samples.

[0112] Here, the graph neural network model can have corresponding neural network layers (e.g. Figure 3 The computation layer 310 and decoding layer 320 shown can update the learnable parameters of the corresponding neural network layers by training the graph neural network model, so that the trained graph neural network model can obtain the target embedding features of the nodes in the heterogeneous graph corresponding to the synesthetic network based on the first feature, the second feature and the third feature, and obtain the first prediction power allocation weight and the second prediction power allocation weight based on the target embedding features of the nodes in the heterogeneous graph.

[0113] In some embodiments, a graph neural network model can be trained based on training samples and a target loss function. Specifically, the graph neural network model can be trained based on training samples with the goal of reducing the value of the target loss function. For example, appropriate optimization algorithms can be used to update the parameters of the graph neural network model to minimize the value of the target loss function, thereby completing the training of the graph neural network model.

[0114] In some embodiments, the target loss function may be set based on the mean squared error (MSE) between the first sample power allocation weights of the sample communication user and the first predicted power allocation weights of the sample communication user obtained using a graph neural network model, and the mean squared error between the second sample power allocation weights of the sample perception target and the second predicted power allocation weights of the sample perception target obtained using a graph neural network model.

[0115] As one implementation, the target loss function may include one or more sub-loss functions, each corresponding to an edge in the heterogeneous graph corresponding to the sample synesthetic network. Each sub-loss function is configured based on the mean square error between the predicted power assignment weight of the corresponding edge and the sample power label of the corresponding edge, as well as the task priority weight of the corresponding edge. Here, the edges in the heterogeneous graph corresponding to the sample synesthetic network include edges between sample UAV base station nodes and associated sample communication user nodes, and edges between sample UAV base station nodes and associated sample sensing target nodes.

[0116] For example, the target loss function It can be set based on the following formula: ; here, This represents the edges in the heterogeneous graph corresponding to the sample synesthetic network. Indicates the number of edges. Let the edge (i,j) between node i and node j be a candidate associated edge. Let (i,j) be the task priority weight coefficient, where... It can be determined based on the task priority weight coefficient of the communication user connected by edge (i,j) or the task priority weight coefficient of the sensing target. Assign weights (e.g., first sample power assignment weights or second sample power assignment weights) to the sample power corresponding to edge (i,j) in the label part of the training sample. The predicted power allocation weights (e.g., the first predicted power allocation weights or the second predicted power allocation weights) are obtained based on the graph neural network model and correspond to the edge (i,j).

[0117] In some embodiments according to this disclosure, the training of a graph neural network model can be based on a Markov Decision Process (MDP).

[0118] Specifically, the state at time t The graph state at time t can be set, where the graph state can include the characteristics of each node in the integrated sensing network. Here, when the node is a UAV base station node, the node's characteristics can be the first characteristic; when the node is a communication user node, the node's characteristics can be the second characteristic; and when the node is a sensing target node, the node's characteristics can be the third characteristic.

[0119] Action at time t It can be set as the predicted power allocation weight or the target power allocation weight of the node associated with the UAV base station obtained based on the graph neural network model. Here, the target power allocation weight can be obtained, for example, by normalizing the predicted power allocation weight.

[0120] Reward function at time t It can be based on the state implement The rewards are set based on the communication performance bonus, perception performance bonus, priority bonus, backhaul load penalty, and total power penalty.

[0121] In some embodiments, the reward function It can be set based on the following formula: .

[0122] here, and The weights are adjustable, meaning they can be adjusted according to the specific scenario. Used to reward improved communication and perception performance Used to reward high-priority task guarantees. Used to punish return overload. Used to penalize total system power consumption, in some embodiments It can be greater than .

[0123] For priority-weighted communication rates, For priority-weighted sensing rates, where, and They represent Time Communication Users Communication rate and sensing target The rate of perception.

