Multi-carrier data on-orbit detection system based on LEO satellite

Abstract

The invention provides a multi-carrier data on-orbit detection system based on an LEO satellite, and relates to the field of satellite monitoring, the system comprises a data acquisition and preprocessing module, a task scheduling module, a cluster resource management module, a task processing module, a data transmission module and a ground station processing module; obtaining and preprocessing a carrier signal to obtain task data; acquiring satellite node state information; according to the satellite node state information, tasks are distributed in combination with a task scheduling algorithm MDL based on dynamic scheduling; processing the distributed task to obtain a filtering result; and monitoring the driving state of the carrier according to the received filtering result. In-orbit data analysis is carried out on a carrier signal, and maximization of the ratio of normal data in downlink transmission data is taken as a target under the condition that the maximum tolerant transmission delay is met.

Description

Technical Field

[0001] This application relates to the field of satellite monitoring, and particularly to a multi-vehicle data on-orbit detection system based on LEO satellites. Background Art

[0002] In modern fields such as aviation and shipping, monitoring the operating conditions of multiple vehicles is crucial for ensuring safety and improving traffic management efficiency. Currently, there are many technologies for monitoring vehicle information.

[0003] Traditional vehicle information monitoring technologies broadcast various status information of vehicles in plain text, lacking message encryption and authentication mechanisms. It is easy to conduct spoofing and interference attacks on the monitoring system using simple hardware or software-defined radio (SDR). Although space-based vehicle signal monitoring systems offer significant advantages, their implementation still faces challenges. Related research has shown that the migration of vehicle signal monitoring systems from ground-based to space-based has not fundamentally solved the security problem.

[0004] Space-based vehicle signal monitoring systems transfer the signal receiving devices originally installed on the ground to satellites. Their related protocols have not been modified, and the positions of satellites are publicly available for easy positioning. In this case, although the attacker is far from the satellite, by simply increasing the power of the attack device, the attack can be completed, seriously disrupting the traffic system and having a greater impact on the monitoring system. Therefore, how to ensure the secure transmission of space-based vehicle signal monitoring systems has become an important challenge.

[0005] In recent years, with the gradual improvement of the payload resources of LEO satellites, deploying data analysis models on satellites and using multi-satellite collaboration for on-orbit processing of data security analysis has become an effective method. Using space computing for vehicle information detection can quickly identify normal data and abnormal data under attack in vehicle signals, filter abnormal data, ensure the transmission of normal data to the ground station, enhance the security of the traffic system, and improve the operating efficiency of the monitoring system. However, resources such as computing power and energy on satellites are limited, and the topology between satellites will change dynamically with the operating orbit. In addition, the high transmission delay of satellites further limits the tolerance time for on-orbit processing.

[0006] Therefore, how to perform efficient real-time satellite resource perception and efficient space computing task allocation strategies has become the key challenge for the data transmission ability of spaceborne signal monitoring systems. Summary of the Invention

[0007] The purpose of the present invention is to provide a multi-vehicle data on-orbit detection system based on LEO satellites to solve the problem that the existing space-based vehicle signal detection system is difficult to accurately schedule computing tasks, resulting in the inability to effectively filter abnormal data.

[0008] The above object of the present application is achieved by the following technical solutions:

[0009] The system includes: a data acquisition and preprocessing module, a task scheduling module, a cluster resource management module, a task processing module, a data transmission module, and a ground station processing module;

[0010] The data acquisition and preprocessing module, the task scheduling module, the cluster resource management module, the task processing module, the data transmission module, and the ground station processing module are sequentially connected in order; the task scheduling module is connected to the task processing module;

[0011] The data acquisition and preprocessing module is used to obtain vehicle signals and perform preprocessing to obtain task data; the vehicle signals include: ship signals and aircraft signals;

[0012] The cluster resource management module is used to obtain satellite node status information;

[0013] The task scheduling module is used to receive task data for analysis and accept satellite node status information; the task data is the preprocessed vehicle signal;

[0014] The task scheduling module is further used to allocate tasks to the task processing module for execution according to the satellite node status information, in combination with the task scheduling algorithm MDL based on dynamic scheduling;

[0015] The task processing module is used to perform abnormal data filtering processing on the received tasks, obtain a filtering result and input it to the data transmission module;

[0016] The data transmission module is used to transmit the filtering result to the ground station processing module through the constructed optimal space-ground link;

[0017] The ground station processing module is used to monitor the driving status of the vehicle according to the received filtering result.

