Priority scheduling generation method and system for Beidou satellite short message communication

By generating multi-dimensional feature data and using graph theory optimization algorithms, the priority score and transmission sequence of BeiDou short messages are dynamically calculated, solving the problem of single priority judgment in the BeiDou short message communication system and realizing efficient resource utilization and adaptive scheduling.

CN121462069AActive Publication Date: 2026-02-03HUAXIN ZHENGNENG GRP CO LTD

Patent Information

Application Number
CN202610018064.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-01-08
Publication Date
2026-02-03
Estimated Expiration
2046-01-08

AI Technical Summary

Technical Problem

The BeiDou short message communication system lacks a comprehensive consideration of the urgency, timeliness, network load, and channel quality of message content in its scheduling decisions, resulting in a single and static priority judgment that fails to reflect the real-time urgency of services.

Method used

A multi-dimensional feature data generation method is adopted, and a weighted scoring model is used to calculate the dynamic priority score of the message. A conflict graph is constructed and a graph theory algorithm is used to allocate the candidate message set. Combined with a genetic algorithm, the sending sequence and power allocation are optimized to achieve dynamic scheduling.

Benefits of technology

It significantly improves the transmission guarantee capability of critical messages and the efficiency of system resource utilization, and enhances the adaptability in high-concurrency scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121462069A_ABST
    Figure CN121462069A_ABST
Patent Text Reader

Abstract

The invention relates to a priority scheduling generation method and system for Beidou satellite short message communication, and the method comprises the steps: collecting multi-dimensional features, such as message service types, user identities, spatio-temporal information, channel states and the like, dynamically calculating the priority score of each message through a configurable weighting model, and then, carrying out the dynamic calculation of the priority score of each message; the method comprises the following steps of: abstracting a message into a conflict graph vertex, establishing an edge according to a receiving conflict relationship, distributing candidate message sets which are not conflicted with each other for each scheduling time slot by adopting an independent set selection algorithm, and performing joint optimization on a message sending sequence and power distribution by using a genetic algorithm in each time slot by taking a maximized comprehensive utility value as a target, and generating a final scheduling instruction. Through dynamic evaluation, conflict avoidance and resource collaboration, intelligence and self-adaption of a scheduling strategy are realized, and the guarantee capability of a system for key messages and the overall resource utilization efficiency are remarkably improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of satellite communication technology, and in particular to a priority scheduling generation method and system for BeiDou satellite short message communication. Background Technology

[0002] With the widespread application of the BeiDou Navigation Satellite System, its unique short message communication function has become an indispensable communication tool in critical scenarios such as emergency rescue, ocean navigation, border patrol, and field operations. This function allows users to transmit short text messages bidirectionally via satellite links in areas lacking terrestrial network coverage. However, BeiDou short message communication resources are highly limited, manifested in limited channel capacity, strict power constraints, and the system's real-time concurrent processing capabilities.

[0003] Currently, in practical deployments, the BeiDou short message communication system typically employs relatively simple scheduling strategies to handle multi-user access issues. A common existing technology is based on a first-come, first-served queuing rule, supplemented by static priority allocation based on fixed user identity categories. In this mode, all arriving messages enter the queue in chronological order or according to the pre-assigned level of their sending terminal to await service. Another similar technology is a fixed beam or fixed time slot allocation scheme, which allocates fixed communication resources to specific areas or user groups, lacking the ability to dynamically adjust according to real-time service demands.

[0004] Regarding the aforementioned technical methods, the priority judgment dimensions are singular and static, relying solely on arrival time or fixed user categories. They do not comprehensively consider multi-dimensional dynamic information such as the urgency and timeliness of the message content itself, the current network load status, and channel quality, making it difficult for scheduling decisions to truly reflect the real-time urgency of the service. Summary of the Invention

[0005] To address the problem that prioritization decisions are often based on a single, static dimension, relying solely on arrival time or fixed user categories without considering multi-dimensional dynamic information such as the urgency and timeliness of the message content, current network load, and channel quality, which makes it difficult for scheduling decisions to accurately reflect the real-time urgency of services, this application provides a priority scheduling generation method and system for BeiDou satellite short message communication.

[0006] Firstly, this application provides a priority scheduling generation method for BeiDou satellite short message communication, which adopts the following technical solution: including: Based on the set of short messages to be sent and satellite network status information, corresponding multi-dimensional feature data is generated for each message. The multi-dimensional feature data includes at least message service type features, sending user identity features, message spatiotemporal urgency features, and current channel quality features. Based on the multi-dimensional feature data, a dynamic priority score for each message is calculated and output through a configurable weighted scoring model. The weighted scoring model is configured with weight coefficients for different feature dimensions to comprehensively reflect the real-time urgency of message transmission. Based on the dynamic priority score, the message is abstracted as the vertex of the graph, and edges are constructed between conflicting message vertices according to the conflict relationship constraints of the message to generate a conflict graph; based on the conflict graph and the dynamic priority score, the independent set selection algorithm in graph theory is used to allocate a set of non-conflicting candidate messages to each scheduling time slot, and the mapping relationship between the time slot and the candidate message set is determined. Based on the target time slot and its corresponding candidate packet set, with the goal of maximizing the total utility value of packet scheduling within the target time slot, under the conditions of time slot length and power constraints, an optimization algorithm is used to sort and adjust the power of the candidate packet set, solve and output the final transmission sequence and power allocation scheme of the packets within the target time slot; Based on the final transmission sequence and the power allocation scheme, the satellite payload is controlled to perform short message transmission, and the transmission result data is collected as performance feedback.

[0007] Preferably, the step of generating corresponding multi-dimensional feature data for each message based on the set of short messages to be sent and satellite network status information includes: Parse the set of short messages to be sent and extract the original feature tuple of each message. The original feature tuple includes at least the service type tag of the message identifier, the identity identifier of the sending terminal, the message generation timestamp, the message content length, and the message payload data. The satellite network status information is obtained, and the current network status parameters are processed. The network status parameters include at least: the real-time channel gain of the satellite and each receiving terminal, the current load rate of each communication beam, the system background noise power spectral density, and the status of available time and frequency resource blocks. Based on the message generation timestamp and the geographical location information of the sending terminal, combined with the preset scenario policy library, the spatiotemporal context weight of each message is determined. The scenario policy library defines the business urgency level corresponding to different time periods and geographical areas. By integrating the original feature tuples, the network state parameters, and the spatiotemporal context weights, structured multi-dimensional feature data is generated for each packet. Specifically, the packet service type feature is mapped to a numerical priority baseline value based on the service type label; the sending user identity feature is generated based on the sending terminal identity identifier and a pre-stored user service level mapping table to produce a corresponding service level value; the packet spatiotemporal urgency feature calculates the waiting time based on the packet generation timestamp and, combined with the spatiotemporal context weights, generates a dynamically increasing urgency coefficient; and the current channel quality feature calculates the estimated signal-to-interference-plus-noise ratio (SIR) or bit error rate (BER) based on the real-time channel gain and background noise power spectral density.