[0124] To ensure that the indicator function is satisfied for high-priority tasks, it can be set as follows: .in, For indicator functions, if If the expression inside is true, then ,otherwise . and Let these represent the set of communication users with high task priority and the set of sensing targets with high task priority, respectively. For example, This refers to the set of communication users whose task priority weight coefficient is greater than or equal to the specified weight coefficient. This refers to the set of perceived targets whose task priority weight coefficient is greater than or equal to the specified weight coefficient. Indicates communication user The communication rate QoS requirement threshold, Representing the perceived target The perceived rate QoS requirement threshold.

[0125] The return load function can be set as follows: . This can represent the communication user at time t. The backhaul link capacity occupied by the communication data This can represent the perception of the target at time t. The backhaul link capacity occupied by the sensing data It can represent the backhaul link load capacity of drone base station n at time t, and N can represent the total number of drone base stations in the integrated sensing network.

[0126] This can represent the total power consumption of the system at time t, specifically, It can be set to: , Let n be the set of unmanned aerial vehicle (UAV) base stations. Let be the power used by drone base station n at time t. This is achieved through the reward function. Settings This initiative can encourage the efficient use of power resources and improve system energy efficiency.

[0127] Thus, the parameters of a graph neural network can be updated based on a Markov decision process with the goal of maximizing the cumulative reward obtained from the reward function. As some implementation methods, policy gradient algorithms (such as PPO and DDPG) can be used to continuously update the parameters of the GNN.

[0128] In the above embodiments, based on Markov decision processes, the graph neural network model can make high-reward decision-making power allocation strategies under various complex network states, that is, output high-reward decision power allocation results. This end-to-end online training method can eliminate the need for manual annotation, such as manually annotating power allocation weight labels, thus reducing labor costs.

[0129] Embodiments of this disclosure also provide a power distribution device for a sensor-integrated network, which can be configured to perform the power distribution method of any of the above embodiments.

[0130] In some embodiments, the power distribution apparatus may include a module for performing the power distribution method of any of the above embodiments.

[0131] Figure 9 A schematic diagram of a power distribution apparatus according to some embodiments of the present disclosure is shown. For example... Figure 9 As shown, the power distribution device 900 according to some embodiments of the present disclosure may include an acquisition module 910, a weight determination module 920 and a power determination module 930.

[0132] The acquisition module 910 can be configured to acquire a first feature of the UAV base station, a second feature of the communication user, and a third feature of the sensing target, wherein the first feature includes the total power constraint of the UAV base station. The weight determination module 920 can be configured to use a graph neural network model to obtain target embedding features of nodes in a heterogeneous graph based on the first, second, and third features, and to obtain a first predicted power allocation weight for the communication user associated with the UAV base station and a second predicted power allocation weight for the sensing target associated with the UAV base station based on the target embedding features of the nodes in the heterogeneous graph. The heterogeneous graph corresponds to a sensing-interactive network, and the UAV base station, communication user, and sensing target constitute nodes in the heterogeneous graph. The edges in the heterogeneous graph include edges between UAV base station nodes and associated communication user nodes, and edges between UAV base station nodes and associated sensing target nodes. The power determination module 930 can be configured to determine a first target power value of the communication user associated with the UAV base station based on the total power constraint and the first predicted power allocation weight, and to determine a second target power value of the sensing target associated with the UAV base station based on the total power constraint and the second predicted power allocation weight.

[0133] In some embodiments, the power distribution device 900 may further include other modules to perform the power distribution method of any of the above embodiments.

[0134] The specific operation of the various modules in the power distribution device 900 can be found in the detailed explanation of the power distribution method above, and will not be repeated here.

[0135] Figure 10 A schematic diagram of a power distribution device for a sensing-integrated network according to some embodiments of the present disclosure is shown. Figure 10 As shown, the power distribution device 1000 may include a memory 1010 and a processor 1020 coupled to the memory 1010. The processor 1020 may be configured to execute the power distribution method of any of the foregoing embodiments based on instructions stored in the memory 1010.

