Intelligent deployment and control ball networking method and system of distributed ad hoc network

Abstract

This invention relates to the field of communication network management technology, specifically to a method and system for intelligent deployment of distributed self-organizing networks, comprising the following steps: acquiring signal path trajectory construction time and direction records, filtering path-consistent nodes to generate adjacency sets, extracting signal power features to generate identity tags, and analyzing image displacement aggregation to generate occlusion regions. In this invention, by introducing a parallel construction method of signal propagation duration and direction vectors among multiple nodes, a trajectory combination with multi-dimensional path features is established. Consistency filtering is performed based on path matching degree and signal arrival order. Power change behavior is extracted and frequency domain response analysis is completed, forming a recognizable node signal feature expression. The directionality and density change trends of image displacement trajectories are integrated to mark dynamically occluded regions, achieving a synergistic improvement in the structural transparency, node identification accuracy, and connection redundancy fault tolerance of the communication network in complex scenarios.

Description

Technical Field

[0001] This invention relates to the field of communication network management technology, and in particular to a method and system for intelligent deployment of distributed self-organizing networks. Background Technology

[0002] The field of communication network management technology involves the organization, configuration, monitoring, and maintenance of various resources, structures, and behaviors within a communication network. Core aspects include network topology planning and management, dynamic connection and scheduling of communication nodes, access control of terminal devices, communication path optimization, fault detection and recovery strategies, and coordinated management of bandwidth and latency. During communication network management, it is also necessary to coordinate multiple communication protocols, support compatible access for various types of terminal devices, and achieve complete transmission of communication data and real-time monitoring of network status. This is especially crucial in mobile, multi-terminal, and high-density deployment scenarios, which place higher demands on network autonomy, reliability, and flexibility. Traditional distributed self-organizing network intelligent surveillance sphere networking methods refer to connecting surveillance sphere devices with image acquisition capabilities to the network and relying on inter-device communication capabilities to establish a dynamic network topology to achieve coverage and collaborative work within the monitored area. This technical issue mainly addresses the problem of rapid networking of monitoring equipment in complex or temporary environments. Traditional methods use communication protocols based on fixed network topology configuration, frequency hopping technology, and neighbor discovery mechanisms based on node signal identification for connection establishment and link maintenance. Specifically, it relies on node routing update strategies and physical layer access control processes in network topology management to establish communication paths through hop-by-hop connections between nodes, thereby achieving interconnection between devices.

[0003] In dynamic environments with dense node deployments, frequent occlusion, or rapid topology changes, existing technologies establish links through static topology configuration and beacon neighbor identification mechanisms. They rely on hop-by-hop route updates and physical access processes to maintain connections, making it difficult to reflect the directionality and stability of propagation characteristics of path relationships between nodes in real time. They also lack the ability to structurally model differences in node behavior and the impact of spatial occlusion. Especially when there is occlusion in areas with overlapping views, the connectivity between nodes is prone to failure due to misjudgment of occlusion, leading to frequent local path reconstruction, decreased connection stability, and reduced collaborative coverage efficiency. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides a method for intelligent deployment ball networking in a distributed self-organizing network, comprising the following steps:

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a distributed self-organizing network intelligent deployment ball networking method, comprising the following steps:

[0006] S1: Obtain the sending time of the control ball node and the receiving time of the receiving node during the broadcast signal transmission process, perform matching on the sending and receiving records of the same broadcast signal, construct a propagation duration record sequence based on the difference, and generate a broadcast signal path trajectory group;

[0007] S2: Using the received signal path corresponding to the target node in the broadcast signal path trajectory group, the signal propagation duration and direction path are combined in pairs to construct duration and direction reference items. Path matching degree judgment and signal arrival sequence consistency detection are performed on the reference items corresponding to the target node and the receiving node to generate a stable set of adjacent nodes.

[0008] S3: Using the set of stable adjacent nodes, extract the signal transmission record of the initial period of the control ball node's transmission, obtain the power change time data of the radio frequency output channel in the short period before and after transmission, perform spectrum transformation on the power change time data, and generate a node identity feature tag set;

[0009] S4: Using the node identity feature tag set, call the periodic image collection content to obtain the pixel coordinates of the edge region and the pixel displacement trajectory between consecutive frames, perform direction aggregation processing on the trajectory, and generate a dynamic occlusion feature region set.

[0010] As a further embodiment of the present invention, the broadcast signal path trajectory group includes a time path sequence, a direction path sequence, and a path matching reference model; the stable adjacent node set includes a high matching degree node index, path consistency feature quantity, and target node reception frequency statistics; the node identity feature tag set includes a frequency domain energy distribution pattern, power spectral density curve features, and signal structure behavior encoding; and the dynamic occlusion feature region set includes a direction consistency identifier block, a pixel density change spectrum, and inter-frame occlusion trend feature parameters.

[0011] As a further aspect of the present invention, the specific steps of S1 are as follows:

[0012] S101: Obtain the sending time of the control ball node and the receiving time of the receiving node during the broadcast signal transmission process, calculate the time difference between each group of sending and receiving times, extract the broadcast signal index corresponding to the time difference, match the sending record and the receiving record according to the same signal index, establish a time difference sequence based on the matching result, call each group of data frames in the time difference sequence and aggregate them in sequence to generate a propagation duration record sequence;

[0013] S102: According to the propagation duration recording sequence, call the recorded signal incident angle information, perform polar coordinate transformation on the incident angle data frame, and perform cosine direction derivation operation on the transformation result and the spatial position parameter set of the receiving node to generate a direction vector sequence set;

[0014] S103: Based on the set of direction vector sequences, perform path alignment matching between the propagation duration data and direction vector data corresponding to the same signal index, perform index concatenation operation between the propagation time sequence and direction vector sequence of each broadcast signal, merge the broadcast signal path information, and generate a broadcast signal path trajectory group.

[0015] As a further aspect of the present invention, the specific steps of S2 are as follows:

[0016] S201: Call the received signal path corresponding to the target node in the broadcast signal path trajectory group, and perform combination matching using a dual indexing method based on the propagation duration and corresponding direction vector data of each broadcast signal in the path. Establish key-value correspondence between the propagation duration data and the direction vector data according to the signal index, and generate a duration-direction reference item sequence.

[0017] S202: Based on the duration direction reference item sequence, select the reference items corresponding to the target node and the receiving node respectively, compare the corresponding paths between the two using the path index relocation method, and perform node sequence consistency detection and propagation duration difference calculation operations on the corresponding path group, extract the nodes with path consistency scores higher than the preset path matching threshold, and generate a path consistent node index set.

[0018] S203: Call the broadcast records between the receiving node and the target node in the path-consistent node index set, extract and normalize the signal arrival time difference in the broadcast records, perform secondary index filtering based on the encoding result and the path-consistent node index, aggregate the receiving node numbers that meet the filtering conditions, and generate a stable neighbor node set.

