Fragmented transmission generation method and system for low-orbit satellite terminal data of Internet of Things

Abstract

The invention relates to a fragmented transmission generation method and system for Internet of Things low-orbit satellite terminal data, and the method comprises the steps: actively sensing and quantifying the state of a data queue through an Internet of Things terminal, calculating and reporting a transmission priority and a local resource demand degree, converging all terminal information through a low-orbit satellite node, and transmitting the collected terminal information to the Internet of Things terminal. An algorithm based on cooperative multi-agent reinforcement learning is utilized to carry out joint optimization scheduling, a resource allocation strategy including fragment granularity guidance is generated and issued, and the Internet of Things terminal adaptively calculates the size of an optimal fragment according to the resource allocation strategy, cuts data and packages the data into data fragment units which are provided with priority labels and can be independently transmitted, and sends the data fragment units to the Internet of Things terminal. And finally, sending is completed on the precise resources allocated by the satellite. According to the method and the device, cross-layer accurate matching of data generation and resource scheduling is realized, the system throughput is remarkably improved, the service time delay is reduced, and fairness among massive terminals is guaranteed.

Description

Technical Field

[0001] This application relates to the field of data transmission technology, and in particular to a method and system for generating fragmented transmission data from low-orbit satellite terminals for the Internet of Things. Background Technology

[0002] In recent years, with the rapid evolution of integrated air-space-ground information networks, low Earth orbit (LEO) satellite constellations, with their advantages of wide-area coverage, low transmission latency, and flexible deployment, have become the core infrastructure for achieving seamless global Internet of Things (IoT) access and data backhaul. Hundreds of millions of IoT terminals are distributed across land, sea, and air, generating a large volume of traffic characterized by bursts, intermittency, and small data packets. However, LEO satellite network resources are extremely limited, and their high-speed movement leads to rapidly changing channel conditions.

[0003] Currently, existing technologies primarily focus on satellite-side network resource management, concentrating on utilizing advanced beam hopping and beamforming technologies, combined with artificial intelligence algorithms, to achieve real-time dynamic scheduling of beam shape, pointing, power, and bandwidth. For example, through multi-agent collaborative learning, optimal beam illumination patterns and resource allocation schemes are generated based on the real-time service demand distribution of ground users (or user clusters), aiming to maximize system throughput, ensure fairness, and reduce queuing latency.

[0004] Regarding the aforementioned technologies, the optimization decisions made by the satellite-side scheduling algorithm are not conducive to timely feedback on the data encapsulation process at the terminal side; conversely, the data characteristics generated at the terminal side can only be reported in a coarse-grained manner through limited signaling. This fragmentation of cross-layer information can easily lead to a mismatch between resource supply and data demand.

[0005] Based on this, this application provides a method and system for generating fragmented transmission data from low-orbit satellite terminals for the Internet of Things. Summary of the Invention

[0006] In order to improve the optimization decisions made by the satellite-side scheduling algorithm, it is not convenient to have a timely feedback effect on the data encapsulation process on the terminal side; conversely, the data characteristics generated by the terminal side can only be reported in a coarse-grained manner through limited signaling. This fragmentation of cross-layer information can easily cause a mismatch between resource supply and data demand. This application provides a fragmented transmission generation method and system for IoT low-orbit satellite terminal data.

[0007] Firstly, this application provides a method for generating fragmented transmission data from a low-Earth orbit satellite terminal for the Internet of Things, employing the following technical solution: including: The local data queue is monitored by an IoT terminal to obtain the queuing delay and data volume of the data to be transmitted; based on the queuing delay and data volume, the transmission priority and local resource demand of the data to be transmitted are calculated; and the terminal status information containing the transmission priority and the local resource demand is sent to the corresponding low-orbit satellite node. The system receives terminal status information from multiple IoT terminals within its coverage area via low-orbit satellite nodes. Based on the received terminal status information, a resource scheduling algorithm is executed to jointly optimize system throughput, transmission latency, and user fairness. A resource allocation strategy is generated for each IoT terminal, which includes at least the equivalent bandwidth and fragmentation granularity parameters allocated to the corresponding terminal. The resource allocation strategy is then distributed to the corresponding IoT terminal. The resource allocation strategy is received from low-Earth orbit satellite nodes via an IoT terminal. Based on the fragment granularity parameter and equivalent bandwidth in the resource allocation strategy, the data to be transmitted is cut into one or more independent data fragments. Each data fragment is encapsulated with an independent metadata header to form an independently transmittable fragment unit. The metadata header contains at least a fragment priority tag derived from the transmission priority. Based on the time and frequency resources allocated to it by the IoT terminal on the low-Earth orbit satellite node, the fragment unit that can be transmitted independently is sent; the fragment unit is received by the low-Earth orbit satellite node, and a differentiated forwarding, caching or packet loss strategy is executed according to the fragment priority tag in the metadata header.