[0008] Preferably, the step of calculating and outputting the dynamic priority score of each message based on the multi-dimensional feature data and using a configurable weighted scoring model includes: Based on the current system operation mode, external input is loaded from or received from a preset weight strategy library to determine the current weight coefficients corresponding to each feature dimension in the weighted scoring model; wherein, the system operation mode includes at least a normal operation mode, an emergency communication mode, and a key support mode, and the weight coefficients of the same feature dimension are different in different modes; Based on the original values ​​of each dimension in the multi-dimensional feature data, the original values ​​of each dimension are normalized and mapped to a unified numerical range to obtain the normalized scores corresponding to each feature dimension. Among them, the normalization of the message spatiotemporal urgency feature adopts a function that makes the corresponding score increase non-linearly with the increase of message waiting time, so as to ensure that messages close to the preset survival time threshold obtain significantly higher urgency scores. The initial value of the dynamic priority score for each message is calculated by multiplying the current weight coefficient by the normalized score of the corresponding feature dimension and summing the results. The spatiotemporal context weights contained in the multidimensional feature data are used as adjustment factors and applied to the initial value of the dynamic priority score. The initial value of the dynamic priority score is then adjusted by gain to obtain the final dynamic priority score. The dynamic priority scores of all the calculated messages are sorted and used as the priority basis for subsequent scheduling of time slot resources.

[0009] Preferably, based on the dynamic priority score, the message is abstracted as a vertex of a graph, and edges are constructed between conflicting message vertices according to the conflict relationship constraints to generate a conflict graph; based on the conflict graph and the dynamic priority score, an independent set selection algorithm in graph theory is used to allocate a set of non-conflicting candidate messages to each scheduling slot, and the mapping relationship between the slot and the candidate message set is determined, including: Based on the conflict graph, the dynamic priority scores of all packet vertices, the total number of available scheduling slots, and the maximum number of packets that each slot can accommodate, for each scheduling slot to be allocated, the vertex with the highest dynamic priority score is selected from the set of packet vertices that have not yet been allocated to any slot, and used as the starting vertex of the current slot candidate set; among the remaining unallocated vertices, the vertex with the highest dynamic priority score is selected again, and it is determined whether there is an edge between the corresponding vertex and all selected vertices in the current slot candidate set in the conflict graph; if there is no edge, the corresponding vertex is added to the current slot candidate set; if there is an edge, the corresponding vertex is skipped, and the next highest priority vertex is checked, until the current slot candidate set reaches its maximum capacity or all unallocated vertices have been checked; The determined set of candidate messages is bound to the current scheduling time slot, forming and updating the mapping relationship between the time slot and the set of candidate messages, while the selected vertices are marked as allocated; Repeat the vertex allocation steps to allocate a candidate packet set for the next available time slot until all packet vertices have been allocated at least once or all available time slots have been used up, thus completing the pre-allocation of all scheduled time slots.

[0010] Preferably, based on the target time slot and its corresponding candidate packet set, with the objective of maximizing the total utility value of packet scheduling within the target time slot, and under the conditions of time slot length and power constraints, an optimization algorithm is used to sort and adjust the power of the candidate packet set, and solve and output the final transmission sequence and power allocation scheme of the packets within the target time slot, including: Using the candidate message set within the target time slot as the optimization object, a time slot resource optimization model is established. The optimization objective is defined as maximizing the sum of the comprehensive utility values ​​of all scheduled messages within the target time slot. The comprehensive utility value is determined at least by the dynamic priority score of the corresponding message and its expected transmission success rate within the time slot. The constraints include at least the following: the sum of the estimated transmission durations of all selected messages in the candidate message set is not greater than the fixed duration of the target time slot; and the sum of the transmission powers allocated to all selected messages in the candidate message set is not greater than the power budget allocated to the target time slot or the corresponding beam. A preset optimization algorithm is used to solve the resource optimization model within the target time slot. The optimization algorithm performs joint or iterative optimization on the message transmission order and the optional transmission power level. The optimization algorithm is a genetic algorithm, which iteratively searches for the optimal or near-optimal transmission sequence and power allocation combination through selection, crossover, and mutation operations, and incorporates the dynamic priority score or its derivative value into the fitness function to guide the search direction. Based on the solution results of the optimization algorithm, the final transmission sequence and the power allocation scheme for the target time slot are determined. The final transmission sequence specifies the order in which the packets are processed, and the power allocation scheme assigns a transmission power value to each packet in the final transmission sequence.

[0011] Preferably, the step of using a preset optimization algorithm to solve the resource optimization model within the target time slot includes: An initial population containing multiple individuals is randomly generated. Each individual is represented by a chromosome code that represents a candidate scheduling scheme. The structure of the chromosome code is designed to simultaneously represent the transmission order of the messages in the candidate message set and the transmission power level or specific power value assigned to each message. For each individual in the current population, based on the transmission order and power allocation in the corresponding chromosome decoding, combined with the estimated channel conditions, simulate its transmission process in the target time slot, and determine whether it meets the preset time slot length constraints and power constraints. The sum of the overall utility values ​​of the set of messages that the scheme represented by the target individual can successfully send is calculated, and this value is used as the fitness value of the target individual; for individuals that violate the constraints, a penalty term is introduced into the fitness calculation to reduce their fitness. Based on the fitness value of an individual, a roulette wheel selection or tournament selection method is used to select individuals with high fitness from the current population as parent individuals to generate the next generation of the population. A crossover operation is performed on the selected parent individuals with a preset crossover probability, exchanging some gene segments of the chromosomes of the two parent individuals to generate new offspring individuals. A mutation operation is performed on the generated new individuals with a preset mutation probability, randomly changing the specific gene values ​​in their chromosomes that represent the sending order or power allocation, introducing new search possibilities. The new individuals generated by selection, crossover, and mutation are combined to form a new generation of population, and iterative optimization is carried out until the preset maximum number of iterations is reached or the fitness value no longer shows significant improvement in multiple consecutive generations. The individual with the highest fitness value obtained during the iteration process is determined as the solution result of the optimization algorithm.

[0012] Preferably, determining the final transmission sequence and the power allocation scheme for the target time slot based on the solution results of the optimization algorithm includes: The optimization algorithm is a genetic algorithm. The optimal or near-optimal solution chromosome output by the optimization algorithm is obtained. According to the preset encoding rules, the optimal or near-optimal solution chromosome is decoded to determine the candidate message sending order sequence and the power allocation value corresponding to each message. The transmitted sequence and power allocation values ​​obtained from decoding are verified based on time slot length constraints and power constraints; If the estimated total duration of transmitting all messages in the sequential sequence exceeds the fixed duration of the time slot, then the first part of the messages are truncated in sequence so that the total duration of the truncated sequence meets the time slot length constraint. If the total allocated power exceeds the power budget, the power allocation value of each message is scaled proportionally according to a preset ratio until the power constraint is met, and a verified and corrected compliant solution is obtained. Based on the transmission order and power allocation in the compliant solution, combined with the estimated channel conditions, the set of messages that can be successfully transmitted under the corresponding scheme is calculated, and the sum of the comprehensive utility values ​​of all messages in the corresponding message set is calculated. The transmission order in the compliant solution is determined as the final transmission sequence, and the corresponding power allocation value is determined as the power allocation scheme; A complete scheduling instruction, containing the final transmission sequence and the power allocation scheme, is determined in a predetermined data format for execution by the satellite payload.