[0019] As a further aspect of the present invention, for each corresponding path in the path-consistent node index set, a bidirectional traversal method is used to sequentially compare the direction vector data included in the reference items corresponding to the target node and the receiving node, and during the comparison process, the change in the angle between the direction vector of any node and the direction vector of the previous node is determined. If the change in the angle is less than a preset direction change threshold, the node is marked as a sequentially consistent node.

[0020] In the propagation duration difference calculation operation, the propagation duration difference corresponding to nodes with consistent order is calculated by weighted accumulation. The weight of the propagation duration difference is determined based on the cosine value of the angle between the corresponding direction vector and the reference direction. When the cosine value is greater than the preset direction consistency threshold, a high weight is assigned.

[0021] The propagation duration difference after weighted accumulation is compared with the standard deviation of the propagation duration of the corresponding path. When the ratio is lower than the set duration consistency judgment threshold, the sequentially consistent node is included in the path consistent node index set.

[0022] As a further aspect of the present invention, the specific steps of S3 are as follows:

[0023] S301: Call the number information of the receiving node in the set of stable adjacent nodes, locate the signal transmission record of the control ball node within a set time interval before and after the receiving time of the corresponding adjacent node, extract the radio frequency output channel number associated with each record, and retrieve the channel power sampling value sequence within a fixed time window before and after transmission, aggregate them according to the sampling time order, and generate a channel power time series.

[0024] S302: Based on the channel power time series, the data sequence is divided into segments of equal length. Fourier transform is performed on each power segment to extract the frequency domain amplitude vector and the corresponding frequency index to form a frequency domain data frame. Frequency index sorting and amplitude normalization operations are performed on the frequency domain data frame. The frequency domain data frames are spliced ​​together in time order to construct a two-dimensional frequency domain spectrum and generate a frequency domain response profile.

[0025] S303: Call the amplitude trajectory of the frequency segment in the frequency domain response contour map, perform same-frequency segment matching rate calculation on the contour map between adjacent nodes, extract the frequency index with a matching rate higher than the preset segment comparison judgment threshold, perform feature aggregation on the behavior waveform pattern corresponding to the frequency index, bind the node number and signal behavior features, and generate a node identity feature label set.

[0026] As a further aspect of the present invention, the specific steps of S4 are as follows:

[0027] S401: Call the identification information associated with the target node in the node identity feature tag set, obtain the corresponding periodic image collection content, locate the contour boundary of the edge region in each frame image, extract the pixel coordinate set of the pixel block in the edge region, perform position difference operation on the pixel block with the same coordinate position between adjacent frames, construct the continuous inter-frame displacement path of the pixel block in time order, and generate a set of pixel displacement trajectories.

[0028] S402: Based on the start and end coordinates of each trajectory in the pixel displacement trajectory set, perform displacement vector angle calculation, quantize and group the angle values, perform numbering and classification processing on the trajectories in the same direction, perform continuous intra-frame consistency judgment on the classified pixel block set, filter pixel block regions with consistent offset directions, and generate a set of continuously offset pixel blocks in the same direction.

[0029] S403: Call the numbering information of the pixel block region in the continuously offset pixel block set, perform frame-by-frame difference operation on the pixel density value of the pixel block region in consecutive frames and record the density difference sequence, calculate the density difference trend change rate in time order, perform index tag aggregation on the pixel block region with continuously increasing trend change rate, and generate a dynamic occlusion feature region set.

[0030] As a further aspect of the present invention, the method further includes step S5:

[0031] S5: Based on the pixel occlusion region position marked by the dynamic occlusion feature region set, combined with the direction vector information of stable adjacent nodes, construct the intersection region between multiple nodes, determine the proportion of the intersection region in the real-time node view and cross-validate it with the redundancy information of the collaborative nodes, perform the attribution classification processing of the view of the main view node, and generate the intelligent deployment ball network connectivity table.

[0032] The intelligent deployment ball network connectivity table includes the main view node index, the collaborative view overlap evaluation value, and the node connectivity state matrix.

[0033] As a further aspect of the present invention, the specific steps of S5 are as follows:

[0034] S501: Based on the pixel coordinate range of the occluded area in the set of dynamic occlusion feature areas, call the direction vector information corresponding to the nodes in the set of stable adjacent nodes, perform a direction vector reverse extension positioning operation on the occluded area, extract the adjacent node number and direction vector combination of the pointing area, and perform cross projection superposition on the direction vector in the image coordinate system to generate a multi-node cross area mapping group.

[0035] S502: Based on the node number corresponding to the region in the multi-node cross region mapping group, retrieve the visible region mask of the real-time frame image of the node, calculate the ratio of the pixel value occupied by the cross region in the node image to the total number of pixels in the image, construct the region proportion parameter of the cross region in the view, count the number of visible cross regions associated with the node, and generate a cross region view proportion and redundancy parameter set.

[0036] S503: Call the cross-region view proportion and redundancy parameter set, perform region classification operation on the main view node, determine whether the node is a main view node according to the proportion threshold, and mark the redundancy status of the node in combination with the redundancy count parameter. Map the visual coverage relationship between the nodes with node numbers and mark the connectivity status to generate the intelligent deployment ball network connectivity table.

[0037] A distributed, self-organizing intelligent deployment ball network system includes:

[0038] The trajectory extraction module obtains the transmission time parameters of the broadcast signal from the control ball node and the reception time parameters of the receiving node, calls the signal propagation start and end time to calculate the propagation time length sequence, combines the signal incident angle information recorded by the direction sensing device of the receiving node to obtain the direction path parameters, and matches the propagation time length sequence and direction path parameters according to the signal identifier to obtain the broadcast signal path trajectory group.

[0039] Based on the broadcast signal path trajectory group, the path filtering module extracts the path data between the target node and the receiving node, obtains the propagation time length parameter and direction path parameter in each pair of paths, performs a combined judgment of the time length difference and the direction angle difference, and obtains a set of stable adjacent nodes.

[0040] The adjacency steady-state module calls the set of stable adjacency nodes, extracts the signal transmit power, receive power and signal frequency band parameters, calculates the signal power change trend under the same frequency band, judges the degree of stability in communication based on the number of times the receiving node is in the continuous communication record and the magnitude of the power fluctuation trend, and generates a set of node identity feature tags.

[0041] The signal tag module uses the node identity feature tag set to extract the power change time record at the beginning of the signal transmission stage, performs frequency domain feature transformation on the time record to form a frequency domain response spectrum, and determines the number of features, intervals and energy change amplitudes in the frequency band of the response spectrum to obtain a set of dynamic occlusion feature regions.