[0018] Optionally, the preprocessing steps of the cluster resource management module include:

[0019] Let the set of original vehicle signal data received be where represents the m-dimensional feature vector of the i-th type of vehicle, and the preprocessing process is defined as:

[0020]

[0021] where Φ(·) is a multi-type vehicle data fusion function, and φ j (·) includes signal denoising, format standardization, and spatio-temporal alignment; the output is intermediate data in a unified format;

[0022] The preprocessed data Encapsulated as a task unit where τ i is the computational amount of each task in the task data, is the maximum tolerable time delay, and w i is the priority weight of each task in the task data.

[0023] Optionally, the state matrix S(t) is continuously updated through the cluster resource management module. The state matrix S(t) contains the resource states of N satellites at time t:

[0024]

[0025] where c i (t) is the remaining computing resource, m i (t) is the memory occupancy rate, and e i (t) is the energy reserve;

[0026] The satellite node state information includes: the remaining computing resource, the remaining computing resource, and the energy reserve;

[0027] Construct a real-time resource topology map through the cluster resource management module where v represents the set of all satellite nodes in the satellite cluster; ε represents the set of inter-satellite communication links between satellite nodes;

[0028] The edge weight of the resource topology map is the inter-satellite communication delay d jk (t), and k represents another satellite not equal to j.

[0029] Optionally, the specific steps of allocating tasks to the task processing module according to the satellite node state information and combining with the dynamic scheduling task scheduling algorithm MDL include:

[0030] S1: Perform pre-classification processing on tasks according to the maximum tolerable time delay of task data, specifically including:

[0031] If the computational amount τ of the task in the task data i is greater than the satellite's maximum computing capacity limit τ high , then split this task into multiple subtasks;

[0032] According to the maximum tolerable time delay of the task combined with the preset threshold δ t , mark the urgency of the task;

[0033] Adjust the priority weight w of the task according to the urgency of the task i ;

[0034] S2: Sort the tasks according to the maximum tolerable latency of the tasks and the adjusted priority weights, specifically including: according to the maximum tolerable latency T max , sort the tasks in ascending order; according to the priority weight w i of the tasks, sort the tasks in descending order;

[0035] Predict the total load of the satellite node cluster according to the task scheduling algorithm MDL based on dynamic scheduling and the satellite node status information;

[0036] S3: Dynamically allocate the tasks through the sorted tasks and the total load of the satellite node cluster. The specific steps are as follows:

[0037] Calculate the computing time and transmission time required for the active node j in the satellite node cluster to process the task, and verify whether the latency constraint of the task can be satisfied. The latency constraint is determined by the following formula:

[0038]

[0039] where d jk is the transmission delay between node j and other active nodes, B k is the bandwidth of the node; τ represents the computing requirement of the task; η j represents the computing efficiency of satellite j;

[0040] If the sum of the computing latency t comp and the transmission latency t trans does not exceed the maximum tolerable latency T max of the task, then this node is regarded as a feasible node and can carry the task;

[0041] Based on the multi-objective optimization strategy, determine the optimal node from the feasible nodes; the optimization objectives of the multi-objective optimization strategy include: the remaining computing resources c of the node, the remaining energy resources e of the node, the computing requirement τ of the task, and the communication delay d j，g of the task to the ground station;

[0042] Allocate the sorted tasks to the optimal node.