[0013] Secondly, this application discloses a priority scheduling generation device for BeiDou satellite short message communication, which adopts the following technical solution, including: The feature generation module is used to generate corresponding multi-dimensional feature data for each message based on the set of short messages to be sent and satellite network status information. The multi-dimensional feature data includes at least message service type features, sending user identity features, message spatiotemporal urgency features, and current channel quality features. The dynamic weighting module is used to calculate and output the dynamic priority score of each message based on the multi-dimensional feature data and through a configurable weighted scoring model. The weighted scoring model is configured with weight coefficients for different feature dimensions to comprehensively reflect the real-time urgency of message transmission. The candidate message module is used to abstract messages as vertices of a graph based on the dynamic priority score, and construct edges between conflicting message vertices according to the conflict relationship constraints of the messages to generate a conflict graph; based on the conflict graph and the dynamic priority score, the independent set selection algorithm in graph theory is used to allocate a set of non-conflicting candidate messages to each scheduling time slot, and determine the mapping relationship between the time slot and the candidate message set. The sequence adjustment module is used to maximize the total utility value of message scheduling within the target time slot based on the target time slot and its corresponding candidate message set. Under the condition of satisfying the time slot length and power constraints, it uses an optimization algorithm to sort and adjust the power of the candidate message set, solves and outputs the final transmission sequence and power allocation scheme of the messages within the target time slot. The message sending module is used to control the satellite payload to send short messages based on the final sending sequence and the power allocation scheme, and to collect sending result data as performance feedback.

[0014] Thirdly, this application also provides a control device, the device comprising: It includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed, such as the priority scheduling generation method for BeiDou satellite short message communication described above.

[0015] Fourthly, this application also provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed as described above in the priority scheduling generation method for BeiDou satellite short message communication.

[0016] In summary, this application's system collects multi-dimensional feature data in real time, including message content, user identity, waiting time, channel status, and network load. It dynamically calculates the priority score of each message using a configurable weighted scoring model, accurately quantifying its real-time transmission urgency. Subsequently, messages are abstracted as vertices of a conflict graph, edges are constructed based on constraints such as reception conflicts, and a graph theory algorithm is used to allocate a set of non-conflicting high-priority candidate messages to each scheduling slot, achieving resource conflict reduction and coarse-grained allocation in the spatial dimension. Next, for the candidate set within each slot, aiming to maximize the overall utility value, optimization methods such as genetic algorithms are used to jointly solve for the optimal transmission sequence and power allocation under the conditions of slot length and power constraints. Finally, the system executes scheduling and collects performance feedback to dynamically adjust model weights and algorithm parameters. Through dynamic priority evaluation, conflict-reducing resource planning, and global joint optimization, the system significantly improves its ability to guarantee critical messages, overall resource utilization efficiency, and adaptability to dynamic environments in high-concurrency congestion scenarios. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a priority scheduling generation method for BeiDou satellite short message communication.

[0018] Figure 2 This is a structural block diagram of a priority scheduling generation device for BeiDou satellite short message communication. Detailed Implementation

[0019] The following combination Figure 1 - Figure 2 This application will be described in further detail.

[0020] When faced with sudden, high-concurrency service demands, traditional static or simple priority scheduling methods are insufficient to simultaneously ensure the timeliness of critical information and the overall efficiency of the system under resource-constrained conditions. The shortcomings of existing technologies in multi-dimensional dynamic evaluation, interference coordination avoidance, and global resource optimization constitute the main bottlenecks to improving system performance.

[0021] To address the aforementioned issues, this application proposes a priority scheduling generation method for BeiDou satellite short message communication. The execution entity is a control system. First, it dynamically quantifies the real-time urgency of messages by fusing multi-source information. Then, it decomposes the complex scheduling problem into two levels: slot-level conflict and queue-level utility, solving them sequentially using graph theory and optimization algorithms. Finally, it achieves adaptive adjustment of the strategy through execution feedback.

[0022] Reference Figure 1 The embodiments of this application include at least steps S10 to S50.

[0023] S10: Based on the set of short messages to be sent and the satellite network status information, generate corresponding multi-dimensional feature data for each message. The multi-dimensional feature data includes at least message service type features, sender identity features, message spatiotemporal urgency features, and current channel quality features.

[0024] S20, based on multi-dimensional feature data, calculates and outputs the dynamic priority score of each message through a configurable weighted scoring model. The weighted scoring model is configured with weight coefficients for different feature dimensions to comprehensively reflect the real-time urgency of message transmission.

[0025] S30, based on dynamic priority scores, abstracts messages as vertices of a graph, and constructs edges between conflicting message vertices according to the conflict relationship constraints of the messages to generate a conflict graph; based on the conflict graph and dynamic priority scores, the independent set selection algorithm in graph theory is used to allocate a set of non-conflicting candidate messages to each scheduling time slot, and determines the mapping relationship between the time slot and the candidate message set.

[0026] S40: Based on the target time slot and its corresponding candidate packet set, with the goal of maximizing the total utility value of packet scheduling within the target time slot, under the conditions of time slot length and power constraints, an optimization algorithm is used to sort and adjust the power of the candidate packet set, solve and output the final transmission sequence and power allocation scheme of the packets within the target time slot.

[0027] S50, based on the final transmission sequence and power allocation scheme, controls the satellite payload to perform short message transmission and collects transmission result data as performance feedback.

[0028] Specifically, by collecting dynamic information from multiple dimensions such as message content, user, time, and channel, a feature profile that comprehensively depicts the urgency of business and resource requirements is constructed. Then, a configurable weighted scoring model is used to fuse these features and generate a dynamically changing priority score for each message, thus transforming complex scheduling requirements into quantifiable and comparable numerical criteria. Next, the scheme decomposes the scheduling problem into two levels: First, based on graph theory, messages are abstracted into a conflict graph, and edges are constructed according to constraints. An independent set selection algorithm is used at the time slot level for coarse-grained resource allocation to avoid conflicts, ensuring that simultaneously scheduled messages do not interfere with each other. Then, within each time slot, with the goal of maximizing utility, an optimization algorithm is used to sort the transmission sequence and finely adjust the power of the allocated candidate messages to achieve optimal resource utilization. Finally, the system executes the scheduling scheme and collects feedback data, forming a closed-loop optimization. This continuous process achieves multi-dimensional awareness and automated scheduling from a global perspective, significantly improving the transmission guarantee capability of critical messages, system resource utilization efficiency, and the overall service adaptability in high-concurrency scenarios.

[0029] In some embodiments, step S10 specifically includes the following steps: parsing the set of short messages to be sent, extracting the original feature tuple for each message, wherein the original feature tuple includes at least the service type label of the message identifier, the identity identifier of the sending terminal, the message generation timestamp, the message content length, and the message payload data; obtaining satellite network status information, processing the current network status parameters, wherein the network status parameters include at least: the real-time channel gain of the satellite and each receiving terminal, the current load rate of each communication beam, the system background noise power spectral density, and the status of available time-frequency resource blocks; determining the spatiotemporal context weight of each message based on the message generation timestamp and the geographical location information of the sending terminal, combined with a preset scenario policy library, wherein the scenario... The policy library defines the service urgency levels corresponding to different time periods and geographical regions; it integrates the original feature tuples, network state parameters, and spatiotemporal context weights to generate structured multi-dimensional feature data for each message; among them, the message service type feature is mapped to a numerical priority baseline value based on the service type label; the sending user identity feature is generated based on the sending terminal identity identifier and combined with a pre-stored user service level mapping table to generate the corresponding service level value; the message spatiotemporal urgency feature calculates the waiting time based on the message generation timestamp and generates a dynamically increasing urgency coefficient based on the spatiotemporal context weights; and the current channel quality feature calculates the estimated signal-to-interference-plus-noise ratio or bit error rate value based on the real-time channel gain and background noise power spectral density.