[0136] Specifically, processor 1020 can perform various actions and processes according to instructions stored in memory 1010. Processor 1020 can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. The general-purpose processor can be a microprocessor or any conventional processor, and can be an x86 architecture or an ARM architecture, etc.

[0137] Memory 1010 stores executable instructions that, when executed by processor 1020, implement the power allocation method described above. Memory 1010 may be volatile memory or non-volatile memory, or may include both. Non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory may be random access memory (RAM), which serves as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (SLDRAM), and direct memory bus random access memory (DR RAM). It should be noted that the memory of the methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0138] Embodiments of this disclosure also provide an aerial platform station. In some embodiments, the aerial platform station can be configured to perform the power distribution method described in any of the above embodiments. As some implementations, the aerial platform station may include the power distribution device 900 described in any of the above embodiments. As some implementations, the aerial platform station may include the power distribution device 1000 described in any of the above embodiments.

[0139] Embodiments of this disclosure also provide a sensing-integrated network. In some embodiments, the sensing-integrated network may include the high-altitude platform station described above. In some embodiments, the sensing-integrated network may further include unmanned aerial vehicle (UAV) base stations, communication users, and sensing targets.

[0140] This disclosure also proposes a non-transitory computer-readable storage medium storing instructions that, when executed by a processor, can implement the steps of the power allocation method described above.

[0141] Similarly, the non-transitory computer-readable storage media in the embodiments of this disclosure are intended to include, but are not limited to, the above and any other suitable types of memory.

[0142] This disclosure also proposes a computer program product that may include instructions that, when executed by a processor, implement the steps of the power allocation method as described above.

[0143] Instructions can be any set of instructions that will be executed directly by one or more processors, such as machine code, or any set of instructions that will be executed indirectly, such as a script. The terms “instruction,” “application,” “procedure,” “step,” “program,” and “computer program” used herein are used interchangeably. Instructions can be stored in object code format for direct processing by one or more processors, or stored in any other computer language, including scripts or sets of independent source code modules that are interpreted on demand or compiled in advance. The function, methods, and routines of instructions are explained in more detail in other parts of this document.

[0144] Figure 11A schematic block diagram of a computer system 1100 on which embodiments of the present disclosure may be implemented is shown. The computer system 1100 includes a bus 1110 or other communication mechanism for transmitting information, and a processing means 1120 coupled to the bus 1110 for processing information. The computer system 1100 also includes a storage means 1130 coupled to the bus 1110 for storing instructions to be executed by the processing means 1120. The storage means 1130 may include random access memory (RAM) or other dynamic storage devices. The storage means 1130 may be used to store temporary variables or other intermediate information during the execution of instructions to be executed by the processing means 1120. The storage means 1130 may include a read-only memory (ROM) or other static storage device for storing static information and instructions for the processing means 1120. The storage means 1130 may include means such as a magnetic disk or optical disk for storing information and instructions. Computer system 1100 may be coupled via bus 1110 to output device 1140 for providing output to a user, such as, but not limited to, a display (such as a cathode ray tube (CRT) or liquid crystal display (LCD)), speakers, etc. Input device 1150, such as a keyboard, mouse, microphone, etc., is coupled to bus 1110 for transmitting information and command selections to processing device 1120. Computer system 1100 may execute embodiments of this disclosure. Consistent with certain implementations of this disclosure, computer system 1100 provides results by executing one or more sequences of one or more instructions contained in storage device 1130 in response to processing device 1120. Such instructions may be read into storage device 1130 from another computer-readable medium, such as storage device 1130. Execution of the sequence of instructions contained in storage device 1130 causes processing device 1120 to perform the methods described herein. Alternatively, hard-wired circuitry may be used in place of or in combination with software instructions to implement the teachings. Therefore, implementations of this disclosure are not limited to any particular combination of hardware circuitry and software. In various embodiments, computer system 1100 can be connected across a network to one or more other computer systems, such as computer system 1100, to form a networked system via network interface 1160. This network may include a private network or a public network such as the Internet. In a networked system, one or more computer systems may store data and supply data to other computer systems. As used herein, the term "computer-readable medium" refers to any medium that participates in providing instructions to processing device 1120 for execution. Such media can take many forms, including but not limited to non-volatile media, volatile media, and transmission media. Non-volatile media include, for example, optical discs or magnetic disks. Volatile media include dynamic memory such as RAM. Transmission media include coaxial cables, copper wires, and optical fibers, including wiring that includes bus 1110.Common forms of computer-readable media or computer program products include, for example, floppy disks, flexible disks, hard disks, magnetic tapes, or any other magnetic media, CD-ROMs, digital video discs (DVDs), Blu-ray discs, any other optical media, thumb drives, memory cards, RAM, PROMs and EPROMs, fast EPROMs, any other memory chips or cartridges, or any other tangible media from which a computer can read. Various forms of computer-readable media may be involved when carrying one or more sequences of one or more instructions to processing device 1120 for execution. For example, instructions may initially be carried on a disk of a remote computer. The remote computer may load the instructions into its dynamic memory and transmit the instructions over a telephone line using a modem. A modem local to computer system 1100 may receive data over a telephone line and convert the data into an infrared signal using an infrared transmitter. An infrared detector coupled to bus 1110 may receive the data carried in the infrared signal and place the data on bus 1110. Bus 1110 carries the data to storage device 1130, from which processing device 1120 retrieves and executes the instructions. Optionally, the instructions received by the storage device 1130 may be stored on the storage device 1130 before or after they are executed by the processing device 1120.

