Sensing data processing method, communication system, communication device, and storage medium

Abstract

The present application provides a sensing data processing method, a communication system, a communication device, and a storage medium. The method comprises: on the basis of a plurality of pieces of fused sensing data, determining a sensing range overlapping region where a first target jointly sensed by a plurality of sensing nodes is located, wherein the plurality of pieces of fused sensing data come from a plurality of computing nodes, and the fused sensing data are obtained by the computing nodes by performing data fusion on the basis of sensing data of a plurality of subordinate sensing nodes (S210); then performing region allocation on the basis of the sensing range overlapping region to obtain a region allocation result (S220); and broadcasting the region allocation result to the plurality of computing nodes (S230).

Description

Sensing data processing methods, communication systems, devices and storage media

[0001] Cross-references to related applications

[0002] This application is based on and claims priority to Chinese Patent Application No. 2024116408103, filed on November 15, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The embodiments of this application relate to, but are not limited to, the field of communication technology, and in particular to a sensing data processing method, communication system, device, and storage medium. Background Technology

[0004] In the current integrated sensing and computing network architecture, sensing nodes and computing nodes are tightly integrated, forming network units with both sensing and computing functions. These network units not only have the ability to autonomously collect sensing data, but also process and transmit the data. However, because the deployment of sensing nodes is often based on considerations of coverage breadth and redundancy, the sensing ranges of different sensing nodes may overlap, especially in complex environments or critical monitoring areas. This overlap in sensing ranges means that the same target may be captured by multiple sensing nodes simultaneously, resulting in a large amount of redundant data. This not only wastes computing resources but may also affect the real-time performance of data processing due to processing delays. Therefore, how to efficiently and accurately process data within overlapping sensing ranges has become a critical problem that urgently needs to be solved. Summary of the Invention

[0005] This application provides a sensing data processing method, a communication system, a device, and a storage medium.

[0006] On one hand, embodiments of this application provide a sensing data processing method, the method comprising: determining, based on multiple fused sensing data, an overlapping area of ​​sensing ranges of a first target jointly sensed by multiple sensing nodes, wherein the multiple fused sensing data come from multiple computing nodes, and the fused sensing data is obtained by the computing nodes through data fusion based on the sensing data of their subordinate multiple sensing nodes; performing region allocation based on the overlapping area of ​​sensing ranges to obtain a region allocation result; and broadcasting the region allocation result to the multiple computing nodes.

[0007] On the other hand, this application embodiment also provides a communication system, including multiple computing nodes and multiple sensing nodes; one of the multiple computing nodes determines the overlapping area of ​​the sensing range of a first target jointly sensed by the multiple sensing nodes based on multiple fused sensing data, performs region allocation based on the overlapping area of ​​the sensing range, obtains a region allocation result, and then broadcasts the region allocation result to the multiple computing nodes; wherein, the multiple fused sensing data comes from the multiple computing nodes, and the fused sensing data is obtained by the computing node through data fusion based on the sensing data of its subordinate multiple sensing nodes.

[0008] On the other hand, embodiments of this application also provide a communication device, including: at least one processor; at least one memory for storing at least one program; at least one program is executed by at least one processor to implement the perception data processing method as described above when executed.

[0009] On the other hand, embodiments of this application also provide a computer-readable storage medium storing computer-executable instructions for performing the sensory data processing method described above.

[0010] On the other hand, embodiments of this application also provide a computer program product, including a computer program or computer instructions, the computer program or computer instructions being stored in a computer-readable storage medium, a processor of a device reading the computer program or computer instructions from the computer-readable storage medium, and the processor executing the computer program or computer instructions to cause the device to perform the sensing data processing method as described above. Attached Figure Description

[0011] Figure 1 is a diagram of the integrated network architecture for inductive computing provided in an embodiment of this application;

[0012] Figure 2 is a flowchart of a sensing data processing method provided in an embodiment of this application;

[0013] Figure 3 is a schematic diagram of the connection architecture between computing nodes and sensing nodes provided in an embodiment of this application;

[0014] Figure 4 is a schematic diagram of the computing node distribution provided in an embodiment of this application;

[0015] Figure 5 is a detailed flowchart of step S210 in Figure 2 provided in an embodiment of this application;

[0016] Figure 6 is a detailed flowchart of step S220 in Figure 2 provided in an embodiment of this application;

[0017] Figure 7 is a schematic diagram of the region allocation results provided in a specific example of this application;

[0018] Figure 8 is a flowchart of the perception data fusion and reporting process provided in an embodiment of this application;

[0019] Figure 9 is a schematic diagram of the coverage of a sensing node provided in a specific example of this application;

[0020] Figure 10 is a schematic diagram of the coverage of a sensing node provided in another specific example of this application;

[0021] Figure 11 is a schematic diagram of the coverage of a sensing node provided in another specific example of this application;

[0022] Figure 12 is a schematic diagram of a communication system architecture provided in an embodiment of this application. Detailed Implementation

[0023] To make the objectives, technical methods, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0024] It should be noted that although the flowchart shows a logical order, in some cases, the steps shown or described may be performed in a different order than that shown in the flowchart. In the description of the specification, claims, and the foregoing drawings, "multiple" means two or more; "greater than," "less than," and "exceeding" are understood to exclude the stated number; "above," "below," and "within" are understood to include the stated number. The use of terms such as "first" and "second" is merely for distinguishing technical features and should not be construed as indicating or implying relative importance, or implicitly specifying the number of indicated technical features or their sequential relationship.

[0025] With the rapid development of wireless communication technology, the integrated sensing, communication, and computing network architecture is gradually becoming an important direction for future network development. As shown in Figure 1, in the integrated sensing, communication, and computing network architecture, the access network side supports communication, sensing, and computing functions, while the core network side mainly consists of sensing network elements and communication network elements. Communication network elements refer to devices or nodes in the communication network used to implement core services such as data transmission, signaling processing, and network management. They are responsible for processing and forwarding large amounts of data traffic, and also need to support various signaling processing and network management functions to ensure the normal operation of the entire communication network. Sensing network elements refer to devices or nodes in the sensing network responsible for various sensing-related functions such as sensing authorization, capability interaction, network element selection, control, and data processing. The core network side supports functions such as communication mobility management, sensing management, and control. This architecture aims to achieve more efficient and intelligent network services by integrating communication, sensing, and computing capabilities. It should be noted that the functional entities of communication network elements and sensing network elements can also be deployed to the access network; this embodiment does not specifically limit this.

[0026] However, the design of sensor network architectures and solutions faces multiple challenges, which are particularly prominent in current integrated sensor-computation network architectures. First, some scenarios require high-complexity algorithms to process sensor data, but the processing power of a single node is limited. Therefore, computing nodes must be introduced to achieve collaboration between sensing and computation. This requires the network to not only efficiently collect sensor data but also flexibly allocate computing power among nodes to cope with complex data processing needs. Second, in different sensor application scenarios, depending on the specific requirements of the sensing task, it is necessary not only to carefully select the nodes that perform the sensing task but also to designate computing nodes to process the measurement data collected by the sensing nodes. More importantly, in actual network deployments, the sensing ranges of sensing nodes managed by different computing nodes often overlap, leading to complex challenges in data processing and node collaboration. Specifically, when sensing ranges overlap, different computing nodes may collect discrepancies or redundancies regarding the same sensing target. This phenomenon not only significantly increases the difficulty of data fusion and processing but may also affect the accuracy and real-time performance of the data. Furthermore, when multiple computing nodes collaboratively process this data, the lack of an effective collaboration mechanism may lead to resource waste and reduced efficiency.