[0042] The occlusion modeling module calls the dynamic occlusion feature region set to obtain the edge pixel positions and displacement trajectories in continuous image frames, counts the aggregation degree of the offset direction of pixels in continuous frames, filters pixel regions with consistent directions, performs the attribution classification processing of the viewpoint of the main viewpoint node, and obtains the intelligent deployment ball network connectivity table.

[0043] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0044] In this invention, by introducing a parallel construction method of signal propagation duration and direction vector among multiple nodes, a trajectory combination with multi-dimensional path characteristics is established. Consistency screening is performed based on path matching degree and signal arrival order as standards. Power change behavior is extracted and frequency domain response analysis is completed to form a recognizable node signal feature expression. The directionality and density change trend of image displacement trajectory are fused, dynamic occlusion areas are marked, and node viewpoint attribution mapping is constructed based on this and combined with directional intersection relationship. Node connectivity results with spatiotemporal cross-verification capability and viewpoint classification mechanism are generated, enhancing the comprehensive response capability to multiple factors such as spatial occlusion, node identity, and path stability during network topology establishment. This achieves a synergistic improvement in the structural transparency, node identification accuracy, and connection redundancy fault tolerance capability of communication networks in complex scenarios. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a schematic diagram of the steps of the present invention;

[0047] Figure 2 This is a detailed schematic diagram of S1 of the present invention;

[0048] Figure 3 This is a detailed schematic diagram of S2 of the present invention;

[0049] Figure 4 This is a detailed schematic diagram of S3 of the present invention;

[0050] Figure 5 This is a detailed schematic diagram of S4 of the present invention;

[0051] Figure 6 This is a detailed schematic diagram of S5 of the present invention;

[0052] Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0053] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0054] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0055] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0056] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0057] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0058] Please see Figure 1 This invention provides a method for intelligent deployment of a distributed self-organizing network for ballistic surveillance, comprising the following steps:

[0059] S1: Obtain the sending time of the control ball node and the receiving time of the receiving node during the broadcast signal transmission process. Perform matching on the sending and receiving records of the same broadcast signal. Construct a propagation duration record sequence based on the difference. Call the signal incident angle recorded by the direction sensing device of the receiving node to construct the direction vector. Construct a time path and direction path parallel record for each broadcast signal to generate a broadcast signal path trajectory group.

[0060] S2: Using the received signal path corresponding to the target node in the broadcast signal path trajectory group, the signal propagation duration and direction path are combined in pairs to construct duration and direction reference items. Path matching degree judgment and signal arrival sequence consistency detection are performed on the reference items corresponding to the target node and the receiving node. Receiving nodes with path consistency higher than the threshold judgment value are filtered out. Broadcast records between the receiving node and the target node are called to generate a stable set of adjacent nodes.

[0061] S3: Using a set of stable adjacent nodes, extract the signal transmission records of the initial transmission period of the control ball node, obtain the power change time data of the RF output channel in the short time before and after transmission, perform spectrum transformation on the power change time data, construct a frequency domain response profile map, perform matching rate segment comparison processing on the profile map, verify the source attributes of the node signal structure behavior, and generate a node identity feature tag set.

[0062] S4: By using the node identity feature tag set, call the periodic image collection content to obtain the pixel coordinates of the edge region and the pixel displacement trajectory between consecutive frames, perform direction aggregation processing on the trajectory, filter the pixel block region that continuously shifts in the same direction, perform pixel density change trend statistics on the pixel block region, and generate a dynamic occlusion feature region set.

[0063] S5: Based on the pixel occlusion region position marked by the dynamic occlusion feature region set, combined with the direction vector information of stable adjacent nodes, the region is constructed to cross the region between multiple nodes. The proportion of the cross region in the real-time node view is judged and cross-validated with the redundancy information of the collaborative nodes. The view of the main view node is classified and processed to generate the intelligent deployment ball network connectivity table.

[0064] The broadcast signal path trajectory group includes a time path sequence, a directional path sequence, and a path matching reference model. The stable adjacent node set includes a high-matching node index, path consistency features, and target node reception frequency statistics. The node identity feature tag set includes a frequency domain energy distribution pattern, power spectral density curve features, and signal structure behavior encoding. The dynamic occlusion feature region set includes a directional consistency identifier block, a pixel density change map, and inter-frame occlusion trend feature parameters. The intelligent deployment ball network connectivity table includes a main viewpoint node index, a cooperative viewpoint overlap evaluation value, and a node connectivity state matrix.

[0065] Please see Figure 2 The specific steps of S1 are as follows:

[0066] S101: Obtain the sending time of the control ball node and the receiving time of the receiving node during the broadcast signal transmission process, calculate the time difference between each group of sending and receiving times, extract the broadcast signal index corresponding to the time difference, match the sending record and the receiving record according to the same signal index, establish a time difference sequence based on the matching result, call each group of data frames in the time difference sequence and aggregate them in sequence to generate a propagation duration record sequence;

[0067] This is achieved by integrating a high-precision clock module within the surveillance sphere and combining it with a time synchronization mechanism. When a node transmits a broadcast signal, it automatically records the current time and binds it to a unique signal number. This number is sent along with the signal. Upon receiving the signal, the receiving node immediately records its local reception time and extracts the signal number to ensure correct data association. After recording the signal reception from multiple receiving nodes, a one-to-one or one-to-many time record lookup table is established by mapping the transmission and reception times using the signal number. For example, when the surveillance sphere transmits a broadcast signal with the number T001, the transmission time is set to 10.000 seconds. Points A, B, and C record reception times of 10.004, 10.006, and 10.005 seconds, respectively, corresponding to time differences of 0.004, 0.006, and 0.005 seconds. By traversing the number and corresponding time records, the time difference values ​​are extracted and a time difference sequence list indexed by the number is constructed. Traversing this sequence, the time difference values ​​under the same number are aggregated. Aggregation can be achieved by sorting by number and combining them sequentially. For example, the time difference under number T001 is arranged into a group of data frames according to the receiving node order, and the data with numbers T002, T003, etc. are processed sequentially to generate a propagation duration record sequence.

[0068] S102: Based on the propagation duration record sequence, call the recorded signal incident angle information, perform polar coordinate transformation on the incident angle data frame, and perform cosine direction derivation operation on the transformation result and the spatial position parameter set of the receiving node to generate a direction vector sequence set;

[0069] The receiving node retrieves the signal incident angle information corresponding to its number. The receiving node is equipped with a direction sensing device, which synchronously measures the incident direction of the signal when receiving the signal and records it, binding it with the signal number. The incident angle information is organized into a data frame and combined with the propagation duration data to form relative position information. At this point, the incident angle data needs to be converted into a direction pointing data structure to form a direction vector that can be calculated or stored. Each incident angle data is fused with the corresponding node spatial position. The spatial position can be obtained from preset coordinates, on-site deployment data, or calibration information during installation. After the incident angle and coordinates are combined, the direction vector is calculated to represent the signal arrival path. This vector is the unit path line extending from the receiving point towards the signal source direction. Specifically, for each numbered broadcast signal, a set of receiving node direction vector sequences is formed. Multiple broadcast signals will form a set of direction vector sequence sequences.