[0145] According to various embodiments, instructions configured to be executed by processing device 920 to perform a method are stored on a computer-readable medium. The computer-readable medium may be a device for storing digital information. For example, the computer-readable medium includes a compact disc read-only memory (CD-ROM) as known in the art for storing software. The computer-readable medium is accessed by a processor adapted to execute the instructions configured to be executed.

[0146] In this disclosed technical solution, communication users, sensing targets, and UAV base stations are uniformly modeled as heterogeneous graph nodes. Edges are constructed between "UAV base stations and communication users" and "UAV base stations and sensing targets." Utilizing the message passing and aggregation mechanism of graph neural networks, the complex power competition and coordination relationships between various communication links and sensing links are explored. This overcomes the limitations of traditional solutions that allocate power to communication and sensing links according to a preset fixed power ratio or optimize them separately. It achieves multi-dimensional joint power scheduling and globally optimal power allocation across services and links, effectively optimizing overall system energy consumption and improving the efficiency of related task processing. This significantly enhances the system's task completion capability and reliability in emergency scenarios. According to this disclosed solution, it has strong generalization ability, robustness, and flexibility, and can adapt to complex and dynamically changing integrated communication and sensing networks to achieve intelligent power allocation under dynamic service distribution and channel fluctuations. Furthermore, according to some embodiments of this disclosure, the power allocation weight is optimized by utilizing task priority weights. When the total system power is limited, power resources can be intelligently tilted towards high-urgency tasks, ensuring the lower limit of the performance of critical communication and sensing services. This effectively solves the problems of "egalitarianism" and improper power resource allocation caused by uniform or static allocation.

[0147] The embodiments disclosed herein can be applied to electronic devices such as terminal devices, computer systems, and servers, and can operate together with a wide range of other general-purpose or special-purpose computing system environments or configurations. Examples of well-known terminal devices, computing systems, environments, and / or configurations suitable for use with electronic devices such as terminal devices, computer systems, and servers include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputer systems, mainframe computer systems, and distributed cloud computing environments including any of the above systems, etc.