[0030] The calculation of real-time channel gain h integrates factors such as satellite antenna gain, path loss, rain attenuation, and shadow fading, and can be modeled as follows: ,in, The deterministic attenuation factor, determined by antenna gain and path loss, is SF, which is the shadowing fading factor following a log-normal distribution. Based on this channel gain and background noise power spectral density N0, the signal-to-interference-plus-noise ratio (SNR) of the communication link under interference can be predicted. For positioning users, the Cramer-Rao lower bound of their code phase measurement error characterizes their theoretically achievable accuracy; this error variance... With transmission power The signal bandwidth B and interference level are related factors, and the formula is as follows: ; in, This represents the interference power from other communication or location signals. This model provides a basis for subsequent power optimization.

[0031] Specifically, by systematically collecting and fusing multi-source information, a precise and dynamic data foundation is built for subsequent scheduling. First, key metadata is extracted from the original packets, while simultaneously acquiring network channel status in real time. Time and geographic location information are then transformed into context weights through a scenario policy library. Finally, this information is structurally fused to generate feature data encompassing four dimensions: service type, user identity, spatiotemporal urgency, and channel quality. This transforms raw, discrete state information into standardized, quantifiable feature vectors, providing comprehensive, real-time, and standardized input for subsequent dynamic priority calculations. This ensures that scheduling decisions reflect the actual urgency of packets and the instantaneous resource status of the system.

[0032] In some embodiments, step S20 specifically includes the following steps: Based on the current system operating mode, loading or receiving external input from a preset weighting strategy library to determine the current weight coefficients corresponding to each feature dimension in the weighted scoring model; wherein, the system operating mode includes at least a normal operating mode, an emergency communication mode, and a key support mode, and the weight coefficients for the same feature dimension differ under different modes; based on the original values ​​of each dimension in the multi-dimensional feature data, normalizing the original values ​​of each dimension separately, mapping them to a unified numerical range, and obtaining the normalized score corresponding to each feature dimension; wherein, the normalization of the message spatiotemporal urgency feature... The algorithm employs a function that causes the corresponding score to increase non-linearly with the message waiting time, ensuring that messages approaching the preset lifetime threshold receive significantly higher urgency scores. The initial dynamic priority score for each message is calculated by multiplying the current weight coefficient by the normalized score of the corresponding feature dimension and summing the results. The spatiotemporal context weights contained in the multi-dimensional feature data are then applied as adjustment factors to the initial dynamic priority score, resulting in the final dynamic priority score. Finally, the dynamic priority scores of all calculated messages are sorted and used as the priority basis for subsequent scheduling of time slot resources.

[0033] Let the normalized scores of message service type characteristics, sender identity characteristics, message spatiotemporal urgency characteristics, and current channel quality characteristics of message i be respectively... The corresponding weighting coefficient is The initial value of the dynamic priority score. The calculation is as follows: ; right The normalization uses a non-linear function to reflect the urgency of the timeliness, for example: , where λ is the urgency coefficient.

[0034] Furthermore, in conjunction with the message Spatiotemporal context weights The final dynamic priority score is obtained. : ,in, This is a context-dependent moderating factor.

[0035] Specifically, multi-dimensional features are transformed into a unified dynamic priority score, providing a quantitative basis for core scheduling decisions. First, the weight coefficients of each feature dimension are dynamically adjusted according to different system modes such as emergency and normal operation. Then, the original feature values ​​are normalized, with a non-linear function applied to the waiting time to significantly improve the score of ultra-urgent messages. Next, a weighted sum is performed to obtain the base score, and spatiotemporal context weights are introduced for final gain adjustment. The purpose of this process is to construct an adaptive priority quantification model, which not only responds to different macro-level scenarios through weight configuration but also ensures that the scheduling system can accurately identify and prioritize the most urgent and critical messages through non-linear processing and context fusion, thereby precisely mapping complex business requirements into executable scheduling criteria.

[0036] In some embodiments, step S30 specifically includes the following steps: based on the conflict graph, the dynamic priority scores of all packet vertices, the total number of available scheduling slots, and the maximum number of packets that each slot can accommodate, for each scheduling slot to be allocated, select the vertex with the highest dynamic priority score from the set of packet vertices that have not yet been allocated to any slot, and use it as the starting vertex of the current slot candidate set; among the remaining unallocated vertices, continue to select the vertex with the highest dynamic priority score, and determine whether there is an edge between the corresponding vertex and all selected vertices in the current slot candidate set in the conflict graph; if no edge exists, then... Add the corresponding vertex to the candidate set of the current time slot; if an edge exists, skip the corresponding vertex and continue to check the next highest priority vertex until the candidate set of the current time slot reaches its maximum capacity or all unallocated vertices have been checked; bind the determined candidate message set to the current scheduling time slot to form and update the mapping relationship between the time slot and the candidate message set, and mark the selected vertex as allocated; repeat the vertex allocation steps to allocate the candidate message set for the next available time slot until all message vertices have been allocated at least once or all available time slots have been used up, completing the pre-allocation of all scheduling time slots.

[0037] Specifically, dynamic priority sorting is combined with a conflict graph structure, employing a greedy strategy to pre-allocate time slot-level resources. This process begins with the highest-priority message, iteratively selecting a set of non-conflicting messages for each time slot until the time slot capacity is filled or no more messages are available, ultimately forming a complete mapping between time slots and message sets. This transforms abstract priority sorting into a conflict-free physical resource allocation scheme. While strictly adhering to receiver conflict constraints, it ensures that transmission actions within each time slot do not interfere with each other. Simultaneously, by prioritizing the allocation of high-priority messages, it lays the foundation for efficient resource utilization and prioritized transmission of high-value messages in subsequent fine-grained scheduling.

[0038] In some embodiments, step S40 specifically includes the following steps: taking the candidate message set within the target time slot as the object to be optimized, establishing a resource optimization model within the time slot, the optimization objective is defined as maximizing the sum of the comprehensive utility values ​​of all scheduled messages within the target time slot, the comprehensive utility value being determined at least by the dynamic priority score of the corresponding message and its expected transmission success rate within the time slot; the constraints at least include: the sum of the estimated transmission durations of all selected messages in the candidate message set is not greater than the fixed duration of the target time slot, and the sum of the allocated transmit power of all selected messages in the candidate message set is not greater than the power budget allocated to the target time slot or the corresponding beam. The optimization algorithm is used to solve the resource optimization model within the target time slot. The optimization algorithm performs joint or iterative optimization on the message transmission order and the selectable transmission power levels. The optimization algorithm is a genetic algorithm that iteratively searches for the optimal or near-optimal combination of transmission sequence and power allocation through selection, crossover, and mutation operations, and incorporates the dynamic priority score or its derivative value into the fitness function to guide the search direction. Based on the solution results of the optimization algorithm, the final transmission sequence and power allocation scheme for the target time slot are determined. The final transmission sequence clarifies the order in which messages are processed, and the power allocation scheme assigns a transmission power value to each message in the final transmission sequence.

[0039] The time-slot resource optimization model is formally defined as follows, with the optimization objective being: ; in, This is the set of candidate messages for this time slot. The utility value of message j (which can be obtained by mapping its dynamic priority score S(j)). This is an indicator function that takes a value of 1 when message j meets the successful decoding condition, and 0 otherwise. The successful decoding condition is related to the received signal-to-interference-plus-noise ratio (SINR). Depends on the allocated power And channel gain, etc.

[0040] Time slot length constraint: ,in, For the selected subset of sent messages, For the estimated transmission time of message j, The time slot has a fixed length.