[0070] S103: Based on the direction vector sequence set, perform path alignment matching between the propagation duration data and direction vector data corresponding to the same signal index, perform index concatenation operation between the propagation time sequence and direction vector sequence of each broadcast signal, merge the broadcast signal path information, and generate a broadcast signal path trajectory group.

[0071] For broadcast signals with the same number, the propagation duration and direction vector are mapped one-to-one with the receiving node number to form path data units. The path data units under the number are categorized and combined to form a complete path group for the broadcast signal. Each unit in the path group represents the propagation characteristics of the signal from the control sphere to a certain receiving node, including the propagation duration and path direction. To ensure the logical integrity and timing correctness of the path data, a path index is established according to the signal number. The path data groups are then formed into an index structure. The structure is used to call the drawing or stored program to generate and maintain the path trajectory group. The path is set in signal number T005, corresponding to four receiving nodes, forming four path vector units. After merging, it becomes the path group of T005. In the data formed by multiple signal numbers, the path groups are compared to extract the spatial path continuity and direction change characteristics to generate the broadcast signal path trajectory group.

[0072] Please see Figure 3 The specific steps of S2 are as follows:

[0073] S201: Call the received signal path corresponding to the target node in the broadcast signal path trajectory group, and perform combination matching using a dual indexing method based on the propagation duration and corresponding direction vector data of each broadcast signal in the path. Establish key-value correspondence between the propagation duration data and the direction vector data according to the signal index, and generate a duration and direction reference item sequence.

[0074] The path records containing the target node are selected from the path trajectory group. Each record contains propagation duration information and direction vector data related to the target node. The data is organized by signal number index in the path trajectory structure. By traversing the path trajectory group, it is determined whether each broadcast path contains the target node as the receiver. If it does, the complete record of the path is extracted and cached as the target node path set. Dual indexes are established for the propagation duration information and direction vector information in each record of the path set. The signal number is used as the primary index key. At the same time, a mapping table is constructed with the receiver node number as the secondary index. Two key-value pair mapping tables are generated for matching operations. The propagation duration dataset is constructed as a structure with the number as the key and the propagation time as the value. At the same time, the direction vector data is constructed as a structure with the number as the key and the vector value as the value. The two sets of structures are combined and matched in parallel to generate a sequence of duration and direction reference items.

[0075] S202: Based on the time-direction reference item sequence, select the reference items corresponding to the target node and the receiving node respectively, compare the corresponding paths between the two using the path index relocation method, and perform node sequence consistency detection and propagation time difference calculation operations on the corresponding path group. Extract the nodes whose path consistency scores are higher than the preset path matching threshold and generate a path consistent node index set.

[0076] The path consistency score is calculated using the following formula:

[0077] ;

[0078] in, This represents the path consistency score, where N represents the total number of node pairs. This represents the propagation time of the i-th node pair along the propagation path at the receiving node. This represents the propagation time of the i-th node pair along the propagation path to the target node. This represents the propagation time of the (i+1)th node pair on the path to the receiving node. This represents the propagation time of the (i+1)th node pair on the path to the target node. denoted by , represents the difference in propagation distance between the corresponding nodes of the receiving path and the target path in the i-th node pair; α represents an empirical factor for adjusting the propagation distance difference; and δ represents a positive constant. The node-to-path difference weight decay factor;

[0079] The formula's calculation logic is as follows: By normalizing the propagation time difference between corresponding nodes in the target path and the receiving path, and then combining this with the weighted attenuation factor of the propagation distance between nodes, a quantitative assessment of the overall path matching degree is achieved. The formula first takes the absolute difference in propagation time for each pair of nodes. This reflects the time difference between the two paths at the nodes, and is then calculated by summing the propagation intervals between the preceding and following nodes and adding an empirical term. To construct a normalization factor and prevent differences from being amplified or distorted, a normalization factor is introduced. Distance decay weighting reduces the impact of errors from distant nodes, and the results of the nodes are averaged to obtain the path consistency score.

[0080] The path consistency score represents the degree of matching between the receiving path and the target path in time and space. The lower the score, the smaller the difference between the two paths in terms of propagation time and propagation distance, and the better the consistency. The overall path matching effect is reflected by weighted average, which combines the node propagation time difference and distance difference.

[0081] Parameter definition and data acquisition method description:

[0082]

[0083]

[0084] Explanation of quantification of non-numerical data:

[0085] parameter Although derived from location coordinates, the actual data is numerical and does not require direct quantization. However, if the original input is a node identifier (such as NodeA, NodeB), it needs to be converted into a numerical distance, as shown in the following example:

[0086] NodeA has coordinates (10, 20), and NodeB has coordinates (13, 24). The distance difference is:

[0087] ;

[0088] experience Set as This value can be obtained by testing the minimum propagation interval in a stable network scenario. The purpose of setting this value is to prevent the denominator from becoming unstable due to extremely small propagation time differences.

[0089] Control coefficient The influence of distance on propagation time was fitted experimentally, with the value range being [value missing]. This is set to The results are derived from the linear fitting of the average distance change between the five groups of nodes and the propagation delay difference.

[0090] Weighting factors This reflects the reduced impact of propagation distance on time differences in the evaluation. If the distance is 1, then the weight of this item is 1; as the distance increases, the weight decreases.

[0091] Parameter settings and actual calculation examples:

[0092] Based on actual network monitoring data, the parameters are set as follows (taking 3 nodes as an example):

[0093] Table 1: Propagation Delay and Distance Difference of Path Nodes

[0094] As shown in Table 1, the propagation time and distance difference for each node were obtained from actual measurements and simulation calculations.

[0095] Substitute the terms into the calculation:

[0096] For the first node:

[0097] ;

[0098] ;

[0099] ;

[0100] ;

[0101] Node 2:

[0102] ;

[0103] 6 + 6 + 1 = 13;

[0104] ;

[0105] ;

[0106] 3rd node:

[0107] ;

[0108] 6 + 7 + 1 = 14;

[0109] ;

[0110] ;

[0111] Substitute into the formula to calculate:

[0112] ;

[0113] The results indicate that the weighted consistency score between the monitored path and the target path in terms of propagation delay and distance differences is 0.1024. If the threshold for good expected consistency is set at 0.15 (determined by regression analysis of the experimental system's average value), then: ,

[0114] Therefore, the node is included in the subsequent path similarity index set;

[0115] The advantage of the formula lies in the propagation of distance weighting factors. The introduction of this approach not only considers time differences but also regulates spatial distance, effectively avoiding the amplification of the impact of distant nodes with high differences on the overall consistency evaluation, improving the robustness and selection accuracy of path recognition, and enabling better identification of path sequences with local consistency rather than relying solely on similar global absolute values ​​in overall path matching, thus significantly improving the intelligence level of path reasoning.