[0148] In this disclosure, electronic devices such as terminal devices, computer systems, and servers can be described in the general context of computer system executable instructions (such as program modules) executed by a computer system. Typically, program modules can include routines, programs, object programs, components, logic, data structures, etc., which perform specific tasks or implement specific abstract data types. Computer systems / servers can be implemented in a distributed cloud computing environment, where tasks are performed by remote processing devices linked via a communication network. In a distributed cloud computing environment, program modules can reside on local or remote computing system storage media, including storage devices.

[0149] The technical solutions of the various embodiments of this application can be applied to various existing or future-developed communication systems, such as: 5th generation (5G) or new radio (NR) systems, long term evolution (LTE) systems, LTE frequency division duplex (FDD) systems, LTE time division duplex (TDD) systems, and sixth generation mobile communication systems.

[0150] The technical solutions provided in this application can also be applied to device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, machine-to-machine (M2M) communication, machine-type communication (MTC), and Internet of Things (IoT) communication systems or other communication systems.

[0151] In this disclosure, terminal equipment, also referred to as user terminal, terminal, etc., is a device with wireless transceiver capabilities that can communicate with one or more core networks (CNs) via access network devices in a radio access network (RAN). Terminal equipment can be deployed on land, including indoors or outdoors, handheld, wearable, or vehicle-mounted; it can also be deployed on water, such as on ships; and it can be deployed in the air, such as on airplanes, balloons, or satellites. Terminal equipment can be a mobile phone, tablet computer, computer with wireless transceiver capabilities, virtual reality (VR) terminal equipment, augmented reality (AR) terminal equipment, wireless terminals in industrial control, wireless terminals in self-driving, wireless terminals in remote medical care, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, wireless terminals in smart homes, etc.

[0152] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0153] In general, the various exemplary embodiments of this disclosure can be implemented in hardware or dedicated circuitry, software, firmware, logic, or any combination thereof. Some aspects can be implemented in hardware, while others can be implemented in firmware or software that can be executed by a controller, microprocessor, or other computing device. When aspects of embodiments of this disclosure are illustrated or described as block diagrams, flowcharts, or using some other graphical representation, it will be understood that the blocks, apparatuses, systems, techniques, or methods described herein can be implemented as non-limiting examples in hardware, software, firmware, dedicated circuitry or logic, general-purpose hardware or controllers or other computing devices, or some combination thereof.

[0154] As used herein, the term “exemplary” means “used as an example, instance, or illustration” and not as a “model” to be exactly copied. Any implementation described herein by example is not necessarily to be construed as preferred or advantageous over other implementations.

[0155] Furthermore, terms such as “first,” “second,” etc., may be used in this document for reference purposes only and are not intended to be limiting. For example, unless the context clearly indicates otherwise, the words “first,” “second,” and other such numerical terms relating to structures or elements do not imply order or sequence.

[0156] It should also be understood that when the term “including / contains” is used herein, it indicates the presence of the indicated feature, whole, step, operation, unit and / or component, but does not preclude the presence or addition of one or more other features, wholes, steps, operations, units and / or components and / or combinations thereof.

[0157] In this disclosure, the term “provide” is used broadly to cover all ways of obtaining an object, and therefore “provide an object” includes, but is not limited to, “purchasing,” “preparing / manufacturing,” “arranging / setting up,” “installing / assembling,” and / or “ordering” an object.

[0158] As used herein, the term “and / or” includes any and all combinations of one or more of the listed items in association. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of this disclosure. As used herein, the singular forms “a,” “an,” and “the” are also intended to include the plural forms unless the context clearly indicates otherwise.

[0159] Those skilled in the art will recognize that the boundaries between the above operations are merely illustrative. Multiple operations may be combined into a single operation, a single operation may be distributed among additional operations, and operations may be performed with at least partial overlap in time. Moreover, alternative embodiments may include multiple instances of a particular operation, and the order of operations may be changed in various other embodiments. However, other modifications, variations, and substitutions are equally possible. Aspects and elements of all the embodiments disclosed above may be combined in any way and / or in combination with aspects or elements of other embodiments to provide multiple additional embodiments. Therefore, this specification and the accompanying drawings should be considered illustrative rather than restrictive.