[0041] Power constraints: , The transmit power allocated to message j, The total power budget for that time slot or beam.

[0042] Quality of Service Constraint (Optional): For communication messages, a certain bit error rate is required. Below the threshold ,Right now The bit error rate can be approximated as: K These are constants related to the modulation method.

[0043] Specifically, for a set of candidate packets already allocated to the same time slot, under strict constraints of time slot length and power budget, a mathematical model is established with the goal of maximizing overall utility. A genetic algorithm is then used to jointly and intelligently optimize packet transmission order and power allocation. This process first formalizes the scheduling problem as a constrained optimization problem. Then, through iterative search simulating biological evolution, an approximate optimal solution is found that maximizes the likelihood of successful transmission of high-priority packets while maintaining the highest resource utilization. Fine-grained scheduling is then performed at the micro-level. By optimizing sorting and power control, the overall system utility is maximized within limited time and energy resources, thus transforming the coarse-grained resource allocation planned in the preceding steps into an executable transmission scheme.

[0044] Furthermore, the optimization algorithm's solution process specifically includes the following steps: First, an initial population containing multiple individuals is randomly generated. Each individual is represented by a chromosome code, indicating a candidate scheduling scheme. The chromosome code structure is designed to simultaneously represent the transmission order of messages in the candidate message set and the transmission power level or specific power value allocated to each message. For each individual in the current population, based on the transmission order and power allocation in the corresponding chromosome decoding, combined with the estimated channel conditions, its transmission process within the target time slot is simulated to determine whether it meets the preset time slot length and power constraints. Second, the sum of the comprehensive utility values ​​of the message set that the scheme represented by the target individual can successfully transmit is calculated, and this value is used as the fitness value of the target individual. Third, for individuals that violate the constraints, a penalty term is introduced in the fitness calculation to reduce the... The fitness of individuals is determined by a selection process. Based on the fitness value of an individual, a roulette wheel selection or tournament selection method is used to select individuals with high fitness from the current population as parents for the next generation. A crossover operation is performed on the selected parents with a preset crossover probability, exchanging parts of the chromosomes of the two parents to generate new offspring. A mutation operation is performed on the generated new individuals with a preset mutation probability, randomly changing specific gene values ​​in their chromosomes that represent sending order or power allocation, introducing new search possibilities. The new individuals generated by selection, crossover, and mutation are combined to form a new generation population, and iterative optimization is performed until a preset maximum number of iterations is reached or the fitness value no longer shows significant improvement over multiple generations. The individual with the highest fitness value obtained during the iteration process is determined as the solution result of the optimization algorithm.

[0045] The fitness function F is formulated as follows: ; in, This is a subset of messages that were successfully decoded during the simulation transmission process; , This serves as a penalty factor for violations of slot length and power constraints. and These represent the total time and total power consumption of the corresponding solution for this individual, respectively. This function guides the algorithm to search for an efficient solution that satisfies the constraints.

[0046] Specifically, the algorithm starts with a randomly generated initial population of solutions, calculates the fitness (i.e., overall utility) of each solution through simulated transmission, and selects the best solutions for "crossover" and "mutation" to generate a new generation of solutions. This process is iterated in a loop, gradually approaching the optimal solution. Its core function is to find, in an efficient and automated search method, the message transmission sequence and power allocation combination that maximizes multiple constraints such as timeliness and power while maximizing system utility in all possible solution spaces, thus solving complex optimization problems that are difficult to handle manually or with simple rules.

[0047] Furthermore, considering how to transform the theoretical output of the genetic algorithm into executable scheduling instructions, the corresponding processing steps are as follows: The optimization algorithm is a genetic algorithm; the optimal or near-optimal solution chromosome output by the optimization algorithm is obtained; according to the preset encoding rules, the optimal or near-optimal solution chromosome is decoded to determine the representative candidate message transmission sequence and the power allocation value corresponding to each message; the decoded transmission sequence and power allocation value are verified based on the time slot length constraint and power constraint; if the estimated total duration of all messages in the transmission sequence exceeds the fixed time slot duration, the first part of the messages is truncated according to the sequence order, so that the total duration of the truncated sequence is reduced. The system ensures that the time slot length constraint is met. If the total allocated power exceeds the power budget, the power allocation value of each message is scaled proportionally according to a preset ratio until the power constraint is met, resulting in a verified and corrected compliant solution. Based on the transmission order and power allocation in the compliant solution, combined with the estimated channel conditions, the set of messages that can be successfully transmitted under the corresponding scheme is calculated, and the sum of the comprehensive utility values ​​of all messages in the corresponding message set is calculated. The transmission order in the compliant solution is determined as the final transmission sequence, and the corresponding power allocation value is determined as the power allocation scheme. A complete scheduling instruction containing the final transmission sequence and power allocation scheme is determined in a predetermined data format for execution by the satellite payload.

[0048] Specifically, the optimal solution chromosome is decoded, the obtained sequence and power values ​​are verified and corrected for compliance, the estimated performance under this scheme is calculated, and finally the formatted output is converted into payload-executable instructions. This ensures that the efficient solution found by the optimization algorithm strictly meets the system's time and energy constraints, and the decision results are unambiguously converted into control commands for the satellite hardware, thus completing the transition from intelligent computation to physical execution.

[0049] The implementation principle of the priority scheduling generation method for BeiDou satellite short message communication in this application is as follows: In this application, the system collects multi-dimensional feature data such as message content, user identity, waiting time, channel status, and network load in real time; through a configurable weighted scoring model, the priority score of each message is dynamically calculated to accurately quantify its real-time transmission urgency; subsequently, the message is abstracted as the vertex of a conflict graph, edges are constructed based on constraints such as receiving conflicts, and a graph theory algorithm is used to allocate a set of non-conflicting high-priority candidate messages to each scheduling time slot, realizing the reduction of resource conflicts and coarse-grained allocation in the spatial dimension; next, for the candidate set in each time slot, with the goal of maximizing the comprehensive utility value, optimization methods such as genetic algorithms are used to jointly solve the optimal solution for transmission sequence and power allocation under the conditions of time slot length and power constraints; finally, the system executes scheduling and collects performance feedback to dynamically adjust model weights and algorithm parameters, thereby significantly improving the system's ability to guarantee key messages, overall resource utilization efficiency, and adaptability to dynamic environments in high-concurrency congestion scenarios through dynamic priority evaluation, conflict reduction resource planning, and global joint optimization.

[0050] Figure 1 This is a flowchart illustrating a priority scheduling generation method for BeiDou satellite short message communication in one embodiment. It should be understood that, although... Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows; unless explicitly stated otherwise, there is no strict order requirement for the execution of these steps, and they can be executed in other orders; and Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0051] Based on the same technical concept, referring to Figure 2 This application also provides a priority scheduling generation device for BeiDou satellite short message communication, which adopts the following technical solution: the device includes: The feature generation module is used to generate corresponding multi-dimensional feature data for each message based on the set of short messages to be sent and satellite network status information. The multi-dimensional feature data includes at least message service type features, sender identity features, message spatiotemporal urgency features, and current channel quality features. The dynamic weighting module is used to calculate and output the dynamic priority score of each message based on multi-dimensional feature data and through a configurable weighted scoring model. The weighted scoring model is configured with weight coefficients for different feature dimensions to comprehensively reflect the real-time urgency of message transmission. The candidate message module is used to abstract messages as vertices of a graph based on dynamic priority scores, and to construct edges between conflicting message vertices according to the conflict relationship constraints of the messages to generate a conflict graph. Based on the conflict graph and dynamic priority scores, the independent set selection algorithm in graph theory is used to allocate a set of non-conflicting candidate messages to each scheduling time slot, and to determine the mapping relationship between the time slot and the candidate message set. The sequence adjustment module is used to maximize the total utility value of message scheduling within the target time slot based on the target time slot and its corresponding candidate message set. Under the condition of satisfying the time slot length and power constraints, it uses an optimization algorithm to sort and adjust the power of the candidate message set, solves and outputs the final transmission sequence and power allocation scheme of the messages within the target time slot. The message transmission module is used to control the satellite payload to send short messages based on the final transmission sequence and power allocation scheme, and to collect transmission result data as performance feedback.