[0116] S203: Call the broadcast records between the receiving node and the target node in the path-consistent node index set, extract and normalize the signal arrival time difference in the broadcast records, perform secondary index filtering based on the encoding result and the path-consistent node index, aggregate the receiving node numbers that meet the filtering conditions, and generate a stable neighbor node set.

[0117] The system retrieves the broadcast records of each node in the index set between the target node and the target node, extracts the signal arrival time difference from the records, which represents the relative difference in reception time between the target node and the remaining nodes under a certain broadcast event. This time difference reflects the degree of propagation synchronization between nodes. The extracted time difference is normalized to map time differences of different magnitudes to a unified interval, typically set to 0 to 1. After processing, a standardized encoded value is obtained. Then, a secondary screening is performed based on the encoded value and the node index with consistent path. That is, based on the similar path structure, nodes with large differences are filtered out from the perspective of propagation time, and node numbers with close time differences are retained. This reflects the group of receiving nodes that are highly consistent with the target node in terms of structural position and propagation characteristics among multiple broadcast signals. It is the key input data for constructing a local propagation model or network topology mapping, forming a stable set of adjacent nodes.

[0118] Please see Figure 4 The specific steps of S3 are as follows:

[0119] S301: Call the number information of the receiving node in the stable adjacent node set, locate the signal transmission record of the deployed ball node within a set time interval before and after the receiving time of the corresponding adjacent node, extract the radio frequency output channel number associated with each record, and retrieve the channel power sampling value sequence within a fixed time window before and after transmission, aggregate them according to the sampling time order, and generate the channel power time series.

[0120] By performing a location operation on each number, the signal transmission records of the deployed ball node within the time range before and after the signal received by the receiving node are queried in the signal database. The time range is set to a fixed-length time window, 10 milliseconds before and after the reception time. The system searches for a transmission record corresponding to the reception event of an adjacent node within this time window. If a record exists, the radio frequency output channel number used in the record is extracted. Each record contains the transmission channel number, transmission timestamp, and remaining parameter information. Centered on the transmission time of each record, within a fixed time period of 5 milliseconds before and after, the power sampling data of the channel within the time period is retrieved. The power data is the result of multi-point sampling within a continuous time period, stored at a fixed frequency, such as once every 0.1 milliseconds, forming a power value sequence composed of multiple power points. The sampled values ​​are aggregated by sorting by timestamp, and the power data belonging to the same time window are spliced ​​together to form a channel power time series.

[0121] S302: Based on the channel power time series, the data sequence is divided into segments of equal length. Fourier transform is performed on each power segment to extract the frequency domain amplitude vector and the corresponding frequency index to form a frequency domain data frame. Frequency index sorting and amplitude normalization operations are performed on the frequency domain data frame. The frequency domain data frames are spliced ​​together in time order to construct a two-dimensional frequency domain spectrum and generate a frequency domain response profile.

[0122] The channel power time series is divided into segments of equal length, with the segment length determined by set parameters, such as 100 sampling points per segment. The entire sequence is divided into several segments, and a Fourier transform is applied to each segment to convert it into a frequency domain representation. The amplitude distribution and corresponding frequency index values ​​of the segments in the frequency domain are obtained, and the results together form a frequency domain data frame. Each data frame contains data pairs with frequency and amplitude. The frequency index values ​​within each frequency domain data frame are sorted in ascending order to ensure that the segments are arranged consistently on the frequency axis. Then, the amplitude data is normalized to make different segments comparable in the amplitude dimension. The normalization method is to set the maximum amplitude of each segment to 1 and scale the remaining amplitudes proportionally. After normalization, the frequency domain data frames arranged in chronological order are spliced ​​together to construct a two-dimensional frequency domain spectrum. The horizontal axis is the time segment number, and the vertical axis is the frequency index. Each point in the spectrum represents the standardized amplitude of a frequency point in a certain time segment, generating a frequency domain response profile.

[0123] S303: Call the amplitude trajectory of the frequency segment in the frequency domain response contour map, perform same-frequency segment matching rate calculation on the contour map between adjacent nodes, extract the frequency index with a matching rate higher than the preset segment comparison judgment threshold, perform feature aggregation on the behavior waveform pattern corresponding to the frequency index, bind the node number and signal behavior features, and generate a node identity feature tag set.

[0124] For the frequency domain profile maps of any two adjacent nodes in a stable set of adjacent nodes, a matching rate calculation operation for the same frequency segment is performed. The frequency segments in the two profile maps are aligned, and the frequency axis is divided into the same segment range. A frequency index is set with 10Hz as a segment. The amplitude change trajectory of the corresponding segment in the same time segment is compared segment by segment. If the amplitude change trends of the two nodes in a certain segment are similar, the segment is considered a match. The matching status of each frequency segment is recorded, and the overall segment matching rate is calculated, which is the proportion of the number of matching segments to the total number of segments. The proportion is compared with the set comparison judgment threshold. For example, if the threshold is set to 0.75, if the matching rate reaches 0.8, the frequency segment is retained. After extracting the frequency index that meets the matching requirements, the amplitude change sequence covered by the frequency index is called, and the behavior trajectory in different time periods is aggregated and analyzed to generate the behavior waveform pattern corresponding to the frequency index. The pattern is bound to the corresponding receiving node number and recorded in the signal behavior database. The node and the corresponding behavior waveform pattern are structurally combined to generate a node identity feature tag set.

[0125] Please see Figure 5 The specific steps of S4 are as follows:

[0126] S401: Call the identification information associated with the target node in the node identity feature tag set, obtain the corresponding periodic image collection content, locate the contour boundary of the edge region in each frame image, extract the pixel coordinate set of the pixel block in the edge region, perform position difference operation on the pixel block with the same coordinate position between adjacent frames, construct the continuous inter-frame displacement path of the pixel block in time order, and generate a set of pixel displacement trajectories.

[0127] The system collects periodic images corresponding to the location and identification of data. These images come from image acquisition devices deployed in the controlled area. A sequence of image frames is captured at fixed intervals to form an image stream. Image frames related to the activity time period of the target node are extracted from this stream. The edge regions of each frame are processed using image processing algorithms to perform edge contour localization. The edge region is defined as an area within a certain pixel range from the image boundary. A contour extraction method is used to extract significant boundary contours within this region, obtaining a set of pixel blocks constituting the edge region. For each pixel block, a set of pixel coordinates in the image is extracted, including coordinates of the center position and boundary points. After edge extraction, a difference calculation operation is performed on pixel blocks between consecutive image frames. This mainly involves comparing the positional changes of pixel blocks with the same coordinates in adjacent frames to generate pixel displacement information. The displacements of each pixel block in consecutive frames are connected sequentially over time to form pixel displacement trajectories composed of multiple frames. These trajectories represent the minute movements of objects or backgrounds in the edge region of the image over time, providing basic data for analyzing behavioral features such as occlusion and dynamic recognition, and generating a set of pixel displacement trajectories.