[0160] While specific embodiments of this disclosure have been described in detail by way of example, those skilled in the art should understand that the examples are for illustrative purposes only and not intended to limit the scope of this disclosure. The various embodiments disclosed herein can be combined in any way without departing from the spirit and scope of this disclosure. Those skilled in the art should also understand that various modifications can be made to the embodiments without departing from the scope and spirit of this disclosure. The scope of this disclosure is defined by the appended claims.

Claims

1. A power allocation method for a sensor-integrated network, comprising: Acquire a first feature of the UAV base station, a second feature of the communication user, and a third feature of the perceived target, wherein the first feature includes the total power constraint of the UAV base station; Using a graph neural network model, target embedding features of nodes in a heterogeneous graph are obtained based on the first feature, the second feature, and the third feature. Based on the target embedding features of nodes in the heterogeneous graph, a first predicted power allocation weight for communication users associated with the UAV base station and a second predicted power allocation weight for sensing targets associated with the UAV base station are obtained. The heterogeneous graph corresponds to the integrated sensing network. The UAV base station, communication users, and sensing targets constitute nodes in the heterogeneous graph. The edges in the heterogeneous graph include the edges between UAV base station nodes and associated communication user nodes, and the edges between UAV base station nodes and associated sensing target nodes. The first target power value of the communication user associated with the UAV base station is determined based on the total power constraint and the first predicted power allocation weight, and the second target power value of the sensing target associated with the UAV base station is determined based on the total power constraint and the second predicted power allocation weight.

2. The power allocation method according to claim 1, wherein, The first feature also includes at least one of the following: the location of the drone base station, the current power used, the total backhaul link capacity, and the backhaul link capacity already occupied; The second feature includes at least one of the following: the location of the communication user, the communication rate quality of service requirement threshold, and the task priority weight coefficient; The third feature includes at least one of the following: the location of the perceived target, the perception rate service quality requirement threshold, and the task priority weight coefficient.

3. The power distribution method according to claim 1, wherein, The graph neural network model includes multiple computational layers. Using the graph neural network model, target embedding features of nodes in the heterogeneous graph are obtained based on the first feature, the second feature, and the third feature. Based on the target embedding features of nodes in the heterogeneous graph, a first predicted power allocation weight for communication users associated with the UAV base station and a second predicted power allocation weight for sensing targets associated with the UAV base station are obtained, including: In each of the multi-layered computational layers, for each node in the heterogeneous graph, the aggregation result of the node is obtained by using the embedding features of the node's neighboring nodes obtained in the previous computational layer. By utilizing the embedding features obtained by the node in the previous computational layer and the aggregation result of the node, the embedding features of the node in this computational layer are obtained. Among them, the embedding features of the node in the last computational layer of the multi-layered computational system are used as the target embedding features of the node. In the first computing layer of the multi-layer computing layers, the initial embedding feature of a node is used as the embedding feature obtained by the node in the previous computing layer. The initial embedding feature of the UAV base station node is obtained based on the first feature, the initial embedding feature of the communication user node is obtained based on the second feature, and the initial embedding feature of the sensing target node is obtained based on the third feature.

4. The power distribution method according to claim 3, wherein, The graph neural network model is a graph attention neural network model. It utilizes the embedding features of the node's neighboring nodes obtained in the previous computational layer to obtain the aggregation result for that node, including: For each neighboring node of the node, the attention weight coefficient between the node and the neighboring node is obtained based on the task priority weight vector, the attention weight vector, the embedding features obtained by the node in the previous computation layer, and the embedding features obtained by the neighboring node in the previous computation layer. The task priority weight vector is obtained based on the task priority weight coefficient of the node and the task priority weight coefficient of the neighboring node. Based on the embedding features of each neighboring node obtained in the previous computation layer and the attention weight coefficients between the node and each neighboring node, the aggregation result of the node is obtained.