[0052] In some embodiments, the feature generation module is specifically used to parse the set of short messages to be sent and extract the original feature tuple of each message. The original feature tuple includes at least the service type label of the message identifier, the identity identifier of the sending terminal, the message generation timestamp, the message content length, and the message payload data. The satellite network status information is acquired and the current network status parameters are processed. The network status parameters include at least the real-time channel gain of the satellite and each receiving terminal, the current load rate of each communication beam, the system background noise power spectral density, and the status of available time and frequency resource blocks. Based on the message generation timestamp and the geographical location information of the sending terminal, combined with the preset scenario policy library, the spatiotemporal context weight of each message is determined. The scenario policy library defines the business urgency level corresponding to different time periods and geographical areas. By integrating the original feature tuples, network state parameters, and spatiotemporal context weights, structured multi-dimensional feature data is generated for each packet. Among them, the packet service type feature is mapped to a numerical priority benchmark value based on the service type label; the sending user identity feature is generated based on the sending terminal identity identifier and combined with a pre-stored user service level mapping table to generate a corresponding service level value; the packet spatiotemporal urgency feature is calculated based on the packet generation timestamp to calculate the waiting time and combined with the spatiotemporal context weights to generate a dynamically increasing urgency coefficient; and the current channel quality feature is calculated based on the real-time channel gain and background noise power spectral density to obtain the estimated signal-to-interference-plus-noise ratio or bit error rate value.

[0053] In some embodiments, the dynamic weighting module is specifically used to load or receive external input from a preset weighting strategy library according to the current system operating mode, and determine the current weight coefficients corresponding to each feature dimension in the weighted scoring model; wherein, the system operating mode includes at least a normal operating mode, an emergency communication mode, and a key support mode, and the weight coefficients of the same feature dimension are different in different modes; Based on the original values ​​of each dimension in the multi-dimensional feature data, the original values ​​of each dimension are normalized and mapped to a unified numerical range to obtain the normalized scores corresponding to each feature dimension. Among them, the normalization of the message spatiotemporal urgency feature adopts a function that makes the corresponding score increase non-linearly with the increase of message waiting time, so as to ensure that messages close to the preset survival time threshold obtain significantly higher urgency scores. The initial value of the dynamic priority score for each message is calculated by multiplying the current weight coefficient by the normalized score of the corresponding feature dimension and summing the results. The spatiotemporal context weights contained in the multidimensional feature data are used as adjustment factors and applied to the initial value of the dynamic priority score. The gain of the initial value of the dynamic priority score is adjusted to obtain the final dynamic priority score. The dynamic priority scores of all the calculated messages are sorted and used as the priority basis for subsequent scheduling of time slot resources.

[0054] In some embodiments, the candidate message module is specifically used to, based on the conflict graph, the dynamic priority scores of all message vertices, the total number of available scheduling time slots, and the maximum number of messages that each time slot can accommodate, select the vertex with the highest dynamic priority score from the set of message vertices that have not yet been assigned to any time slot for each scheduling time slot to be allocated, and use it as the starting vertex of the current time slot candidate set; among the remaining unallocated vertices, continue to select the vertex with the highest dynamic priority score, and determine whether there is an edge between the corresponding vertex and all selected vertices in the current time slot candidate set in the conflict graph; if there is no edge, add the corresponding vertex to the current time slot candidate set; if there is an edge, skip the corresponding vertex and continue to check the next highest priority vertex, until the current time slot candidate set reaches the maximum capacity or all unallocated vertices have been checked; The determined set of candidate messages is bound to the current scheduling time slot, forming and updating the mapping relationship between the time slot and the set of candidate messages, while the selected vertices are marked as allocated; Repeat the vertex allocation steps to allocate a candidate packet set for the next available time slot until all packet vertices have been allocated at least once or all available time slots have been used up, thus completing the pre-allocation of all scheduled time slots.

[0055] In some embodiments, the sequence adjustment module is specifically used to establish a resource optimization model within a time slot, taking the candidate message set within the target time slot as the optimization object. The optimization objective is defined as maximizing the sum of the comprehensive utility values ​​of all scheduled messages within the target time slot. The comprehensive utility value is determined at least by the dynamic priority score of the corresponding message and its expected transmission success rate within the time slot. The constraints include at least the following: the sum of the estimated transmission durations of all selected messages in the candidate message set is not greater than the fixed duration of the target time slot; and the sum of the transmission powers allocated to all selected messages in the candidate message set is not greater than the power budget allocated to the target time slot or the corresponding beam. The preset optimization algorithm is used to solve the resource optimization model within the target time slot. The optimization algorithm performs joint or iterative optimization on the message sending order and the optional transmission power level. The optimization algorithm is a genetic algorithm, which iteratively searches for the optimal or near-optimal combination of sending sequence and power allocation through selection, crossover and mutation operations, and incorporates the dynamic priority score or its derivative value into the fitness function to guide the search direction. Based on the solution results of the optimization algorithm, the final transmission sequence and power allocation scheme for the target time slot are determined. The final transmission sequence clarifies the order in which the messages are processed, and the power allocation scheme assigns a transmission power value to each message in the final transmission sequence.

[0056] In some embodiments, the sequence adjustment module is further configured to randomly generate an initial population containing multiple individuals, each individual being represented by a chromosome code representing a candidate scheduling scheme; the structure of the chromosome code is designed to simultaneously represent the transmission order of messages in the candidate message set and the transmission power level or specific power value assigned to each message. For each individual in the current population, based on the transmission order and power allocation in the corresponding chromosome decoding, combined with the estimated channel conditions, simulate its transmission process in the target time slot, and determine whether it meets the preset time slot length constraints and power constraints. The sum of the overall utility values ​​of the set of messages that the scheme represented by the target individual can successfully send is calculated, and this value is used as the fitness value of the target individual; for individuals that violate the constraints, a penalty term is introduced into the fitness calculation to reduce their fitness. Based on the fitness value of an individual, a roulette wheel selection or tournament selection method is used to select individuals with high fitness from the current population as parent individuals to generate the next generation of the population. A crossover operation is performed on the selected parent individuals with a preset crossover probability, exchanging some gene segments of the chromosomes of the two parent individuals to generate new offspring individuals. A mutation operation is performed on the generated new individuals with a preset mutation probability, randomly changing the specific gene values ​​in their chromosomes that represent the sending order or power allocation, introducing new search possibilities. The new individuals generated by selection, crossover, and mutation are combined to form a new generation of population, and iterative optimization is carried out until the preset maximum number of iterations is reached or the fitness value no longer shows significant improvement in multiple consecutive generations. The individual with the highest fitness value obtained during the iteration process is determined as the solution result of the optimization algorithm.