[0043] Optionally, the task processing module consists of a satellite constellation cluster that distributes and deploys a data analysis model.

[0044] Optionally, the data transmission module constructs an optimal satellite-ground link using an adaptive link selection method;

[0045] The data transmission module is used to select the current optimal satellite-ground communication link according to the satellite orbital position and the link status;

[0046] The data transmission module is also used to transmit the filtering result to the ground station processing module according to the currently optimal satellite-ground communication link;

[0047] The link status includes: bandwidth, latency, and bit error rate.

[0048] An electronic device includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions. The user interface and the network interface are used to communicate with other devices. The processor is used to execute the instructions stored in the memory so that the electronic device executes a multi-vehicle data on-orbit detection system based on LEO satellites.

[0049] A computer-readable storage medium stores instructions that, when executed, implement a multi-vehicle data on-orbit detection system based on LEO satellites.

[0050] The beneficial effects brought by the technical solution provided in this application are as follows:

[0051] 1. Utilize the computing resources of satellites to perform on-satellite analysis of multi-vehicle signals, improve the analysis and processing efficiency of vehicle abnormal data, and give full play to the advantages of the LEO satellite constellation and improve the management efficiency of ground control terminals.

[0052] 2. Design a task scheduling algorithm MDL based on dynamic scheduling, and propose a task scheduling system with three stages: task pre-classification, resource prediction and sorting, and dynamic task allocation to achieve efficient signal detection dynamic task scheduling.

[0053] 3. Design a distributed data detection model deployment plan based on the limited computing resources of LEO satellites to give full play to the advantages of collaborative computing of satellite constellation clusters. Description of the Drawings

[0054] The following will further illustrate this application in conjunction with the drawings. In the drawings:

[0055] Figure 1 is the system architecture diagram in the embodiment of this application;

[0056] Figure 2 is the experimental result diagram of the polar region in the embodiment of this application;

[0057] Figure 3 is the experimental result diagram of the ocean area in the embodiment of this application;

[0058] Figure 4 is the experimental result diagram of the high-density vehicle area in the embodiment of this application;

[0059] Figure 5 is the overall pseudo-code diagram of the algorithm in the embodiment of this application;

[0060] Figure 6 It is a schematic structural diagram of an electronic device in an embodiment of the present application. Specific embodiments

[0061] For a clearer understanding of the technical features, objectives, and effects of the present application, the specific embodiments of the present application will now be described in detail with reference to the accompanying drawings.

[0062] An embodiment of the present application provides a multi-vehicle data on-orbit detection system based on LEO satellites.

[0063] Please refer to Figure 1 , Figure 1 It is a system architecture diagram of a multi-vehicle data on-orbit detection system based on LEO satellites in an embodiment of the present application, including:

[0064] The system includes: a data acquisition and preprocessing module, a task scheduling module, a cluster resource management module, a task processing module, a data transmission module, and a ground station processing module;

[0065] The data acquisition and preprocessing module, the task scheduling module, the cluster resource management module, the task processing module, the data transmission module, and the ground station processing module are sequentially connected in series; the task scheduling module is connected to the task processing module;

[0066] The data acquisition and preprocessing module is used to obtain vehicle signals and perform preprocessing to obtain task data; the vehicle signals include: ship signals and aircraft signals;

[0067] The cluster resource management module is used to obtain satellite node status information;

[0068] In an embodiment of the present application, the cluster resource management module is responsible for node management, resource monitoring, status listening, etc. of the entire satellite node cluster.

[0069] The task scheduling module is used to receive task data for analysis and receive satellite node status information; the task data is the preprocessed vehicle signal;

[0070] The task scheduling module is further used to allocate tasks to the task processing module for execution according to the satellite node status information and in combination with the task scheduling algorithm MDL based on dynamic scheduling;

[0071] In one embodiment of the present application, in a LEO satellite constellation, in order to maximize resource utilization and meet the task processing delay, the present invention designs a task scheduling algorithm MDL based on dynamic scheduling. This algorithm realizes computational load balancing and timely processing of critical tasks by dynamically allocating tasks, considering scheduling constraints and task priorities. In order to make full use of the computing resources of each satellite, the algorithm ensures that the computational load borne by each satellite is as balanced as possible when allocating tasks, so that no single satellite is overloaded while others are idle.