[0027] Therefore, how to effectively reduce the complexity of collaborative processing among multiple computing nodes and achieve efficient fusion and processing of sensing data in a network architecture that integrates sensing and computing has become a critical issue that urgently needs to be addressed.

[0028] To effectively fuse sensing data while reducing the complexity of collaborative processing among multiple computing nodes, thereby avoiding redundant sensing within the same area, this application provides a sensing data processing method, communication system, communication device, computer-readable storage medium, and computer program product. First, based on multiple fused sensing data, the overlapping sensing range of a first target jointly sensed by multiple sensing nodes is determined. The fused sensing data originates from multiple computing nodes and is obtained by each computing node through data fusion of sensing data from its subordinate sensing nodes. Then, a region allocation is performed based on the overlapping sensing range, yielding a region allocation result. Subsequently, the region allocation result is broadcast to multiple computing nodes. This application, by determining the overlapping sensing range of a target jointly sensed by multiple sensing nodes, can identify the source of data redundancy, thereby avoiding redundant processing of the same target. Furthermore, by rationally allocating the overlapping sensing range, not only is the waste of computing resources effectively reduced, but the resource utilization efficiency of the entire network system is also improved.

[0029] Based on the above analysis, the embodiments of this application will be further described below with reference to the accompanying drawings.

[0030] Referring to Figure 2, which is a flowchart of a perception data processing method provided in an embodiment of this application, the method includes, but is not limited to, steps S210 to S230.

[0031] Step S210: Based on multiple fused sensing data, determine the overlapping area of ​​the sensing range of the first target jointly sensed by multiple sensing nodes. The multiple fused sensing data come from multiple computing nodes. The fused sensing data is obtained by the computing nodes by fusing the sensing data of their subordinate multiple sensing nodes.

[0032] Step S220: Allocate regions based on the overlapping areas of the sensing range to obtain the region allocation results;

[0033] Step S230: Broadcast the region allocation results to multiple computing nodes.

[0034] For example, in an integrated sensing and computing network, sensing nodes combine communication and sensing functions, widely integrating various devices, modules, and components. They can not only efficiently transmit and receive data, but also process communication signals, capture sensing signals, and effectively convey control commands. Typically, these nodes are deployed on the access network side, including base stations, relay stations, and terminal devices with communication and sensing capabilities. Within a specific sensing range, sensing nodes can accurately monitor and identify various target objects or phenomena. Sensing data is the collection of information acquired by sensing nodes after performing sensing activities on targets (such as the primary target) within their sensing range. Specific target information includes, but is not limited to, target sequence and identification, latitude and longitude information, distance, speed, acceleration, azimuth angle, elevation angle, target type, and target signal strength. On the other hand, computing nodes are the core components in the network responsible for processing communication or sensing information and providing computing resources. These nodes may appear as servers, dedicated computing devices such as computing boards, or even base stations, relay stations, or terminal devices that integrate computing capabilities and resources. In the deployment of an integrated sensing and computing network architecture, sensing nodes and computing nodes are typically configured at the access network layer. It is worth noting that if a sensing node possesses a certain computing capability, it can also act as a computing node. Of course, the implementation of computing nodes is not limited to integration with sensing nodes; they can also exist by independently deploying computing resources. This application does not impose specific restrictions on such deployment methods.

[0035] Understandably, in an integrated sensing and computing network, the sensing elements on the core network side can determine the required sensing area based on the sensing needs of the application layer (such as various applications), and further determine the sensing nodes that should be activated based on the actual coverage of that sensing area. For example, if a specific functional platform is needed to achieve low-altitude monitoring of a certain geographical area, this "low-altitude monitoring" task constitutes a clear sensing requirement of the application layer. When the sensing elements of the core network receive this sensing requirement instruction from the functional platform, they will meticulously plan the specific area to be monitored based on the instruction, and decide accordingly which sensing nodes should be deployed to ensure that these nodes can carry out effective sensing operations in the designated area.

[0036] For example, sensing nodes can establish connections with computing nodes, and computing nodes can collaboratively network and fuse sensing data collected by one or more associated sensing nodes, and report the fused data to the sensing network elements. Network fusion refers to the overall process of forming a network coverage area without duplicate target sensing results through target association and information fusion when constructing a sensor network. Target association refers to matching and associating targets captured by different sensing nodes. Here, "target" broadly refers to any perceptible entity, such as people or objects captured by cameras, objects scanned by radar, or any entity identified by other devices. Taking two sensing nodes (e.g., base station 1 and base station 2) as an example, if each of them senses a target, base station 1 and base station 2 will compare and associate their respective sensed targets to determine whether they sense the same target. Once confirmed, the association of the target is completed. Information fusion refers to the process of fusing and unifying the sensing information of different sensing nodes for the same target according to a unified standard, based on target association. For example, if the coordinates of a target relative to base station 1 are (x1, y1, z1) and the coordinates relative to base station 2 are (x2, y2, z2), after information fusion, the unified coordinates of the target relative to base station 1 and base station 2 will become (x3, y3, z3).

[0037] It should be noted that the connection architecture between computing nodes and sensing nodes exhibits high flexibility and versatility. They support multiple connection modes, including but not limited to the architecture of multiple sensing nodes and one computing node shown in Figure 3, as well as a one-to-one (one sensing node to one computing node) connection architecture. Furthermore, they can adapt to more complex scenarios, such as multiple sensing nodes interconnected with multiple computing nodes, where these computing nodes can be further aggregated into a unified computing node (similar to the concept of a master node). Regardless of the networking architecture, the system can adapt, fully demonstrating its powerful flexibility and wide applicability.

[0038] For example, in a sensor-computing integrated network, specific computing nodes can be selected as central nodes, while other computing nodes surrounding these central nodes are designated as non-central nodes. This layout not only optimizes the network architecture but also ensures efficient data flow and processing. For instance, the specific process for selecting a computing node as a central node is as follows: First, starting from the outermost edge of the entire network, nodes surrounded by other computing nodes are selected and designated as central nodes for a specific area. Simultaneously, other nodes surrounding these central nodes are marked as non-central nodes. To ensure network connectivity and smooth data flow, robust interconnection links are established between central and non-central nodes. This marking and connection process continues until all computing nodes in the network are clearly categorized and connected. Furthermore, if a computing node is connected to a sensing node that is located in a remote area (i.e., geographically dispersed), that computing node will also be specifically identified as a non-central node corresponding to that remote area. In the example in Figure 4, compute node 1, compute node 2, compute node 3, and compute node 4 are selected and marked as central nodes, while the remaining nodes are considered as non-central nodes.