[0128] S402: Based on the start and end coordinates of each trajectory in the pixel displacement trajectory set, perform displacement vector angle calculation, quantize and group the angle values, perform numbering and classification processing on the trajectories in the same direction, perform continuous intra-frame consistency judgment on the classified pixel block set, filter pixel block regions with the same offset direction, and generate a continuous same-direction offset pixel block set.

[0129] The starting and ending coordinates of each trajectory are analyzed. By calculating the angle value formed by the displacement vector between the two points, the movement direction of the pixel block within the entire observation period is determined, which facilitates subsequent classification and processing. The angle value is quantized and grouped, and the direction groups are divided according to equal intervals, such as 30 degrees per direction segment, for a total of 12 direction groups. Each pixel displacement trajectory is assigned to the corresponding numbered direction group according to its direction. Trajectories within the same direction group are numbered and classified. Pixel blocks belonging to the same direction are grouped into a classification unit. The positional characteristics of the pixel blocks in the classification unit in consecutive frames are analyzed, that is, whether the pixel block continuously shifts in the direction in adjacent image frames. If the shift direction remains consistent in multiple frames, it is considered to be continuous and consistent. The trajectory is screened for consistency, and pixel blocks with unstable or large changes in shift direction between frames are removed. Only pixel blocks with a clear direction and stable movement in continuous time are retained, generating a set of pixel blocks with continuous unidirectional shift.

[0130] S403: Call the number information of the pixel block region in the continuously offset pixel block set, perform frame-by-frame difference operation on the pixel density value of the pixel block region in consecutive frames and record the density difference sequence, calculate the density difference trend change rate in time order, perform index tag aggregation on the pixel block region with continuously increasing trend change rate, and generate a dynamic occlusion feature region set.

[0131] The pixel density values ​​of pixel block regions in the set are analyzed within consecutive image frames. Pixel density represents the number of pixels with non-background features per unit area. The density of the same pixel block region in each frame is statistically analyzed, and the difference between adjacent frames is calculated to form a density difference sequence recorded frame by frame. This sequence is used to analyze the content changes of pixel block regions at different time points. The trend change rate of the density difference sequence is calculated to analyze the direction and speed of density value changes. Special attention is paid to continuously increasing trend segments, that is, the pixel density of a certain pixel block region continues to rise in multiple frames, which represents the gradual entry of occlusion or moving objects into the region. The trend change rate curve of each region is calculated by sliding within a set time window. Pixel block regions with a continuous growth trend within a specific time period are selected, and the regions are indexed and marked. Image regions marked as having dynamic change characteristics are aggregated to generate a set of dynamic occlusion feature regions.

[0132] Please see Figure 6 The specific steps of S5 are as follows:

[0133] S501: Based on the pixel coordinate range of the occluded region in the dynamic occlusion feature region set, call the direction vector information corresponding to the node in the stable adjacent node set, perform the direction vector reverse extension positioning operation on the occluded region, extract the adjacent node number and direction vector combination of the pointing region, and perform cross projection superposition on the direction vector in the image coordinate system to generate a multi-node cross region mapping group.

[0134] The system locates the position boundary in the image space and calls the direction vector information of each receiving node in the stable neighboring node set. The direction vector has been bound to each node number in the previous step and is used to represent the observation direction of the node relative to the signal source or image object. Each direction vector is inverted and a reverse extension operation is performed, that is, the vector direction is adjusted from the original direction of the node to the direction pointing to the occluded area. The extension path is extended from the node position to the image boundary along the vector direction. The purpose is to determine whether the observation path of the node crosses the target occluded area. If the extension path intersects or overlaps with the pixel boundary of the occluded area in the image coordinate system, the direction vector is determined to be validly pointing to the occluded area. The node number and the direction vector are combined into a valid identifier record. The system performs detection on the stable neighboring nodes to obtain the nodes pointing to the same occluded area and the combination of direction vectors. The valid vectors are cross-projected in the image coordinate system. Multiple direction vectors are superimposed on the image plane to mark the area covered by the cross projection and generate a multi-node cross area mapping group.

[0135] S502: Based on the node number corresponding to the region in the multi-node cross region mapping group, retrieve the visible region mask of the real-time frame image of the node, calculate the ratio of the pixel value occupied by the cross region in the node image to the total number of pixels in the image, construct the region proportion parameter of the cross region in the view, count the number of visible cross regions associated with the node, and generate a cross region view proportion and redundancy parameter set.

[0136] Extract the node number associated with each group of intersection regions, and call the visible region mask generated in the corresponding real-time frame image for the node. The mask identifies the effective area of ​​the visible range of the node in the current time image frame. The mask is represented in a binary map, where non-zero pixels represent the visible part and zero pixels represent occluded or unreachable areas. Locate the corresponding intersection region boundary on the mask map, and count the number of effective pixels covered by the region. Combined with the total number of pixels in the image frame, calculate the proportion of the intersection region in the node image. The result is expressed as the pixel proportion parameter of the intersection region in the image frame. Calculate the corresponding proportion for each intersection region involved by each node, count the number of intersection regions that each node can currently observe, record the region number and the corresponding pixel proportion value, and integrate the data into a cross-region view proportion and redundancy parameter set.

[0137] S503: Call the cross-region view proportion and redundancy parameter set, perform region classification operation on the main view node, determine whether the node is the main view node according to the proportion threshold, and mark the redundancy status of the node in combination with the redundancy count parameter. Map the visual coverage relationship between the nodes with node numbers and mark the connectivity status to generate the intelligent deployment ball network connectivity table.

[0138] The system performs a primary perspective judgment operation on the regional affiliation of nodes. A regional proportion threshold of 30% is set, and each node's proportion in the intersecting regions is checked one by one to see if it reaches or exceeds the threshold. If it does, the node is determined to be the primary perspective node of the region and is classified as the main observation node of the region. If the proportion does not reach the threshold but still has a certain coverage, it is recorded as an auxiliary perspective node. At the same time, the redundancy parameter is combined to evaluate the number of regions observed by each node in real time. If the number of intersecting regions associated with a node is higher than the average level, the node is marked as a high redundancy node, otherwise it is marked as a low redundancy or unique perspective node. Each node number is mapped with perspective attributes and redundancy marks. A connectivity table between nodes is established with the node number as the key. By identifying the common observation relationship between nodes in the same region, a graph-structured connectivity matrix is ​​constructed to indicate which nodes can achieve information sharing or collaborative coverage through the visual path, generating a connectivity table for the intelligent deployment ball network.