[0072] The task processing module is used to perform abnormal data filtering processing on the received tasks, obtain the filtering results and input them into the data transmission module;

[0073] The data transmission module is used to transmit the filtering results to the ground station processing module through the constructed optimal space-ground link;

[0074] The ground station processing module is used to monitor the driving state of the vehicle according to the received filtering results.

[0075] The preprocessing steps of the cluster resource management module include:

[0076] As an embodiment, the data acquisition and preprocessing module is responsible for receiving signal data from various vehicles (such as airplanes, ships, etc.). Since the vehicle data formats, signal characteristics, etc. may vary, it is necessary to perform preprocessing and integration on the on-board platform, and the integrated data is delivered to the task scheduling module for processing.

[0077] Let the set of original vehicle signal data received be where represents the m-dimensional feature vector of the i-th type of vehicle, and the preprocessing process is defined as:

[0078]

[0079] where φ(·) is a multi-type vehicle data fusion function, and φ j (·) includes signal denoising, format standardization, and spatio-temporal alignment; the output is the intermediate data in a unified format;

[0080] The preprocessed data is encapsulated into a task unit where τ i is the amount of computation for each task in the task data, is the maximum tolerable delay, and w i is the priority weight for each task in the task data.

[0081] The state matrix S(t) is continuously updated through the cluster resource management module. The state matrix S(t) contains the resource states of N satellites at time t:

[0082]

[0083] Among them, c i (t) is the remaining computing resources, m i (t) is the memory occupancy rate, e i (t) is the energy reserve;

[0084] The satellite node status information includes: the remaining computing resources, the remaining computing resources, and the energy reserve;

[0085] Build a real-time resource topology map through the cluster resource management module Among them, v represents the set of all satellite nodes in the satellite cluster; ε represents the set of inter-satellite communication links between satellite nodes;

[0086] The edge weight of the resource topology map is the inter-satellite communication delay d jk (t), and k represents another satellite that is not equal to j.

[0087] The specific steps of allocating tasks to the task processing module according to the satellite node status information and combining with the task scheduling algorithm MDL for dynamic scheduling include:

[0088] As an example, initialize the task allocation matrix according to the resource topology map Initialize a copy of the resource pool. Initialize the satellite active nodes active_nodes, with the screening condition: satellites greater than or equal to the computing resource threshold θ c and the energy threshold θ e of the satellite.

[0089] S1: Perform pre-classification processing on the tasks according to the maximum tolerable delay of the task data, specifically including:

[0090] If the computing amount τ of the task in the task data i is greater than the satellite's maximum computing capacity limit τ high , then split this task into multiple subtasks;

[0091] According to the maximum tolerable delay of the task combined with the preset threshold δ t , mark the urgency of the task;

[0092] According to the urgency of the task, adjust the priority weight w of the task i ;

[0093] In an embodiment of the present application, the weight w of the task i will be multiplied by an emergency task coefficient λ u, so that resources can be preferentially allocated to this task. If the computing requirement of the task is large, that is, the computing volume τ of the task i is greater than the maximum computing capacity limit τ of the satellite high , then the task will be split into multiple subtasks to enable more flexible resource allocation.