[0057] In some embodiments, the sequence adjustment module is further configured to optimize the algorithm as a genetic algorithm, obtain the optimal or near-optimal solution chromosome output by the optimization algorithm, decode the optimal or near-optimal solution chromosome according to a preset encoding rule, and determine the candidate message transmission order sequence represented by the chromosome and the power allocation value corresponding to each message. The transmitted sequence and power allocation values ​​obtained from decoding are verified based on time slot length constraints and power constraints; If the estimated total duration of transmitting all messages in the sequential sequence exceeds the fixed duration of the time slot, then the first part of the messages are truncated in sequence so that the total duration of the truncated sequence meets the time slot length constraint. If the total allocated power exceeds the power budget, the power allocation value of each message is scaled proportionally according to a preset ratio until the power constraint is met, and a verified and corrected compliant solution is obtained. Based on the transmission order and power allocation in the compliant solution, combined with the estimated channel conditions, the set of messages that can be successfully transmitted under the corresponding scheme is calculated, and the sum of the comprehensive utility values ​​of all messages in the corresponding message set is calculated. The transmission order in the compliant solution is determined as the final transmission sequence, and the corresponding power allocation value is determined as the power allocation scheme; Using a predetermined data format, a complete scheduling instruction containing the final transmission sequence and power allocation scheme is determined for execution by the satellite payload.

[0058] This application also discloses a control device.

[0059] Specifically, the control device includes a memory and a processor. The memory stores a computer program that can be loaded by the processor and executed to generate the priority scheduling method for BeiDou satellite short message communication.

[0060] This application also discloses a computer-readable storage medium.

[0061] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executed, such as the priority scheduling generation method for BeiDou satellite short message communication described above. The computer-readable storage medium includes, for example, various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0062] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A priority scheduling generation method for BeiDou satellite short message communication, characterized in that, include: Based on the set of short messages to be sent and satellite network status information, corresponding multi-dimensional feature data is generated for each message. The multi-dimensional feature data includes at least message service type features, sending user identity features, message spatiotemporal urgency features, and current channel quality features. Based on the multi-dimensional feature data, a dynamic priority score for each message is calculated and output through a configurable weighted scoring model. The weighted scoring model is configured with weight coefficients for different feature dimensions to comprehensively reflect the real-time urgency of message transmission. Based on the dynamic priority score, the message is abstracted as the vertex of the graph, and edges are constructed between conflicting message vertices according to the conflict relationship constraints of the message to generate a conflict graph; based on the conflict graph and the dynamic priority score, the independent set selection algorithm in graph theory is used to allocate a set of non-conflicting candidate messages to each scheduling time slot, and the mapping relationship between the time slot and the candidate message set is determined. Based on the target time slot and its corresponding candidate packet set, with the goal of maximizing the total utility value of packet scheduling within the target time slot, under the conditions of time slot length and power constraints, an optimization algorithm is used to sort and adjust the power of the candidate packet set, solve and output the final transmission sequence and power allocation scheme of the packets within the target time slot; Based on the final transmission sequence and the power allocation scheme, the satellite payload is controlled to perform short message transmission, and the transmission result data is collected as performance feedback.

2. The priority scheduling generation method for BeiDou satellite short message communication according to claim 1, characterized in that, Based on the set of short messages to be sent and satellite network status information, multi-dimensional feature data is generated for each message, including: Parse the set of short messages to be sent and extract the original feature tuple of each message. The original feature tuple includes at least the service type tag of the message identifier, the identity identifier of the sending terminal, the message generation timestamp, the message content length, and the message payload data. The satellite network status information is obtained, and the current network status parameters are processed. The network status parameters include at least: the real-time channel gain of the satellite and each receiving terminal, the current load rate of each communication beam, the system background noise power spectral density, and the status of available time and frequency resource blocks. Based on the message generation timestamp and the geographical location information of the sending terminal, combined with the preset scenario policy library, the spatiotemporal context weight of each message is determined. The scenario policy library defines the business urgency level corresponding to different time periods and geographical areas. By integrating the original feature tuples, the network state parameters, and the spatiotemporal context weights, structured multi-dimensional feature data is generated for each packet. Specifically, the packet service type feature is mapped to a numerical priority baseline value based on the service type label; the sending user identity feature is generated based on the sending terminal identity identifier and a pre-stored user service level mapping table to produce a corresponding service level value; the packet spatiotemporal urgency feature calculates the waiting time based on the packet generation timestamp and, combined with the spatiotemporal context weights, generates a dynamically increasing urgency coefficient; and the current channel quality feature calculates the estimated signal-to-interference-plus-noise ratio (SIR) or bit error rate (BER) based on the real-time channel gain and background noise power spectral density.

3. The priority scheduling generation method for BeiDou satellite short message communication according to claim 2, characterized in that, Based on the multi-dimensional feature data, a configurable weighted scoring model is used to calculate and output the dynamic priority score for each message, including: Based on the current system operation mode, external input is loaded from or received from a preset weight strategy library to determine the current weight coefficients corresponding to each feature dimension in the weighted scoring model; wherein, the system operation mode includes at least a normal operation mode, an emergency communication mode, and a key support mode, and the weight coefficients of the same feature dimension are different in different modes; Based on the original values ​​of each dimension in the multi-dimensional feature data, the original values ​​of each dimension are normalized and mapped to a unified numerical range to obtain the normalized scores corresponding to each feature dimension. Among them, the normalization of the message spatiotemporal urgency feature adopts a function that makes the corresponding score increase non-linearly with the increase of message waiting time, so as to ensure that messages close to the preset survival time threshold obtain significantly higher urgency scores. The initial value of the dynamic priority score for each message is calculated by multiplying the current weight coefficient by the normalized score of the corresponding feature dimension and summing the results. The spatiotemporal context weights contained in the multidimensional feature data are used as adjustment factors and applied to the initial value of the dynamic priority score. The initial value of the dynamic priority score is then adjusted by gain to obtain the final dynamic priority score. The dynamic priority scores of all the calculated messages are sorted and used as the priority basis for subsequent scheduling of time slot resources.

4. The priority scheduling generation method for BeiDou satellite short message communication according to claim 3, characterized in that, Based on the dynamic priority score, packets are abstracted as vertices of a graph, and edges are constructed between conflicting packet vertices according to the conflict relationship constraints to generate a conflict graph. Based on the conflict graph and the dynamic priority score, an independent set selection algorithm in graph theory is used to allocate a set of non-conflicting candidate packets to each scheduling slot, and the mapping relationship between the slot and the candidate packet set is determined, including: Based on the conflict graph, the dynamic priority scores of all packet vertices, the total number of available scheduling slots, and the maximum number of packets that each slot can accommodate, for each scheduling slot to be allocated, the vertex with the highest dynamic priority score is selected from the set of packet vertices that have not yet been allocated to any slot, and used as the starting vertex of the current slot candidate set; among the remaining unallocated vertices, the vertex with the highest dynamic priority score is selected again, and it is determined whether there is an edge between the corresponding vertex and all selected vertices in the current slot candidate set in the conflict graph; if there is no edge, the corresponding vertex is added to the current slot candidate set; if there is an edge, the corresponding vertex is skipped, and the next highest priority vertex is checked, until the current slot candidate set reaches its maximum capacity or all unallocated vertices have been checked; The determined set of candidate messages is bound to the current scheduling time slot, forming and updating the mapping relationship between the time slot and the set of candidate messages, while the selected vertices are marked as allocated; Repeat the vertex allocation steps to allocate a candidate packet set for the next available time slot until all packet vertices have been allocated at least once or all available time slots have been used up, thus completing the pre-allocation of all scheduled time slots.