5. The power distribution method according to claim 1, wherein, Using a graph neural network model, target embedding features of nodes in a heterogeneous graph are obtained based on the first feature, the second feature, and the third feature. Based on these target embedding features, a first predicted power allocation weight for communication users associated with a UAV base station and a second predicted power allocation weight for sensing targets associated with a UAV base station are obtained, including: Based on the target embedding features of the node corresponding to the UAV base station, the target embedding features of the node corresponding to the associated communication user, and the task priority weight coefficient of the communication user, the first predicted power allocation weight is obtained; and / or The second predicted power allocation weight is obtained based on the target embedding features of the node corresponding to the UAV base station, the target embedding features of the node corresponding to the associated sensing target, and the task priority weight coefficient of the sensing target.

6. The power allocation method according to claim 1, wherein, Determining a first target power value for a communication user associated with a UAV base station based on the total power constraint and the first predicted power allocation weight, and determining a second target power value for a sensing target associated with a UAV base station based on the total power constraint and the second predicted power allocation weight, includes: For each UAV base station, the first predicted power allocation weight of each communication user associated with the UAV base station and the second predicted power allocation weight of each sensing target associated with the UAV base station are normalized to obtain the corresponding first target power allocation weight, and the second predicted power allocation weight of each sensing target is normalized to obtain the corresponding second target power allocation weight. For each communication user associated with the UAV base station, the first target power value of the communication user is determined according to the total power constraint of the UAV base station and the first target power allocation weight of the communication user; For each sensing target associated with the UAV base station, the second target power value of the sensing target is determined according to the total power constraint of the UAV base station and the second target power allocation weight of the sensing target.

7. The power distribution method according to claim 1, wherein, The power allocation method further includes: Before determining the first target power value of the communication user associated with the UAV base station based on the total power constraint and the first predicted power allocation weight, and before determining the second target power value of the sensing target associated with the UAV base station based on the total power constraint and the second predicted power allocation weight, the first predicted power allocation weight is updated based on the task priority weight coefficient of the communication user, and the updated first predicted power allocation weight is used as the first predicted power allocation weight of the communication user; and the second predicted power allocation weight is updated based on the task priority weight coefficient of the sensing target, and the updated second predicted power allocation weight is used as the second predicted power allocation weight of the sensing target.

8. The power distribution method according to claim 1, wherein, The power allocation method further includes: For each drone base station, the system controls the drone base station to allocate a corresponding first target power value to the communication link of the communication user associated with the drone base station, and controls the drone base station to allocate a corresponding second target power value to the perception link of the perception target associated with the drone base station.

9. The power distribution method according to claim 1, wherein, The power allocation method further includes: For each drone base station, obtain the first initial power value of each communication user associated with the drone base station and the second initial power value of each sensing target associated with the drone base station; For each communication user, the UAV base station is controlled to allocate a corresponding first initial power value to the communication link of the communication user; for each sensing target, the UAV base station is controlled to allocate a corresponding second initial power value to the sensing link of the sensing target. In response to the determination of the first target power value of each communication user and the second target power value of each sensing target, the UAV base station is controlled to adjust the power value of the communication link of each communication user from the corresponding first initial power value to the corresponding first target power value, and the UAV base station is controlled to adjust the power value of the sensing link of each sensing target from the corresponding second initial power value to the corresponding second target power value.