[0094] S2: Sort the tasks according to the maximum tolerable delay and the adjusted priority weight of the tasks, specifically including: Sort the tasks in ascending order according to the maximum tolerable delay T max , and sort the tasks in descending order according to the priority weight w of the tasks i ;

[0095] Predict the total load of the satellite node cluster according to the task scheduling algorithm MDL based on dynamic scheduling and the satellite node status information;

[0096] In an embodiment of the present application, resource prediction and sorting: After the task pre-classification, the algorithm enters the second stage, that is, resource prediction and task sorting. The main task of this stage is to sort the tasks according to the maximum tolerable delay and priority weight of the tasks, and predict the load of each node in the cluster: The tasks will be sorted in ascending order according to their maximum tolerable delay T max , and sorted in descending order according to the priority weight w of the tasks at the same time. Such sorting can ensure that urgent and computationally demanding tasks are scheduled first. To effectively allocate tasks, the algorithm will predict the total load of the cluster. This is achieved by calculating the proportion of the computing resources c of all active nodes. This step helps the scheduling algorithm judge the overall load of the current cluster resources, so as to make reasonable task allocation decisions.

[0097] S3: Dynamically allocate the tasks through the sorted tasks and the total load of the satellite node cluster. The specific steps are as follows:

[0098] In an embodiment of the present application, for each task i, all its active nodes (that is, nodes that meet the activation conditions) will be checked first.

[0099] Calculate the computing time and transmission time required for the active node j in the satellite node cluster to process the task, and verify whether the delay constraint of the task can be satisfied. The delay constraint is determined by the following formula:

[0100]

[0101] where d jk is the transmission delay between node j and other active nodes, B k is the bandwidth of the node; τ represents the computing requirement of the task; η j represents the computing efficiency of satellite j;

[0102] If the sum of the computing delay \(t\) comp and the transmission delay \(t\) trans does not exceed the maximum tolerable delay \(T\) of the task max , then the node is considered a feasible node and can carry the task;

[0103] Based on the multi-objective optimization strategy, the optimal node is determined from the feasible nodes; the optimization objectives of the multi-objective optimization strategy include: the remaining computing resource \(c\) of the node, the remaining energy resource \(e\) of the node, the computing requirement \(\tau\) of the task, and the communication delay \(d\) from the task to the ground station j，g ;

[0104] The sorted tasks are assigned to the optimal node.

[0105] In an embodiment of the present application, this strategy considers multiple factors, including the remaining computing resource \(c\) of the node, the remaining energy resource \(e\) of the node, the computing requirement \(\tau\) of the task, and the communication delay \(d\) from the task to the ground station j，g . The goal is to select a node with the lowest comprehensive score to ensure the load balance and priority of task allocation. After selecting a good node, the task will be assigned to this node and the relevant resources will be reserved. After the task is assigned, the computing resource \(c\) and energy resource \(e\) of the corresponding node will be updated according to the requirements of the task. If the computing resource of the node is lower than the preset threshold \(\theta\) c or the energy resource of the node is lower than the threshold \(\theta\) e , then the node will be removed from the active node list. If there are not enough feasible nodes to assign tasks, the algorithm will adopt a backfilling mechanism. The purpose of the backfilling mechanism is to find available nodes within certain future time windows to ensure that the tasks can ultimately be processed. The overall pseudo-code of the algorithm Figure 5 is shown as follows.

[0106] The task processing module consists of a satellite constellation cluster that distributes and deploys the data analysis model.