[0139] Please see Figure 7 A distributed self-organizing intelligent deployment ball network system includes:

[0140] The trajectory extraction module obtains the transmission time parameters of the broadcast signal from the control ball node and the reception time parameters of the receiving node, calls the signal propagation start and end time to calculate the propagation time length sequence, combines the signal incident angle information recorded by the direction sensing device of the receiving node to obtain the direction path parameters, and matches the propagation time length sequence and direction path parameters according to the signal identifier to obtain the broadcast signal path trajectory group.

[0141] The path filtering module extracts path data between the target node and the receiving node based on the broadcast signal path trajectory group, obtains the propagation time length parameter and direction path parameter in each pair of paths, performs a combined judgment of the time length difference and the direction angle difference, and obtains a set of stable adjacent nodes.

[0142] The adjacency steady-state module calls the set of stable adjacency nodes, extracts the signal transmit power, receive power and signal frequency band parameters, calculates the signal power change trend under the same frequency band, judges the degree of stability in communication based on the number of times the receiving node is in the continuous communication record and the magnitude of the power fluctuation trend, and generates a set of node identity feature tags.

[0143] The signal tag module uses the node identity feature tag set to extract the power change time record at the beginning of the signal transmission, performs frequency domain feature transformation on the time record to form a frequency domain response spectrum, and judges the number, interval and energy change amplitude of features in the frequency band of the response spectrum to obtain a set of dynamic occlusion feature regions.

[0144] The occlusion modeling module calls the dynamic occlusion feature region set to obtain the edge pixel positions and displacement trajectories in continuous image frames, counts the aggregation degree of pixel offset direction in continuous frames, filters pixel regions with consistent directions, performs the attribution classification processing of the viewpoint of the main viewpoint node, and obtains the intelligent deployment ball network connectivity table.

[0145] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for intelligent deployment of satellite navigation networks using a distributed self-organizing network, characterized in that, Includes the following steps: S1: Obtain the sending time of the control ball node and the receiving time of the receiving node during the broadcast signal transmission process. Perform matching on the sending and receiving records of the same broadcast signal according to the signal index. Construct a propagation duration record sequence based on the difference between the sending and receiving times. Generate a direction vector sequence based on the signal incident angle recorded by the receiving node. Associate the propagation duration record corresponding to the same signal index with the direction vector sequence to generate a broadcast signal path trajectory group. S2: Using the received signal path corresponding to the target node in the broadcast signal path trajectory group, the propagation duration of each broadcast signal and the corresponding direction vector are combined into a duration direction reference item. The propagation path structure sequence consistency detection and propagation duration difference calculation operations are performed on the reference items corresponding to the target node and the receiving node respectively. Nodes with path consistency scores lower than the preset path matching threshold are extracted to generate a stable neighbor node set. The path consistency score is calculated using the following formula: Where AS represents the path consistency score, and N represents the total number of node pairs. This represents the propagation time of the i-th node pair along the propagation path to the target node. This represents the propagation time of the (i+1)th node pair on the path to the receiving node. This represents the propagation time of the (i+1)th node pair on the path to the target node. α represents the difference in propagation distance between the corresponding nodes of the receiving path and the target path in the i-th node pair, and α represents an empirical factor for adjusting the propagation distance difference. Represents a positive constant. The node-to-path difference weight decay factor; S3: Using the set of stable adjacent nodes, extract the signal transmission record of the initial period of the control ball node's transmission, obtain the power change time data of the radio frequency output channel in the short period before and after transmission, perform spectrum transformation on the power change time data, and generate a node identity feature tag set; The specific steps for S3 are as follows: S301: Call the number information of the receiving node in the set of stable adjacent nodes, locate the signal transmission record of the control ball node within a set time interval before and after the receiving time of the corresponding adjacent node, extract the radio frequency output channel number associated with each record, and retrieve the channel power sampling value sequence within a fixed time window before and after transmission, aggregate them according to the sampling time order, and generate a channel power time series. S302: Based on the channel power time series, the data sequence is divided into segments of equal length. Fourier transform is performed on each power segment to extract the frequency domain amplitude vector and the corresponding frequency index to form a frequency domain data frame. Frequency index sorting and amplitude normalization operations are performed on the frequency domain data frame. The frequency domain data frames are spliced ​​together in time order to construct a two-dimensional frequency domain spectrum and generate a frequency domain response profile. S303: Call the amplitude trajectory of the frequency segment in the frequency domain response contour map, perform same-frequency segment matching rate calculation on the contour map between adjacent nodes, extract the frequency index with a matching rate higher than the preset segment comparison judgment threshold, perform feature aggregation on the behavior waveform pattern corresponding to the frequency index, bind the node number and signal behavior features, and generate a node identity feature tag set. S4: Using the node identity feature tag set, call the periodic image collection content to obtain the pixel coordinates of the edge region and the pixel displacement trajectory between consecutive frames, perform direction aggregation processing on the trajectory, filter pixel block regions with the same offset direction, and generate a dynamic occlusion feature region set. S5: Based on the pixel occlusion region position marked by the dynamic occlusion feature region set, combined with the direction vector information of stable adjacent nodes, construct the intersection region between multiple nodes, determine the proportion of the intersection region in the real-time node view and cross-validate it with the redundancy information of the collaborative nodes, perform the attribution classification processing of the view of the main view node, and generate the intelligent deployment ball network connectivity table. The intelligent deployment ball network connectivity table includes the main view node index, the collaborative view overlap evaluation value, and the node connectivity state matrix.

2. The intelligent deployment ball network method for distributed self-organizing networks according to claim 1, characterized in that, The broadcast signal path trajectory group includes a time path sequence, a direction path sequence, and a path matching reference model. The stable neighbor node set includes a high-matching-degree node index, path consistency feature quantity, and target node reception frequency statistics. The node identity feature tag set includes a frequency domain energy distribution pattern, power spectral density curve features, and signal structure behavior encoding. The dynamic occlusion feature region set includes a direction consistency identifier block, a pixel density change map, and inter-frame occlusion trend feature parameters.

3. The intelligent deployment ball network method for distributed self-organizing networks according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain the sending time of the control ball node and the receiving time of the receiving node during the broadcast signal transmission process, calculate the time difference between each group of sending and receiving times, extract the broadcast signal index corresponding to the time difference, match the sending record and the receiving record according to the same signal index, establish a time difference sequence based on the matching result, call each group of data frames in the time difference sequence and aggregate them in sequence to generate a propagation duration record sequence; S102: According to the propagation duration recording sequence, call the recorded signal incident angle information, perform polar coordinate transformation on the incident angle data frame, and perform cosine direction derivation operation on the transformation result and the spatial position parameter set of the receiving node to generate a direction vector sequence set; S103: Based on the set of direction vector sequences, perform path alignment matching between the propagation duration data and direction vector data corresponding to the same signal index, perform index concatenation operation between the propagation time sequence and direction vector sequence of each broadcast signal, merge the broadcast signal path information, and generate a broadcast signal path trajectory group.