[0039] For example, fused sensing data can be obtained by a computing node through data fusion of sensing data from multiple subordinate sensing nodes. Specifically, the computing node is connected to multiple sensing nodes, receiving and processing sensing data from these nodes. One of the core tasks in processing this data is data fusion to accurately identify whether these sensing nodes have repeatedly perceived the same target, i.e., to determine whether the data originates from the observation of the same target. Specifically, when multiple fused sensing data include first fused sensing data, and multiple sensing nodes include multiple first sensing nodes, in order to obtain the first fused sensing data, after receiving the first sensing data sent by multiple first sensing nodes, the computing node can first perform correlation judgment processing on the sensing targets corresponding to different first sensing data to obtain a first judgment result. Subsequently, based on the first judgment result, it determines whether these first sensing data need to be fused to obtain the first fused sensing data. In this step, if the first judgment result clearly indicates that different first sensing data are actually observation records of the same sensing target, these scattered data can be merged into a unified first fused sensing data. This fusion process not only reduces data redundancy but also ensures data accuracy and consistency.

[0040] For example, correlation determination is a crucial step in comparing and analyzing the relationships between perceived targets corresponding to multiple sets of first-sensory data. This step aims to determine whether multiple perceived targets point to the same entity. Correlation determination can be performed using methods such as direct comparison and feature matching. Specifically, direct comparison involves directly comparing multiple sets of first-sensory data. This method involves: firstly, preprocessing the multiple sets of first-sensory data to ensure consistency in format, units, and precision for direct comparison; then, comparing the values ​​of each set of first-sensory data and calculating the degree of difference; and finally, determining the correlation of these data based on a preset threshold and the degree of difference. For example, if the degree of difference between two sets of first-sensory data is less than a certain threshold, then the perceived targets corresponding to these two sets of first-sensory data can be considered correlated. On the other hand, feature matching involves extracting features from multiple sets of first-sensory data and matching them based on these features. Features can be temporal characteristics of the data (such as periodicity). By comparing the features of multiple sets of first-sensory data, it can be determined whether they point to the same perceived target. The implementation steps of this method include: firstly, feature extraction is performed on the first-sensory data to extract representative features; then, appropriate matching algorithms (such as distance metrics, similarity calculations, classification algorithms, etc.) are used to compare the features of these first-sensory data; finally, the correlation between the data is determined based on the matching results. For example, if the feature matching degree of two first-sensory data is high, the sensing targets corresponding to these two first-sensory data can be considered to be correlated. It should be noted that in practical applications, the choice of which correlation determination method to use depends on the nature, format, and accuracy of the data, as well as the required judgment accuracy and efficiency. A comprehensive judgment using multiple methods is necessary, and this application does not impose any restrictions on this.

[0041] For example, the plurality of computing nodes includes a first computing node and at least one second computing node, wherein the first computing node is marked as a central node and the second computing node is marked as a non-central node. Each second computing node can establish connections with multiple sensing nodes, and is able to receive sensing data transmitted by these subordinate sensing nodes, and then perform fusion processing on this data to generate corresponding fused sensing data. After completing the data fusion, the second computing node can send this fused sensing data to the first computing node.

[0042] For example, for each computing node marked as a central node (hereinafter referred to as the central computing node, such as the first computing node), at a determined periodic time point (e.g., a period of 60 seconds), all computing nodes marked as non-central nodes (hereinafter referred to as non-central computing nodes, such as the second computing node) can send the latest network-fused sensing data (i.e., fused sensing data) to the central computing node they are connected to. The sensing data includes specific target information, including but not limited to target sequence number and identifier, latitude and longitude information, distance, speed, acceleration, azimuth angle, elevation angle, target type, target signal strength, etc.

[0043] For example, a central computing node can act as the central node of a certain region, establishing an interconnected link system with its surrounding non-central computing nodes to transmit and interact with data. In this system, the central computing node can receive and fuse the first sensing data from multiple subordinate first sensing nodes, and can also receive fused sensing data obtained by non-central computing nodes fusing the second sensing data from multiple second sensing nodes. Furthermore, based on this fused sensing data, it can determine the overlapping area of ​​the sensing range of the target jointly sensed by the first and second sensing nodes. It is understood that since this fused sensing data includes physical parameters such as the target's geographical location (latitude and longitude), distance measurement, velocity value, acceleration value, azimuth angle, and pitch angle, the central computing node can determine the overlapping area of ​​the sensing range of the first target jointly sensed by multiple first and second sensing nodes. As shown in Figure 5, the specific process of step S210 includes, but is not limited to, steps S510 and S520.

[0044] Step S510: Based on the first fused perception data and at least one second fused perception data, perform target fusion on the perception target corresponding to the first fused perception data and the perception target corresponding to at least one second fused perception data to obtain a first target jointly perceived by multiple first perception nodes and multiple second perception nodes;

[0045] Step S520: Determine the overlapping area of ​​the perception range of the first target.

[0046] For example, in step S510, a correlation judgment process can be performed on the sensing targets corresponding to the first and second fused sensing data based on the first and second fused sensing data to obtain a second judgment result. This step aims to determine which sensing targets have potential correlations or similarities, and the second judgment result can indicate which pairs of sensing targets can be considered as different perceptual manifestations of the same target. Subsequently, based on the second judgment result, it is determined whether target fusion is needed for the sensing targets corresponding to the first and second fused sensing data. If the judgment result shows that a certain sensing target in the first and second fused sensing data does indeed belong to the same target, then these two sensing targets will be merged to form a more comprehensive fused target. At the same time, this fused target can also be associated with the corresponding first and second sensing nodes to form target association pairs. These association pairs refer to the pairing relationships formed after different sensing nodes associate the same target.

[0047] It should be noted that a central computing node (e.g., the first computing node), acting as the central node of a certain region, can establish interconnected links with multiple non-central computing nodes (e.g., the second and third computing nodes) to transmit and interact with data. Specifically, after obtaining the first fused sensing data, the second fused sensing data sent by the second computing node, and the third fused sensing data sent by the third computing node, the first computing node can further perform target fusion based on the first, second, and third fused sensing data to obtain a fused target. In this process, a correlation judgment can be performed on the perception targets corresponding to the first, second, and third fused sensing data to obtain a judgment result. This judgment result can indicate which perception targets exhibit different perception behaviors that can be considered the same target. Subsequently, based on this judgment result, the first computing node will perform target fusion on those perception targets determined to be the same target. If the judgment result shows that a certain sensing target in the first fused sensing data, a certain sensing target in the second fused sensing data, and a certain sensing target in the third fused sensing data do indeed belong to the same target, then these three sensing targets will be merged to form a more comprehensive fused target. Simultaneously, this fused target can be associated with the corresponding first, second, and third sensing nodes to form target association pairs. It is worth noting that non-central nodes connected to the central node can report the latest fused sensing data (i.e., the latest fused sensing data) that the computing node has completed network fusion within that period according to an agreed-upon cycle. The central node can maintain a storage space to store all data obtained through the aforementioned periodic association and fusion steps. Once the storage space reaches its capacity limit, the data record with the longest storage time can be automatically removed to ensure efficient utilization of storage resources.

[0048] For example, the central computing node compares and integrates the corresponding sensing targets in the two sets of data, and finally confirms the target jointly observed by the first and second sensing nodes. After successfully fusing the two sets of sensing data and identifying the first jointly perceived target, by further analyzing the target location information (such as latitude, longitude, distance, azimuth, etc.) provided by the first and second sensing data, the overlapping area formed by the intersection of the sensing ranges of the two sensing nodes can be determined where the first target is located.