[0107] In one embodiment of the present application, the task processing module consists of a satellite constellation cluster that distributes and deploys a data analysis model. It is responsible for receiving the task data transmitted by the task scheduling module and executing the corresponding task allocation strategy, and finally delivering the filtering results to the data transmission module. Among them, the model deployment strategy mainly concerns how to efficiently and flexibly deploy the data detection model on the on-board satellite platform to ensure efficient data processing in a resource-constrained environment. The main influencing factors include: the computing resources of the satellite, the diversity of data types, the real-time requirements for data processing, and the energy limitations of the satellite, etc. The computing resources of LEO satellites are limited and the processing capabilities are relatively low, while data detection models usually require a large amount of computing resources. Therefore, the model deployment strategy must take into account the computing power of the satellite. To adapt to various types of data (such as ADS-B, AIS, and other sensor data), a modular model deployment method is adopted on the on-board platform. Each functional module (such as signal feature extraction, deep learning inference, result aggregation, etc.) is divided into multiple relatively independent model units (MUs) to achieve flexible combination and loading. According to the structure of the data detection model, the model is divided into several layers or several segments for collaborative computing on different satellites. If the layered method is adopted, the first few layers can be assigned to satellites with stronger computing capabilities, and the last few layers can be transferred to other satellites to complete the overall inference in a pipeline form; if the segmented method is adopted, the model can be split into multiple functional segments to independently complete a part of the inference or preprocessing operations. During the deployment process, it is necessary to comprehensively consider the real-time computing power, available memory, energy limitations of the satellite, and the throughput of the inter-satellite / space-ground links. When splitting the model, strive to balance between "too fine segmentation granularity resulting in excessive communication overhead" and "too coarse segmentation granularity resulting in excessive single-satellite computing pressure" to ensure the optimal overall processing performance.

[0108] The data transmission module constructs an optimal space-ground link using an adaptive link selection method;

[0109] The data transmission module is used to select the current optimal space-ground communication link according to the satellite orbital position and link status;

[0110] The data transmission module is also used to transmit the filtering results to the ground station processing module according to the current optimal space-ground communication link;

[0111] The link status includes: bandwidth, latency, and bit error rate.

[0112] In one embodiment of the present application, the construction of the optimal space-ground link adopts an adaptive link selection method. For different regions and different time periods, the visibility and link quality between each satellite in the constellation and the ground station are not the same. The data transmission module can select the current optimal space-ground communication link according to the satellite orbital position and link status (such as bandwidth, latency, bit error rate).

[0113] This application provides an embodiment as follows:

[0114] 1. Experimental environment:

[0115] In order to comprehensively evaluate the effectiveness of the dynamic task scheduling strategy (MDL) of the multi-vehicle data on-orbit detection system based on LEO satellite proposed in this invention, a highly simulated experimental environment was constructed to simulate the complex scenes of the real world. The experimental environment consists of a satellite constellation consisting of 72 low earth orbit (LEO) satellites with an orbital altitude of 600 kilometers. Each satellite is equipped with a 4-core ARM processor (main frequency 2.4GHz) and 8GB of memory, with limited on-board computing capabilities. The inter-satellite link adopts laser communication technology, with a maximum transmission bandwidth of 20Gbps, and the end-to-end delay is controlled within 50 milliseconds to ensure the efficiency and real-time performance of data transmission.

[0116] The experimental scenarios cover three typical geographical areas, namely polar regions, open oceans, and high-density vehicle areas. The selection of these scenarios fully considers the complexity of different geographical environments around the world, especially areas that are difficult for ground stations to cover (such as polar regions and open oceans), and areas with dense vehicles and more signal interference (such as high-density vehicle areas). Through the simulation of these scenarios, the performance of the system in different environments can be fully evaluated.

[0117] 2. Contrast Strategy

[0118] In order to quantify the advantages of the present invention more clearly and accurately, two traditional technologies are selected as comparison benchmarks. By comparing with these technologies, the advancement and superiority of the strategy proposed by the system of the present invention are intuitively demonstrated.

[0119] A system with centralized deployment of data detection models (CGD): In this system, each satellite deploys the entire data detection model separately. Although this method avoids the additional overhead of task allocation and aggregation in a distributed system to a certain extent, due to the limited computing power and energy constraints of satellites, complex data detection models may be difficult to run efficiently. When faced with large-scale, high-speed data transmission requirements, centralized deployment may not meet the requirements of real-time and accuracy.