5. The priority scheduling generation method for BeiDou satellite short message communication according to claim 1, characterized in that, Based on the target time slot and its corresponding candidate packet set, with the objective of maximizing the total utility value of packet scheduling within the target time slot, and under the conditions of time slot length and power constraints, an optimization algorithm is used to sort and adjust the power of the candidate packet set, solve for and output the final transmission sequence and power allocation scheme of the packets within the target time slot, including: Using the candidate message set within the target time slot as the optimization object, a time slot resource optimization model is established. The optimization objective is defined as maximizing the sum of the comprehensive utility values ​​of all scheduled messages within the target time slot. The comprehensive utility value is determined at least by the dynamic priority score of the corresponding message and its expected transmission success rate within the time slot. The constraints include at least the following: the sum of the estimated transmission durations of all selected messages in the candidate message set is not greater than the fixed duration of the target time slot; and the sum of the transmission powers allocated to all selected messages in the candidate message set is not greater than the power budget allocated to the target time slot or the corresponding beam. A preset optimization algorithm is used to solve the resource optimization model within the target time slot. The optimization algorithm performs joint or iterative optimization on the message transmission order and the optional transmission power level. The optimization algorithm is a genetic algorithm, which iteratively searches for the optimal or near-optimal transmission sequence and power allocation combination through selection, crossover, and mutation operations, and incorporates the dynamic priority score or its derivative value into the fitness function to guide the search direction. Based on the solution results of the optimization algorithm, the final transmission sequence and the power allocation scheme for the target time slot are determined. The final transmission sequence specifies the order in which the packets are processed, and the power allocation scheme assigns a transmission power value to each packet in the final transmission sequence.

6. The priority scheduling generation method for BeiDou satellite short message communication according to claim 5, characterized in that, The step of using a preset optimization algorithm to solve the resource optimization model within the target time slot includes: An initial population containing multiple individuals is randomly generated. Each individual is represented by a chromosome code that represents a candidate scheduling scheme. The structure of the chromosome code is designed to simultaneously represent the transmission order of the messages in the candidate message set and the transmission power level or specific power value assigned to each message. For each individual in the current population, based on the transmission order and power allocation in the corresponding chromosome decoding, combined with the estimated channel conditions, simulate its transmission process in the target time slot, and determine whether it meets the preset time slot length constraints and power constraints. The sum of the overall utility values ​​of the set of messages that the scheme represented by the target individual can successfully send is calculated, and this value is used as the fitness value of the target individual; for individuals that violate the constraints, a penalty term is introduced into the fitness calculation to reduce their fitness. Based on the fitness value of an individual, a roulette wheel selection or tournament selection method is used to select individuals with high fitness from the current population as parent individuals to generate the next generation of the population. A crossover operation is performed on the selected parent individuals with a preset crossover probability, exchanging some gene segments of the chromosomes of the two parent individuals to generate new offspring individuals. A mutation operation is performed on the generated new individuals with a preset mutation probability, randomly changing the specific gene values ​​in their chromosomes that represent the sending order or power allocation, introducing new search possibilities. The new individuals generated by selection, crossover, and mutation are combined to form a new generation of population, and iterative optimization is carried out until the preset maximum number of iterations is reached or the fitness value no longer shows significant improvement in multiple consecutive generations. The individual with the highest fitness value obtained during the iteration process is determined as the solution result of the optimization algorithm.

7. The priority scheduling generation method for BeiDou satellite short message communication according to claim 6, characterized in that, The step of determining the final transmission sequence and the power allocation scheme for the target time slot based on the solution results of the optimization algorithm includes: The optimization algorithm is a genetic algorithm. The optimal or near-optimal solution chromosome output by the optimization algorithm is obtained. According to the preset encoding rules, the optimal or near-optimal solution chromosome is decoded to determine the candidate message sending order sequence and the power allocation value corresponding to each message. The transmitted sequence and power allocation values ​​obtained from decoding are verified based on time slot length constraints and power constraints; If the estimated total duration of transmitting all messages in the sequential sequence exceeds the fixed duration of the time slot, then the first part of the messages are truncated in sequence so that the total duration of the truncated sequence meets the time slot length constraint. If the total allocated power exceeds the power budget, the power allocation value of each message is scaled proportionally according to a preset ratio until the power constraint is met, and a verified and corrected compliant solution is obtained. Based on the transmission order and power allocation in the compliant solution, combined with the estimated channel conditions, the set of messages that can be successfully transmitted under the corresponding scheme is calculated, and the sum of the comprehensive utility values ​​of all messages in the corresponding message set is calculated. The transmission order in the compliant solution is determined as the final transmission sequence, and the corresponding power allocation value is determined as the power allocation scheme; A complete scheduling instruction, containing the final transmission sequence and the power allocation scheme, is determined in a predetermined data format for execution by the satellite payload.

8. A priority scheduling generation device for BeiDou satellite short message communication, characterized in that, The device includes: The feature generation module is used to generate corresponding multi-dimensional feature data for each message based on the set of short messages to be sent and satellite network status information. The multi-dimensional feature data includes at least message service type features, sending user identity features, message spatiotemporal urgency features, and current channel quality features. The dynamic weighting module is used to calculate and output the dynamic priority score of each message based on the multi-dimensional feature data and through a configurable weighted scoring model. The weighted scoring model is configured with weight coefficients for different feature dimensions to comprehensively reflect the real-time urgency of message transmission. The candidate message module is used to abstract messages as vertices of a graph based on the dynamic priority score, and construct edges between conflicting message vertices according to the conflict relationship constraints of the messages to generate a conflict graph; based on the conflict graph and the dynamic priority score, the independent set selection algorithm in graph theory is used to allocate a set of non-conflicting candidate messages to each scheduling time slot, and determine the mapping relationship between the time slot and the candidate message set. The sequence adjustment module is used to maximize the total utility value of message scheduling within the target time slot based on the target time slot and its corresponding candidate message set. Under the condition of satisfying the time slot length and power constraints, it uses an optimization algorithm to sort and adjust the power of the candidate message set, solves and outputs the final transmission sequence and power allocation scheme of the messages within the target time slot. The message sending module is used to control the satellite payload to send short messages based on the final sending sequence and the power allocation scheme, and to collect sending result data as performance feedback.

9. A control device, characterized in that, The device includes: A memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer program is stored that can be loaded by a processor and executed as described in any one of claims 1 to 7.

Citation Information

Patent Citations

  • Communication message encryption and compression system based on Beidou third generation

    CN116847335A

  • Data transmission optimization method and system based on satellite channel congestion scene

    CN119449144A

  • Bidirectional short message communication method used in network-free scene

    CN120282107A

  • Satellite short message data transmission method and device

    CN120811455A

  • Further enhancements to ultra-reliable low-latency communication (URLLC) in unlicensed spectrum

    US20230164820A1

Cited By

  • Data processing priority scheduling method for high-speed satellite communication and related equipment

    CN121711007A

  • Half-duplex digital time sequence scheduling method and system based on multi-dimensional signaling fusion

    CN121865412A

  • Resource allocation generation method and system for mass terminal access of low earth orbit satellites of Internet of Things

    CN121940885A

  • Resource allocation generation method and system for massive terminal access of low-Earth orbit satellites for Internet of Things

    CN121940885B

  • Enterprise registration self-service terminal remote identity verification method and system

    CN121967094A