10. The power distribution method according to claim 9, wherein, Obtaining the first initial power value of each communication user associated with the UAV base station and the second initial power value of each sensing target associated with the UAV base station includes: Obtain the communication link channel gain of each communication user and the sensing link channel gain of each sensing target; For each communication user, the first initial allocation coefficient of the communication user is obtained based on the first total allocation coefficient and the first sub-allocation coefficient of the communication user, and the first initial power value of the communication user is obtained based on the total power constraint of the UAV base station and the first initial allocation coefficient of the communication user. The first sub-allocation coefficient is obtained based on the communication link channel gain and task priority weight coefficient of the communication user, and the first total allocation coefficient is obtained by summing the first sub-allocation coefficients of each communication user. For each sensing target, a second initial allocation coefficient for the sensing target is obtained based on the second total allocation coefficient and the second sub-allocation coefficient of the sensing target. A second initial power value for the sensing target is obtained based on the total power constraint of the UAV base station and the second initial allocation coefficient of the sensing target. The second sub-allocation coefficient is obtained based on the sensing link channel gain and task priority weight coefficient of the sensing target, and the second total allocation coefficient is obtained by summing the second sub-allocation coefficients of each sensing target.

11. The power distribution method according to claim 1, wherein, The power allocation method further includes: The training samples are obtained. The feature part of the training samples includes the first feature of the sample UAV base station in the sample sensory integration network, the second feature of the sample communication user, and the third feature of the sample perception target. The first feature of the sample includes the total power constraint of the sample UAV base station. The label part of the training samples includes the first sample power allocation weight of the sample communication user associated with the sample human-machine base station and the second sample power allocation weight of the sample perception target associated with the sample human-machine base station. A graph neural network model is trained based on training samples.

12. The power distribution method according to claim 11, wherein, The graph neural network models trained based on training samples include: The graph neural network model is trained based on training samples, with the goal of reducing the value of the target loss function. The target loss function is set based on the mean square error between the first sample power allocation weight of the sample communication user and the first predicted power allocation weight of the sample communication user obtained by the graph neural network model, and the mean square error between the second sample power allocation weight of the sample perception target and the second predicted power allocation weight of the sample perception target obtained by the graph neural network model.

13. The power distribution method according to claim 1, wherein, Graph neural network models are trained based on Markov decision processes, where the reward function in a Markov decision process is based on the state... implement The state at time t is determined by the communication performance reward, perception performance reward, priority reward, backhaul load penalty, and total power penalty. Including the characteristics of each node in the synesthetic network at time t, and the actions at time t. The predicted power allocation weights or target power allocation weights of the nodes associated with the UAV base station are set based on a graph neural network model.

14. A power distribution device for a sensor-integrated network, comprising: The acquisition module is configured to acquire a first feature of the UAV base station, a second feature of the communication user, and a third feature of the perceived target, wherein the first feature includes the total power constraint of the UAV base station; The weight determination module is configured to use a graph neural network model to obtain target embedding features of nodes in a heterogeneous graph based on the first feature, the second feature, and the third feature, and to obtain a first predicted power allocation weight for communication users associated with the UAV base station and a second predicted power allocation weight for sensing targets associated with the UAV base station based on the target embedding features of nodes in the heterogeneous graph. The heterogeneous graph corresponds to the integrated sensing network, and the UAV base station, communication users, and sensing targets constitute nodes in the heterogeneous graph. The edges in the heterogeneous graph include the edges between UAV base station nodes and associated communication user nodes, and the edges between UAV base station nodes and associated sensing target nodes. The power determination module is configured to determine a first target power value of a communication user associated with a UAV base station based on the total power constraint and the first predicted power allocation weight, and to determine a second target power value of a sensing target associated with a UAV base station based on the total power constraint and the second predicted power allocation weight.

15. A power distribution device for a sensor-integrated network, comprising a processor and a memory, wherein the memory stores instructions that, when executed by the processor, implement the steps of the power distribution method according to any one of claims 1 to 13.

16. An aerial platform station, comprising: The power distribution device according to claim 14; or The power distribution device according to claim 15.

17. A non-transitory computer-readable storage medium, wherein, The non-transitory computer-readable storage medium stores instructions that, when executed by a processor, implement the steps of the power allocation method according to any one of claims 1 to 13.

18. A computer program product comprising instructions that, when executed by a processor, implement the steps of the power allocation method according to any one of claims 1 to 13.

Images