[0049] For example, in determining the overlapping area of ​​the sensing range of the first target, the overlapping area between the sensing ranges of multiple first sensing nodes and multiple second sensing nodes can be determined first. It is worth noting that after step S510, the first target has been precisely located within this overlapping area. Subsequently, based on the specific location information of the first target, the actual sensing range of the first target, i.e., the overlapping area of ​​sensing ranges, can be further determined within the identified overlapping area. In other words, this process first relies on defining the sensing ranges of the first and second sensing nodes and determining the overlapping portion between these two sensing ranges, with the first target located within this overlapping area. Then, using the specific location information of the first target, this overlapping area can be further refined to determine the overlapping area of ​​the sensing range of the fused target.

[0050] For example, these fused sensing targets can be stored in computing nodes. For the dataset consisting of all successfully fused sensing targets within the computing node's storage space, a polyhedron composed of overlapping areas can be constructed using the Alpha Shape algorithm or the Beta Shape algorithm based on the latitude, longitude, and height information of the targets. This polyhedron is denoted as the fusion overlapping area (i.e., the sensing range overlapping region). It is worth noting that during the polyhedron construction process, to obtain a suitable polyhedron, when constructing the sensing range overlapping region using the Alpha Shape algorithm, the alpha value can be adjusted according to the density of the target points, for example, set to a value close to the average distance between adjacent target points. For instance, in Figure 7, the polygonal region A composed of multiple black straight lines can be considered a cross-section of the polyhedron, and the black dots represent all successfully fused sensing targets.

[0051] For example, after determining the overlapping area of ​​the sensing range where the fusion target is located, further region allocation can be performed based on the overlapping area of ​​the sensing range to obtain the region allocation result. As shown in Figure 6, the specific process of step S220 may include, but is not limited to, steps S610 to S630.

[0052] Step S610: Determine the number of overlapping regions to be divided;

[0053] Step S620: Based on the number of overlapping regions, the overlapping regions of the sensing range are segmented to obtain the overlapping region segmentation results;

[0054] Step S630: Based on the overlapping region segmentation results, the overlapping regions of the sensing range are allocated to obtain the region allocation results.

[0055] For example, during the information exchange between the central computing node and the non-central computing nodes, the central computing node can also obtain a local first computing resource reserve identifier and receive a second computing resource reserve identifier sent by the non-central computing node. The computing resource reserve identifier indicates whether the corresponding computing node has sufficient computing resources to support network fusion computing in the overlapping area. For example, a true computing resource reserve identifier indicates that the corresponding computing node has sufficient computing resources. A valid computing resource reserve identifier (i.e., a true computing resource reserve identifier) ​​indicates that the corresponding computing node has sufficient computing resources, can receive sensing data from sensing nodes, and perform data fusion on all sensing data. It should be noted that for non-central computing nodes connected to the central computing node, they can report the latest sensing data and computing resource reserve identifier that have been successfully networked within that period to the central computing node according to an agreed-upon period.

[0056] For example, in determining the number of overlapping region segments, the number of valid computing resource surplus identifiers can be determined based on the meanings of the first and second computing resource surplus identifiers; then, the number of overlapping region segments can be determined based on the number of valid computing resource surplus identifiers. For instance, when the first computing resource surplus indicates that its corresponding first computing node has computing resource surplus, and the second computing resource surplus identifier indicates that its corresponding second computing node has computing resource surplus, the number of valid computing resource surplus identifiers is 2. In this case, the number of overlapping region segments can be set to 2 accordingly. That is, there is a direct correspondence between the number of overlapping region segments and the number of valid computing resource surplus identifiers.

[0057] For example, in the process of segmenting overlapping sensing ranges, the computing nodes to be allocated can be determined first based on the effective computing resource reserve identifier. Then, combining the determined number of overlapping area segments and the sensing range of the sensing nodes connected to these nodes to be allocated, the overlapping areas are precisely divided, resulting in at least one overlapping block. These overlapping blocks are all located within the sensing coverage of the sensing nodes connected to their corresponding nodes to be allocated. Next, based on the specific sensing range of each overlapping block, they are rationally allocated to the corresponding computing nodes to be allocated, thus forming the area allocation result. It is worth noting that the effective computing resource reserve identifier ensures that the selected computing nodes have sufficient computing resources to support the network fusion computing in the overlapping areas. It should be explained that in this process, after determining the number of overlapping area segments, further utilizing the sensing range of the sensing nodes connected to these nodes with computing resource reserve to perform a more detailed area division of the overlapping areas where the fused sensing targets are located ensures that each overlapping block can be accurately allocated to the corresponding computing nodes to be allocated, thereby achieving efficient resource utilization and area management.

[0058] For example, when a first computing node marked as the center node is connected to other computing nodes marked as non-center nodes (assuming there are N such nodes), the first computing node uses the Alpha Shape algorithm or the Beta Shape algorithm to construct a polyhedron composed of overlapping areas, i.e., the overlapping region of the perception range. When segmenting the overlapping region of the perception range, it can be divided into M (M≤1+N) overlapping blocks, where M represents the number of computing nodes with remaining computing resources. It is worth noting that in the process of dividing the overlapping region of the perception range into M overlapping blocks, the overlapping region can be segmented in a complete block manner, and target regions composed of computing nodes with the same source node in the target association pair should be grouped together as much as possible. Furthermore, during the initial allocation, multiple overlapping blocks should distribute the entire overlapping region of the perception range as evenly as possible to achieve a balanced resource allocation.

[0059] For example, when allocating overlapping regions of the sensing range based on the overlapping region segmentation results, the allocation principles include at least the following: Allocation by label order: Overlapping blocks are allocated sequentially according to their label order to ensure continuity and orderliness. Priority allocation to nodes with abundant fusion targets and sufficient resources: Overlapping blocks are preferentially allocated to computing nodes with a large number of fusion targets (e.g., the first target) and whose computing resource reserve identifier is true. This principle aims to ensure maximum resource utilization while reducing performance bottlenecks caused by insufficient resources. Avoiding duplicate allocation to the same node (unless necessary): In the initial allocation phase, duplicate allocation of fused overlapping blocks to the same computing node should be avoided as much as possible. However, when the number of nodes with computing resource reserve identifiers is less than a preset threshold M, overlapping blocks can be allocated to the same node. Following these principles ensures that the allocation of overlapping regions is both efficient and reasonable, thereby supporting the stable operation and performance optimization of the entire synesthetic computing network.

[0060] For example, after obtaining the region allocation result, the central computing node can send the result to the non-central computing nodes connected to it. The region allocation result includes, but is not limited to: the overlapping block number, the boundary point information of the overlapping blocks (including specific target information and associated information, which can reconstruct the shape of the overlapping area), and the computing node to which the allocation belongs.