4. The intelligent deployment ball network method for distributed self-organizing networks according to claim 3, characterized in that, The specific steps of S2 are as follows: S201: Call the received signal path corresponding to the target node in the broadcast signal path trajectory group, and perform combination matching using a dual indexing method based on the propagation duration and corresponding direction vector data of each broadcast signal in the path. Establish key-value correspondence between the propagation duration data and the direction vector data according to the signal index, and generate a duration-direction reference item sequence. S202: Based on the time direction reference item sequence, select the reference items corresponding to the target node and the receiving node respectively, compare the corresponding paths between the two using the path index relocation method, and perform node sequence consistency detection and propagation time difference calculation operations on the corresponding path group, extract the nodes whose path consistency scores are lower than the preset path matching threshold, and generate a path consistent node index set. S203: Call the broadcast records between the receiving node and the target node in the path-consistent node index set, extract and normalize the signal arrival time difference in the broadcast records, perform secondary index filtering based on the encoding result and the path-consistent node index, aggregate the receiving node numbers that meet the filtering conditions, and generate a stable neighbor node set.

5. The intelligent deployment ball network method for distributed self-organizing networks according to claim 4, characterized in that, For each corresponding path in the path-consistent node index set, the direction vector data included in the reference items corresponding to the target node and the receiving node are sequentially compared using a bidirectional traversal method. During the comparison process, the change in the angle between the direction vector of any node and the direction vector of the previous node is determined. If the change in the angle is less than a preset direction change threshold, the node is marked as a sequentially consistent node. In the propagation duration difference calculation operation, the propagation duration difference corresponding to nodes with consistent order is calculated by weighted accumulation. The weight of the propagation duration difference is determined based on the cosine value of the angle between the corresponding direction vector and the reference direction. When the cosine value is greater than the preset direction consistency threshold, a high weight is assigned. The propagation duration difference after weighted accumulation is compared with the standard deviation of the propagation duration of the corresponding path. When the ratio is lower than the set duration consistency judgment threshold, the sequentially consistent node is included in the path consistent node index set.

6. The intelligent deployment ball network method for distributed self-organizing networks according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Call the identification information associated with the target node in the node identity feature tag set, obtain the corresponding periodic image collection content, locate the contour boundary of the edge region in each frame image, extract the pixel coordinate set of the pixel block in the edge region, perform position difference operation on the pixel block with the same coordinate position between adjacent frames, construct the continuous inter-frame displacement path of the pixel block in time order, and generate a set of pixel displacement trajectories. S402: Based on the start and end coordinates of each trajectory in the pixel displacement trajectory set, perform displacement vector angle calculation, quantize and group the angle values, perform numbering and classification processing on the trajectories in the same direction, perform continuous intra-frame consistency judgment on the classified pixel block set, filter pixel block regions with consistent offset directions, and generate a set of continuously offset pixel blocks in the same direction. S403: Call the numbering information of the pixel block region in the continuously offset pixel block set, perform frame-by-frame difference operation on the pixel density value of the pixel block region in consecutive frames and record the density difference sequence, calculate the density difference trend change rate in time order, perform index tag aggregation on the pixel block region with continuously increasing trend change rate, and generate a dynamic occlusion feature region set.

7. The intelligent deployment ball network method for distributed self-organizing networks according to claim 1, characterized in that, The specific steps of S5 are as follows: S501: Based on the pixel coordinate range of the occluded area in the set of dynamic occlusion feature areas, call the direction vector information corresponding to the nodes in the set of stable adjacent nodes, perform a direction vector reverse extension positioning operation on the occluded area, extract the adjacent node number and direction vector combination of the pointing area, and perform cross projection superposition on the direction vector in the image coordinate system to generate a multi-node cross area mapping group. S502: Based on the node number corresponding to the region in the multi-node cross region mapping group, retrieve the visible region mask of the real-time frame image of the node, calculate the ratio of the pixel value occupied by the cross region in the node image to the total number of pixels in the image, construct the region proportion parameter of the cross region in the view, count the number of visible cross regions associated with the node, and generate a cross region view proportion and redundancy parameter set. S503: Call the cross-region view proportion and redundancy parameter set, perform region classification operation on the main view node, determine whether the node is a main view node according to the proportion threshold, and mark the redundancy status of the node in combination with the redundancy count parameter. Map the visual coverage relationship between the nodes with node numbers and mark the connectivity status to generate the intelligent deployment ball network connectivity table.

8. A distributed self-organizing intelligent deployment ball network system, characterized in that, The system is used to implement the intelligent deployment ball networking method for a distributed self-organizing network as described in any one of claims 1-7, and the system includes: The trajectory extraction module obtains the sending time of the control ball node and the receiving time of the receiving node during the broadcast signal transmission process. It performs matching on the sending and receiving records of the same broadcast signal according to the signal index, constructs a propagation duration record sequence based on the difference between the sending and receiving times, and generates a direction vector sequence based on the signal incident angle recorded by the receiving node. It associates the propagation duration records corresponding to the same signal index with the direction vector sequence to generate a broadcast signal path trajectory group. The path filtering module uses the received signal path corresponding to the target node in the broadcast signal path trajectory group, combines the propagation duration of each broadcast signal with the corresponding direction vector into a duration direction reference item, performs the propagation path structure sequence consistency detection and propagation duration difference calculation operation on the reference items corresponding to the target node and the receiving node, extracts the nodes whose path consistency score is lower than the preset path matching threshold, and generates a stable neighbor node set. The adjacency steady-state module calls the set of stable adjacency nodes, extracts the signal transmit power, receive power and signal frequency band parameters, calculates the signal power change trend under the same frequency band, judges the degree of stability in communication based on the number of times the receiving node is in the continuous communication record and the magnitude of the power fluctuation trend, and generates a set of node identity feature tags. The signal tag module uses the node identity feature tag set to call the periodic image collection content to obtain the pixel coordinates of the edge region and the pixel displacement trajectory between consecutive frames, performs direction aggregation processing on the trajectory, filters the pixel block region with the same offset direction, and generates a dynamic occlusion feature region set. The occlusion modeling module calls the dynamic occlusion feature region set to obtain the edge pixel positions and displacement trajectories in continuous image frames, counts the aggregation degree of the offset direction of pixels in continuous frames, filters pixel regions with consistent directions, performs the attribution classification processing of the viewpoint of the main viewpoint node, and obtains the intelligent deployment ball network connectivity table.

Images