[0120] Simple Filtering System (SSF): This system performs simple rule filtering operations on the satellite side, such as checking the data format to ensure that the received data meets the basic format requirements. However, the system does not have a complete abnormal data detection capability, cannot conduct in-depth analysis of the data content, and has difficulty identifying complex abnormal data patterns. The data after simple filtering will be directly forwarded to the ground station, which means that the ground station still needs to process a large amount of data that may contain invalid or abnormal information, increasing the processing burden of the ground station and failing to give full play to the satellite's potential in data preprocessing.

[0121] III. Strategy Evaluation

[0122] Test the actual effects of three strategies under three different scenarios. The experimental results are as shown in Figure 2 , Figure 3 and Figure 4 . The experimental results show that the MDL dynamic task scheduling strategy proposed in the present invention has at least a 21% increase in the proportion of effective data transmission in the downlink compared with the traditional strategy, and the comprehensive task processing coverage rate reaches more than 90%, fully verifying the effectiveness of the multi-vehicle data on-orbit detection system based on LEO satellites in improving the proportion of effective data transmission of vehicle signals. Compared with the traditional ground detection system and the spaceborne simple filtering system, the MDL strategy of the system of the present invention shows stronger adaptability and higher performance advantages in complex environments, providing a more advanced and reliable solution for the future multi-vehicle signal monitoring field.

[0123] This application also discloses an electronic device. Referring to Figure 6 , Figure 6 is a schematic structural diagram of an electronic device disclosed in an embodiment of this application. The electronic device 500 may include: at least one processor 501, at least one network interface 504, a user interface 503, a memory 505, and at least one communication bus 502.

[0124] Among them, the communication bus 502 is used to realize the connection and communication between these components.

[0125] Among them, the user interface 503 may include a display screen. Optionally, the user interface 503 may further include a standard wired interface and a wireless interface.

[0126] Among them, the network interface 504 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface).

[0127] This application also discloses a computer-readable storage medium. The computer-readable storage medium stores multiple instructions, and the instructions are suitable for being loaded by a processor to execute the above-mentioned multi-vehicle data on-orbit detection system based on LEO satellites.

[0128] The above are only exemplary embodiments of the present disclosure and should not be used to limit the scope of the present disclosure. That is, all equivalent changes and modifications made in accordance with the teachings of the present disclosure still fall within the scope covered by the present disclosure.

[0129] This application aims to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include well-known common general knowledge or conventional technical means in the technical field not recorded in the present disclosure. The description and embodiments are only regarded as exemplary, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A multi-vehicle data on-orbit detection system based on LEO satellites, characterized in that The system includes: a data acquisition and preprocessing module, a task scheduling module, a cluster resource management module, a task processing module, a data transmission module, and a ground station processing module; The data acquisition and preprocessing module, the task scheduling module, the cluster resource management module, the task processing module, the data transmission module, and the ground station processing module are sequentially connected in order; the task scheduling module is connected to the task processing module; The data acquisition and preprocessing module is used to acquire vehicle signals and perform preprocessing to obtain task data; the vehicle signals include: ship signals and aircraft signals; The cluster resource management module is used to acquire satellite node status information; The task scheduling module is used to receive task data for analysis and accept satellite node status information; the task data is the preprocessed vehicle signal; The task scheduling module is also used to allocate tasks to the task processing module for execution according to the satellite node status information, in combination with the task scheduling algorithm MDL based on dynamic scheduling; The task processing module is used to perform abnormal data filtering processing on the received tasks, obtain a filtering result and input it to the data transmission module; The data transmission module is used to transmit the filtering result to the ground station processing module through the constructed optimal space-ground link; The ground station processing module is used to monitor the driving status of the vehicle according to the received filtering result.

2. The multi-vehicle data on-orbit detection system based on LEO satellites according to claim 1, wherein The preprocessing steps of the cluster resource management module include: Let the received set of original vehicle signal data be where represents the m-dimensional feature vector of the i-th type of vehicle, and the preprocessing process is defined as: where Φ(·) is a multi-type vehicle data fusion function, and φ j (·) includes signal denoising, format standardization, and spatio-temporal alignment; the output is intermediate data in a unified format; The preprocessed data is encapsulated as a task unit where τ i is the computational amount of each task in the task data, is the maximum tolerable delay, and w i is the priority weight of each task in the task data.