[0061] Referring to Figure 7, which is a schematic diagram of the region allocation result provided in a specific example of this application, the figure represents the overlapping of sensing ranges on a horizontal plane at the same height. In this figure, each computing node is equipped with three sensing nodes, each facing one direction and expanding into a 120-degree fan-shaped sensing area. Computing node 1 is the central node, and computing nodes 2 and 3 are non-central nodes. Black dots represent sensing targets that have been successfully associated and fused. It should be noted that some areas may be where two targets are associated as the same target, and areas covered by the three computing nodes may be where three targets are associated as the same target. Taking the allocation result in Figure 7 as an example, assuming that the computing resource reserve indicators of communication computing nodes 1, 2, and 3 are all true, then overlapping block 1 (illustrated as overlapping area 1), overlapping block 2 (illustrated as overlapping area 2), and overlapping block 3 (illustrated as overlapping area 3) can be allocated to computing nodes 1, 2, and 3 respectively. Subsequently, computing node 1 can send the allocation results to computing nodes 2 and 3 to ensure information synchronization and coordination.

[0062] For example, after obtaining the region allocation result and broadcasting it to each non-central computing node connected to it, the central computing node can further perform the networking process within each node, and determine whether the perceived target belongs to the overlapping area of ​​the sensing range and which overlapping block it belongs to after the networking is completed. Targets belonging to a specific overlapping block are then sent to the corresponding computing nodes according to the computing nodes allocated to that overlapping block. Furthermore, after completing the above process, the computing node can also simultaneously receive fused sensing data sent by other computing nodes, and further perform inter-node networking fusion of the collected sensing data with the fused sensing data within its own node, and finally report the fused sensing result to the sensing network element.

[0063] For example, after obtaining the region allocation result, the central computing node will not only broadcast the result to every non-central computing node connected to it, but also send the result to itself for retention or further processing. The further processing may be, for example, verifying the received region allocation result to ensure the accuracy and integrity of the data.

[0064] For example, as shown in Figure 8, after broadcasting the region allocation result to multiple computing nodes, the specific process of the first computing node (central computing node) performing fusion and reporting processing of the sensing data based on the region allocation result may include, but is not limited to, steps S810 to S830.

[0065] Step S810: Perform data fusion on the third sensing data from multiple first sensing nodes to obtain third fused sensing data;

[0066] Step S820: Receive second target data from the second computing node, wherein the second target data is the perception data of the perception target that is within the perception range of the first perception node, and the second target data is determined by the second computing node in the fourth fused perception data, which is obtained by the second computing node after data fusion based on the fourth perception data from multiple second perception nodes;

[0067] Step S830: Perform data fusion and data reporting based on the second target data and the third fused perception data.

[0068] For example, a first computing node receives third-level sensing data from multiple first-level sensing nodes within its responsible area. This sensing data is obtained by each first-level sensing node after sensing targets within its own sensing range. This sensing data contains key information such as the type, location, and status of the sensing targets. Further, the first computing node performs fusion processing on all received third-level sensing data to form third-level fused sensing data. In the third-level fused sensing data, the first computing node identifies first-level target data belonging to the sensing targets within the sensing range of the second-level sensing nodes and sends this first-level target data to the second computing node so that the second computing node can combine it with its own fourth-level fused sensing data for further data fusion and reporting. Subsequently, the first computing node receives second-level target data from the second computing node and performs final data fusion processing based on the received second-level target data and its own fused third-level fused sensing data. Finally, the first computing node reports the fused data to the sensing network element for further analysis and processing.

[0069] For example, in the third fused sensing data, the first computing node can determine the first target data belonging to the sensing target within the sensing range of the second sensing node, and send the first target data to the second computing node, so that the second computing node can perform data fusion and data reporting based on the first target data and the fourth fused sensing data.

[0070] For example, when a computing node (such as the first computing node) faces a situation where its computing power is limited, in order to ensure the continuity of data processing, the overlapping blocks allocated to that node can be flexibly reallocated. Specifically, when the first computing node detects that its local computing resources have reached saturation and it can no longer handle the current workload, it will first initiate a first computing assistance request to the second computing node. This request aims to seek additional computing support to alleviate the computing pressure on the current node. Subsequently, after receiving this request, the second computing node will quickly assess its own computing resource status. If it confirms that it has sufficient computing resource reserves to handle additional computing tasks, the second computing node will send a first computing assistance consent message to the first computing node, indicating its willingness to provide assistance. After receiving the consent message from the second computing node, the first computing node will reallocate the target overlapping blocks originally allocated to it based on this information. Since these target overlapping blocks are those regions that overlap with the second computing node, reallocating them to the second computing node is reasonable and effective. Finally, to ensure that all relevant nodes can obtain the latest overlapping block allocation information in a timely manner, the first computing node will broadcast the redistribution results to ensure that the second computing node and other potentially affected nodes are aware of them and can adjust their working status accordingly.

[0071] For example, similarly, when a second computing node faces a situation of limited computing power, in order to ensure the continuity of data processing, the overlapping blocks allocated to that node can be flexibly reallocated. The specific steps are as follows: When local computing resources are insufficient, the second computing node first initiates a second computing assistance request to its central node, i.e., the first computing node. This request aims to seek support from the first computing node to alleviate the current computing pressure. Subsequently, the second computing node waits for and receives second computing assistance consent information from the first computing node. This consent information is sent by the first computing node based on its confirmation that it has sufficient computing resource reserves, according to the request of the second computing node. After obtaining the consent of the first computing node, the second computing node will reallocate the overlapping blocks. Specifically, it will reallocate the target overlapping blocks that were originally allocated to it and overlap with those of the first computing node to the first computing node. Finally, the second computing node promptly sends the reallocation results to the first computing node.

[0072] It should be noted that,

[0073] The following describes the perception data processing method provided in this application using three specific application scenarios.

[0074] Application Scenario 1:

[0075] The sensing network element determines the sensing area according to the sensing requirements of the application layer, and determines the sensing nodes to be used and activated according to the coverage of the sensing area. It is assumed that the coverage of the activated sensing nodes is represented by the sectors in Figure 9. Among them, computing node 1 is the central node, computing node 2 and computing node 3 are non-central nodes. Each computing node is equipped with 3 sensing nodes. Each sensing node faces one direction and expands a 120-degree sector sensing area. For computing node 1, which is the central node, the following overlapping area determination and allocation scheme can be executed periodically: (1) For computing nodes 2 and 3 connected to the central node, the latest sensing data and computing resource balance identifier of the computing node that has been networked and integrated within the specified period (which can be achieved through parameter configuration, etc., assuming it is 60 seconds) are reported to the central node. The sensing data contains specific target information, including but not limited to target number and identifier, latitude and longitude information, distance, speed, acceleration, azimuth angle, pitch angle, target type and target signal strength, etc. It is assumed that the computing resource balance identifiers of computing node 1, computing node 2 and computing node 3 are all true. (2) The central node performs target association on the target set consisting of the latest network fusion data of its own node and the sensing data reported by non-central nodes, and determines whether the sensing targets belonging to different computing nodes can be identified as the same target, obtains all target association pairs, and performs information fusion on the targets within the target association pairs to obtain the fused sensing targets. (3) The central node maintains a storage space to store all the periodically associated and fused data in the above steps. When the storage is full, the stored data with the longest duration is deleted. All fused targets are similar to the black dots in Figure 9, which represents the fusion situation on the horizontal plane at the same height. (4) The point set consisting of the position information of all fused sensing targets in the storage space is used to construct the polygon or polyhedron composed of overlapping areas using the Alpha shape algorithm, which can obtain the fused overlapping area, as shown in Figure 9, which is a polygonal area B composed of multiple black straight lines. (5) The Alpha shape (i.e., the overlapping area of ​​the sensing range) composed of the fused overlapping area is divided into 3 parts, namely overlapping area 1, overlapping area 2 and overlapping area 3 in the figure. Since the computing resource reserve identifiers of computing node 1, computing node 2 and computing node 3 are all true, according to the regional allocation principle, overlapping block 1, overlapping region 2 and overlapping region 3 can be allocated to computing node 1, computing node 2 and computing node 3 respectively. (6) Computing node 1 broadcasts the regional allocation results obtained in the above steps to all non-central nodes connected to it. The information content includes, but is not limited to: the number of the merged overlapping block, the boundary point information of the merged overlapping block (including specific target information and related information, which can reconstruct the shape of the overlapping area) and the computing node to which it belongs.It is worth noting that after each computing node completes the construction of its corresponding fusion overlap area and receives the area allocation results from other nodes, it can further proceed with the network formation process according to the original procedure. It then judges the sensed targets after the network formation to determine whether they belong to the fusion overlap area and which fusion overlap block they belong to. Targets belonging to a specific fusion overlap block are then sent to the corresponding computing nodes according to the computing nodes allocated to that block. Furthermore, while completing the allocation of its corresponding fusion overlap area, each computing node can also simultaneously receive sensed data sent by other computing nodes. After collection, it performs network fusion and reports the sensed results to the sensed network elements. Taking computing node 2 as an example, after completing internal network fusion, it can receive sensed targets belonging to fusion overlap block 2 from computing nodes 1 and 3, perform network fusion, and then report the sensed results.

[0076] Furthermore, when computing node 1's computing power is limited, the merged overlapping blocks allocated to it can be reallocated. The specific process is as follows: First, computing node 1 proactively sends a computing assistance request to other nodes with potential assistance capabilities, such as computing node 2 and computing node 3. Subsequently, after receiving the request, computing node 3 conducts an internal resource assessment, confirms that it has available idle computing resources, and responds to the request accordingly, indicating that it can provide assistance. Based on computing node 3's response, computing node 1 further selects computing node 3 as the assisting node. Next, computing node 1 reallocates those merged overlapping blocks 1 that share common areas with computing node 3 to computing node 3. Finally, computing node 1 broadcasts this reallocation information to all other relevant computing nodes, ensuring that all nodes are aware of the latest task allocation status in a timely manner.

[0077] Application Scenario 2:

[0078] The sensing network element determines the sensing area based on the sensing requirements of the application layer, and determines the sensing nodes to be used and activated based on the coverage of the sensing area. Assume the coverage of the activated sensing nodes is represented by the sectors in Figure 10. Here, computing node 1 is the central node, and computing nodes 2, 3, and 4 are non-central nodes. Each computing node carries three sensing nodes, each facing one direction and expanding into a 120-degree sector sensing area. For computing node 1, which is the central node, the following overlapping area determination and allocation scheme is periodically executed: (1) For computing nodes 2, 3 and 4 connected to the central node, the latest network fusion completed perception data and computing resource balance identifier of the computing node are reported to the central node according to the agreed period (which can be achieved through parameter configuration, etc.); the perception data contains specific target information, including but not limited to target sequence number and identifier, latitude and longitude information, distance, speed, acceleration, azimuth angle, pitch angle, target type, target signal strength, etc.; it is assumed that the computing resource balance identifiers of computing nodes 1, 2 and 3 are all true, and the computing resource balance identifier of computing node 4 is false. (2) The central node performs target association on the target set composed of the latest network fusion completed data of its own node and the perception data reported by non-central nodes, determines whether the perception targets belonging to different computing nodes can be determined as the same target, obtains all target association pairs, and performs information fusion on the targets within the target association pairs to obtain the fused perception targets. (3) The central node maintains a storage space to store the periodically associated and fused data from all the above steps. When the storage is full, the stored data with the longest duration is deleted. All fused targets are shown as black dots in Figure 10, which represents the fusion situation on the horizontal plane at the same height. (4) The Beta shape algorithm is used to construct polygons or polyhedra composed of overlapping areas from the point set of position information of all fused sensing targets in the storage space. The fused overlapping area is obtained, as shown in Figure 10, which is a polygonal region C composed of multiple black straight lines. (5) The Alpha shape composed of the fused overlapping area is divided into 4 overlapping blocks, namely overlapping region 1, overlapping region 2, overlapping region 3 and overlapping region 4. As shown by the black dividing lines P1, P2 and P3 in the figure. Since the computing resource reserve of computing node 4 is marked as false, meaning that computing node 4 has sufficient computing resources to support the network fusion computing of the overlapping areas, according to the regional allocation principle, overlapping areas 1, 2, 3 and 4 can be allocated to computing nodes 1, 2, 3 and 1 respectively.(6) Computing node 1 broadcasts the region allocation results obtained in the above steps to all non-central nodes connected to it. The information includes, but is not limited to: the number of the fused overlapping block, the boundary point information of the fused overlapping block (including specific target information and associated information, which can reconstruct the shape of the overlapping area), and the computing node to which it belongs. Similarly, after each computing node completes the construction of its corresponding fused overlapping area and receives the region allocation results sent by other nodes, it can proceed with the networking process according to the original procedure, and judge the sensing targets after the networking is completed to determine whether they belong to the fused overlapping area and which fused overlapping block they belong to. The targets belonging to a specific fused overlapping block are then sent to the corresponding computing nodes according to the computing nodes allocated to their respective fused overlapping blocks. When each computing node completes the allocation of its corresponding fused overlapping area, it can also simultaneously collect sensing data sent by other computing nodes, and after the collection is completed, it performs network fusion and reports the sensing results to the sensing network element. Taking computing node 2 as an example, after completing the internal network fusion, it can receive sensing targets from computing nodes 1 and 4 that belong to the fusion overlapping block 2, and report the sensing results after network fusion.

[0079] Furthermore, when computing node 1's computing power is limited, the merged overlapping blocks allocated to it can be reallocated. The specific process is as follows: First, computing node 1 will proactively send computing assistance requests to surrounding computing nodes (such as nodes 2, 3, and 4). Next, upon receiving the request, computing node 3 will assess its own computing resource status. If it confirms the existence of idle computing resources, node 3 will immediately respond to the request, indicating its willingness to provide computing support to node 1. Based on node 3's response, computing node 1 will further make a decision and select node 3 as the assisting node. Subsequently, node 1 will reallocate those merged overlapping blocks 1 that share overlapping areas with node 3 to node 3 to ensure that the task can be effectively distributed and resources can be used rationally. Finally, computing node 1 will broadcast this reallocation information to all other relevant computing nodes.