3. The multi-vehicle data on-orbit detection system based on LEO satellites as claimed in claim 1, wherein, Continuously update the state matrix S(t) through the cluster resource management module. The state matrix S(t) contains the resource status of N satellites at time t: Among them, c i (t) is the remaining computing resources, m i (t) is the memory occupancy rate, e i (t) is the energy reserve; The satellite node status information includes: remaining computing resources, remaining computing resources, and energy reserves; Construct a real-time resource topology map through the cluster resource management module where \(v\) represents the set of all satellite nodes in the satellite cluster; \(\varepsilon\) represents the set of inter-satellite communication links between satellite nodes; The edge weight of the resource topology graph is the inter-satellite communication delay d jk (t), where k represents another satellite that is not equal to j.

4. The multi-vehicle data on-orbit detection system based on LEO satellites according to claim 2, characterized in that, The specific steps of allocating tasks to the task processing module for execution according to the satellite node status information, in combination with the task scheduling algorithm MDL based on dynamic scheduling, include: S1: Perform pre-classification processing on the tasks according to the maximum tolerable delay of the task data, specifically including: If the computational workload τ of the task in the task data i is greater than the maximum computational capacity limit τ of the satellite high , then this task is split into multiple subtasks; According to the maximum tolerable delay of the task Combined with the preset threshold δ i , mark the urgency of the task; Adjust the priority weight ω of the task according to the urgency of the task i ; S2: Sort the tasks according to the maximum tolerable delay and the adjusted priority weights of the tasks, specifically including: according to the maximum tolerable delay T max , sort the tasks in ascending order; according to the priority weight ω of the tasks i sort the tasks in descending order; Predict the total load of the satellite node cluster according to the task scheduling algorithm MDL based on dynamic scheduling and the satellite node status information; S3: Dynamically allocate tasks through the sorted tasks and the total load of the satellite node cluster. The specific steps are as follows: Calculate the computing time and transmission time required for the active node j in the satellite node cluster to process the task, and verify whether it can meet the delay constraint of the task. The delay constraint is determined by the following formula: where d jk is the transmission delay between node j and other active nodes, B k is the bandwidth of the node; τ represents the computing requirement of the task; η j represents the computing efficiency of satellite j; If the sum of the computing delay \(t\) comp and the transmission delay \(t\) trans does not exceed the maximum tolerable delay \(T\) of the task max , then the node is regarded as a feasible node and can carry the task; Determine the optimal node from the feasible nodes based on the multi-objective optimization strategy; the optimization objectives of the multi-objective optimization strategy include: the remaining computing resource c of the node, the remaining energy resource e of the node, the computing requirement τ of the task, and the communication delay d of the task to the ground station j，g ; Allocate the sorted tasks to the optimal node.

5. The on-orbit detection system for multi-vehicle data based on LEO satellites according to claim 1, characterized in that, The task processing module consists of a satellite constellation cluster that distributes and deploys a data analysis model.

6. The multi-vehicle data on-orbit detection system based on LEO satellites according to claim 1, wherein The data transmission module uses an adaptive link selection method to construct an optimal space-ground link; The data transmission module is used to select the current optimal space-ground communication link according to the satellite orbital position and link status; The data transmission module is also used to transmit the filtering result to the ground station processing module according to the current optimal space-ground communication link; The link status includes: bandwidth, delay, and bit error rate.

7. An electronic device, characterized in that, It includes a processor, a memory, a user interface and a network interface. The memory is used for storing instructions. The user interface and the network interface are used for communicating with other devices. The processor is used for executing the instructions stored in the memory so that the electronic device executes the method according to any one of claims 1-6.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions which, when executed by a computer, execute the method according to any one of claims 1-6.

Images