[0080] Application Scenario 3:

[0081] The sensing network element determines the sensing area according to the sensing requirements of the application layer, and determines the sensing nodes to be used and activated according to the coverage of the sensing area. It is assumed that the coverage of the activated sensing nodes is represented by the sectors in Figure 11. Among them, computing node 1 is the central node, computing node 2 is the non-central node, each computing node is equipped with 12 sensing nodes, each sensing node faces one direction and expands a 120-degree sector sensing area. For computing node 1, which is the central node, the following overlapping area determination and allocation scheme is periodically executed: (1) For computing node 2 connected to the central node, according to the agreed period (which can be achieved through parameter configuration, etc.), the latest sensing data and computing resource balance identifier of the computing node that has been networked and integrated within the period is reported to the central node. The sensing data contains specific target information, including but not limited to target number and identifier, latitude and longitude information, distance, speed, acceleration, azimuth angle, pitch angle, target type, target signal strength, etc. It is assumed that the computing resource balance identifiers of computing node 1 and computing node 2 are both true. (2) The central node performs target association on the target set consisting of all targets in the latest network fusion data of its own node and the sensing data reported by non-central nodes, and determines whether sensing targets belonging to different computing nodes can be identified as the same target, obtains all target association pairs, and performs information fusion on the targets within the target association pairs to obtain the fused sensing targets. (3) The central node maintains a storage space to store all the periodically associated and fused data in the above steps. When the storage is full, the stored data with the longest duration is deleted; all fused targets are similar to the black dots in Figure 11, which represents the fusion situation on the horizontal plane at the same height. (4) The point set consisting of the position information of all fused sensing targets in the storage space is used to construct a polygon or polyhedron composed of overlapping areas using the Alpha shape algorithm, which can obtain the fused overlapping area, as shown in Figure 11, which is a polygonal area D composed of multiple black straight lines. (5) The Alpha shape composed of the fused overlapping area is divided into two overlapping blocks, namely overlapping area 1 and overlapping area 2 in the figure. As shown by the black dividing line L in the figure. Since the computing resource reserve flags of computing node 1 and computing node 2 are both true, according to the regional allocation principle, overlapping region 1 and overlapping region 2 can be allocated to computing node 1 and computing node 2 respectively. (6) Computing node 1 broadcasts the regional allocation results obtained in the above steps to all non-central nodes connected to it. The information content includes, but is not limited to: the number of the merged overlapping block, the boundary point information of the merged overlapping block (including specific target information and related information, which can reconstruct the shape of the overlapping area) and the computing node to which it belongs.Similarly, after each computing node completes the construction of its corresponding fusion overlap area and receives the area allocation results from other nodes, it can proceed with the network formation process according to the original procedure. It then judges the perceived targets after network formation to determine whether they belong to the fusion overlap area and which fusion overlap block they belong to. Targets belonging to a specific fusion overlap block are then sent to the corresponding computing nodes according to the computing nodes allocated to that block. While completing the allocation of its corresponding fusion overlap area, each computing node can also simultaneously receive sensing data sent by other computing nodes. After collection, it performs network fusion and reports the sensing results to the sensing network elements. Taking computing node 2 as an example, after completing internal network fusion, it can receive sensing targets belonging to fusion overlap block 2 from computing node 1, perform network fusion, and then report the sensing results.

[0082] In addition, when the computing power of computing node 1 is limited, the fused overlapping blocks allocated to that node can be reallocated. The specific process is as follows: computing node 1 sends a computing assistance request to computing node 2; if there are idle computing resources in node 2, it responds to the request; node 1 selects node 2 as the assisting node, and allocates the fused overlapping block 1 that has common overlap with node 2 to node 2, and broadcasts this information to other nodes.

[0083] Referring to Figure 12, which is a schematic diagram of a communication system architecture provided in an embodiment of this application, the system consists of multiple computing nodes and multiple sensing nodes. The computing nodes include first computing nodes and second computing nodes, while the sensing nodes include multiple first sensing nodes and multiple second sensing nodes. One of these computing nodes (e.g., a first computing node) can determine the overlapping area of ​​the sensing range of a first target jointly sensed by multiple sensing nodes based on fused sensing data received from the multiple computing nodes. This fused sensing data, including first fused sensing data and second fused sensing data, is obtained by each computing node fusing sensing data provided by its subordinate multiple sensing nodes. Specifically, the first fused sensing data is fused by the first computing node based on first sensing data from its subordinate multiple first sensing nodes; while the second fused sensing data is fused by the second computing node based on second sensing data from its subordinate multiple second sensing nodes and then sent to the first computing node. After determining the overlapping area of ​​the sensing range, the first computing node further performs area allocation based on this area and obtains the area allocation result. Finally, this result is broadcast to all computing nodes in the system to achieve information sharing and collaborative processing.

[0084] In addition, one embodiment of this application discloses a communication device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the sensing data processing method as described in any of the preceding embodiments.

[0085] In addition, one embodiment of this application discloses a computer-readable storage medium storing computer-executable instructions for performing the perception data processing method as described in any of the preceding embodiments.

[0086] Furthermore, one embodiment of this application discloses a computer program product, including a computer program or computer instructions stored in a computer-readable storage medium. The processor of the device reads the computer program or computer instructions from the computer-readable storage medium and executes the computer program or computer instructions, causing the device to perform the sensing data processing method as described in any of the preceding embodiments.

[0087] In this embodiment, firstly, based on multiple fused sensing data, the overlapping area of ​​the sensing range of a first target jointly sensed by multiple sensing nodes is determined. The multiple fused sensing data originate from multiple computing nodes, and are obtained by the computing nodes through data fusion of the sensing data from their subordinate sensing nodes. Then, a region allocation is performed based on the overlapping area of ​​the sensing range to obtain the region allocation result. Subsequently, the region allocation result is broadcast to the multiple computing nodes. This embodiment, by determining the overlapping area of ​​the sensing range of a target jointly sensed by multiple sensing nodes, can identify the source of data redundancy, thereby avoiding repeated processing of the same target. Furthermore, by rationally allocating the overlapping area of ​​the sensing range, not only is the waste of computing resources effectively reduced, but the resource utilization efficiency of the entire network system is also improved.

[0088] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0089] The above provides a detailed description of some implementations of this application. However, this application is not limited to the above-described embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the essence of this application. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

Claims

1. A method for processing sensory data, the method comprising: Based on multiple fused sensing data, the overlapping area of ​​the sensing range of the first target jointly sensed by multiple sensing nodes is determined. The multiple fused sensing data comes from multiple computing nodes, and the fused sensing data is obtained by the computing nodes by fusing the sensing data of the multiple subordinate sensing nodes. Region allocation is performed based on the overlapping areas of the sensing ranges to obtain the region allocation results; The region allocation results are broadcast to the multiple computing nodes.

2. The method of claim 1, wherein, The plurality of fused sensing data includes first fused sensing data, and the plurality of sensing nodes includes a plurality of first sensing nodes; The first fused sensing data is obtained according to the following steps: Based on the first sensing data from multiple first sensing nodes, correlation judgment processing is performed on the sensing targets corresponding to different first sensing data to obtain a first judgment result; Based on the first judgment result, the different first sensing data are fused to obtain the first fused sensing data.

3. The method of claim 2, wherein, The step of fusing different first sensing data according to the first judgment result to obtain first fused sensing data includes: If the first judgment result indicates that the perceived targets corresponding to the different first perception data belong to the same target, the different first perception data are fused into one data to obtain the first fused perception data.

4. The method of claim 1, wherein, The plurality of computing nodes includes a first computing node and at least one second computing node, and the method is applied to the first computing node; the plurality of fused sensing data includes second fused sensing data, which is sent to the first computing node by at least one second computing node.

5. The method of claim 1, wherein, The plurality of fused sensing data includes first fused sensing data and at least one second fused sensing data, and the plurality of sensing nodes includes a plurality of first sensing nodes and a plurality of second sensing nodes; The step of determining the overlapping area of ​​the sensing range of the first target jointly sensed by multiple sensing nodes based on multiple fused sensing data includes: Based on the first fused perception data and at least one second fused perception data, the perception target corresponding to the first fused perception data and the perception target corresponding to at least one second fused perception data are fused together to obtain a first target jointly perceived by multiple first perception nodes and multiple second perception nodes. Determine the overlapping area of ​​the perception range where the first target is located.

6. The method of claim 5, wherein, The step of fusing the sensing target corresponding to the first fused sensing data and the sensing target corresponding to at least one second fused sensing data according to the first fused sensing data and at least one second fused sensing data to obtain a first target jointly perceived by multiple first sensing nodes and multiple second sensing nodes includes: Based on the first fused sensing data and at least one second fused sensing data, a correlation judgment process is performed on the sensing target corresponding to the first fused sensing data and the sensing target corresponding to at least one second fused sensing data to obtain a second judgment result; Based on the second judgment result, target fusion is performed on the sensing target corresponding to the first fused sensing data and at least one sensing target corresponding to the second fused sensing data to obtain a first target jointly perceived by multiple first sensing nodes and multiple second sensing nodes.

7. The method of claim 6, wherein, The step of fusing the sensing target corresponding to the first fused sensing data and at least one sensing target corresponding to the second fused sensing data according to the second judgment result to obtain a first target jointly perceived by multiple first sensing nodes and multiple second sensing nodes includes: If the second judgment result is that the sensing target corresponding to the first fused sensing data and at least one sensing target corresponding to the second fused sensing data belong to the same target, the sensing target corresponding to the first fused sensing data and at least one sensing target corresponding to the second fused sensing data are fused into one target to obtain a first target jointly perceived by multiple first sensing nodes and multiple second sensing nodes.

8. The method of claim 5, wherein, Determining the overlapping area of ​​the perception range where the first target is located includes: Determine the overlapping area between the sensing range of the plurality of first sensing nodes and the sensing range of the plurality of second sensing nodes, wherein the first target is located within the overlapping area; Based on the location information of the first target, the overlapping area of ​​the perception range where the first target is located is determined in the overlapping area.

9. The method of claim 1, wherein, The step of allocating regions based on the overlapping areas of the sensing range to obtain region allocation results includes: Determine the number of segments to divide the overlapping region; Based on the number of overlapping regions, the overlapping regions of the sensing range are divided into regions to obtain the overlapping region segmentation result. Based on the overlapping region segmentation results, the overlapping regions of the sensing range are allocated to obtain the region allocation results.

10. The method of claim 9, wherein, The plurality of computing nodes includes a second computing node; determining the number of overlapping region segments includes: Obtain the local first computing resource reserve identifier and the second computing resource reserve identifier from the second computing node; Based on the meaning of the first computing resource surplus identifier and the meaning of the second computing resource surplus identifier, the number of valid computing resource surplus identifiers is determined, wherein the computing node corresponding to the valid computing resource surplus identifier has computing resource surplus. The number of overlapping regions to be divided is determined based on the number of valid computing resource surplus identifiers.

11. The method of claim 10, wherein, The overlapping region segmentation result includes at least one overlapping block; the step of segmenting the overlapping region of the sensing range according to the number of overlapping region segments to obtain the overlapping region segmentation result includes: The computing nodes to be allocated are determined based on the valid computing resource surplus identifier; Based on the number of overlapping regions and the sensing range of the sensing nodes connected to the computing nodes to be assigned, the overlapping regions of the sensing range are divided to obtain at least one overlapping block. The overlapping area is within the sensing range of the sensing node connected to the computing node to be assigned.

12. The method of claim 11, wherein, The step of allocating the overlapping regions of the sensing range according to the overlapping region segmentation result to obtain the region allocation result includes: Based on the sensing range of each overlapping block, each overlapping block is assigned to the corresponding computing node to be assigned, thus obtaining the region allocation result.

13. The method of claim 1, wherein, The plurality of sensing nodes includes a plurality of first sensing nodes and a plurality of second sensing nodes, and the plurality of computing nodes includes a second computing node; the method further includes: Data fusion is performed on the third sensing data from multiple first sensing nodes to obtain third fused sensing data; The second target data is received from the second computing node. The second target data is the perception data of the perception target that is within the perception range of the first perception node. The second target data is determined by the second computing node in the fourth fused perception data. The fourth fused perception data is obtained by the second computing node after data fusion based on the fourth perception data from multiple second perception nodes. Data fusion and data reporting are performed based on the second target data and the third fused perception data.

14. The method of claim 1, wherein, The plurality of computing nodes includes a first computing node and a second computing node, and the method is applied to the first computing node; the method further includes: If there are no spare local computing resources, send a first computing assistance request to the second computing node; Receive a first computing assistance consent message sent by the second computing node, wherein the first computing assistance consent message is sent by the second computing node in accordance with the first computing assistance request when it has computing resource reserves; Based on the first computing assistance agreement information, the target overlapping block allocated to the first computing node is reassigned to the second computing node, and the reassignment result is broadcast to the second computing node. The target overlapping block is the overlapping block where the first computing node and the second computing node overlap.

15. A communication system comprising multiple computing nodes and multiple sensing nodes; One of the multiple computing nodes determines the overlapping area of ​​the perception range of the first target jointly perceived by the multiple perception nodes based on multiple fused perception data, performs region allocation based on the overlapping area of ​​the perception range, obtains the region allocation result, and then broadcasts the region allocation result to the multiple computing nodes. wherein The multiple fused sensing data come from multiple computing nodes, and the fused sensing data is obtained by the computing nodes by fusing the sensing data of the multiple subordinate sensing nodes.

16. The communication system of claim 15, wherein, The plurality of computing nodes include a first computing node and at least one second computing node, one of the plurality of computing nodes being the first computing node; the plurality of sensing nodes include a plurality of first sensing nodes and a plurality of second sensing nodes; and the plurality of fused sensing data includes first fused sensing data and second fused sensing data. The first fused sensing data is obtained by the first computing node through data fusion based on first sensing data from multiple first sensing nodes; The second fused perception data is sent by the at least one second computing node to the first computing node, and the second fused perception data is obtained by data fusion of second perception data from a plurality of the second perception nodes by the at least one second computing node. 17.A communication device comprising: at least one processor; at least one memory configured to store at least one program; the at least one program is configured to be run by the at least one processor to execute the perception data processing method according to any one of claims 1 to 14.

18. A computer-readable storage medium storing computer-executable instructions, wherein, The computer executable instructions are configured to execute the perception data processing method according to any one of claims 1 to 14.

19. A computer program product comprising computer programs or computer instructions, wherein, The computer program or the computer instructions are stored in a computer readable storage medium, and the processor of the communication device reads the computer program or the computer instructions from the computer readable storage medium, and executes the computer program or the computer instructions, so that the communication device executes the perception data processing method according to any one of claims 1 to 14.

Images