[0008] Preferably, the step of monitoring the local data queue through the IoT terminal to obtain the queuing delay and data volume of the data to be transmitted; and calculating the transmission priority and local resource demand of the data to be transmitted based on the queuing delay and data volume, includes: By monitoring its local first-in-first-out data queue through IoT terminals, the queuing time of each data packet to be transmitted in the queue can be obtained. Based on the queuing time of all pending data packets, the overall queuing delay of the pending data packets is calculated. Based on the overall queuing delay, the transmission priority of the data to be transmitted is calculated through a preset priority mapping function. The priority mapping function is configured such that the transmission priority is positively correlated with the overall queuing delay, and its growth rate decreases as the overall queuing delay increases. Based on the IoT terminal, it obtains its own geographical location information, performs geographical clustering with neighboring terminals within the signal coverage area, obtains the geographical coverage area of ​​the cluster, and obtains the highest transmission priority among all terminals in the corresponding cluster. The local resource demand is calculated based on the product of geographical coverage area and the highest transmission priority. The calculated transmission priority and the local resource demand are encapsulated into structured terminal status information through the Internet of Things terminal. The terminal status information is sent to the low-Earth orbit satellite node that provides coverage services via the uplink channel.

[0009] Preferably, the step involves receiving terminal status information from multiple IoT terminals within the coverage area via a low-orbit satellite node; based on all received terminal status information, executing a resource scheduling algorithm to generate a resource allocation strategy for each IoT terminal with the goal of jointly optimizing system throughput, transmission latency, and user fairness, including: The system receives terminal status information from various IoT terminals within the coverage area via low-orbit satellite nodes; it then aggregates the terminal status information from all terminals to form a set of regional terminal statuses at the current moment. Based on the set of terminal states in the region, a resource scheduling algorithm is executed. The resource scheduling algorithm is configured to make joint decisions on beam shape, transmit power and bandwidth through at least one agent. Its decision objective is to maximize a comprehensive utility function weighted by the total system throughput, the negative value of total queuing delay and the fairness index among users. Based on the decision results of the resource scheduling algorithm, a resource allocation strategy containing specific parameters is generated for each IoT terminal through low-orbit satellite nodes. The resource allocation strategy includes at least: a scheduling time slot indicating transmission opportunities, a beam power allocation ratio indicating the degree of power concentration, and a normalized bandwidth allocation scheme indicating the share of spectrum resources. The resource allocation strategy is sent to the corresponding IoT terminal via the downlink control channel.

[0010] Preferably, the step of executing the resource scheduling algorithm based on the regional terminal status set includes: Each communication beam or user cluster within the coverage area is mapped to an independent sub-agent, a policy network with shared parameters is configured for all sub-agents, and a macro-agent is set up to be responsible for centralized training and coordination. Each sub-agent uses the normalized resource demand of the beam or user cluster it manages as its own state input; each sub-agent inputs its own state into the shared policy network, calculates and outputs the preliminary resource allocation action for the corresponding beam or user cluster, the preliminary resource allocation action includes at least the power allocation ratio and the bandwidth allocation ratio. Each sub-agent reports its preliminary resource allocation actions to the macro-agent. Based on preset resource constraints and global optimization objectives, the macro-agent performs conflict detection and coordination arbitration on the preliminary resource allocation actions of each sub-agent, and generates an executable joint resource allocation scheme. The joint resource allocation scheme is executed through low-Earth orbit satellite nodes, i.e., beam illumination and data transmission are performed according to the determined power and bandwidth allocation results; each sub-agent observes the changes in the environmental state after the scheme is executed, and calculates its immediate reward according to the joint optimization objective; each sub-agent constructs the process of this decision and interaction into empirical data including state, action, reward and the state at the next moment. Based on the macro-agent, the experience data generated by all sub-agents are classified according to the number of beams from which they originate and stored in the corresponding distributed experience replay pools. Batch experience data is periodically sampled from each experience replay pool to update the parameters of the shared policy network and the value evaluation network within the macro-agent, so as to achieve continuous optimization learning.

[0011] Preferably, the step of receiving the resource allocation strategy from the low-Earth orbit satellite node via the IoT terminal, and dividing the data to be transmitted into one or more independent data fragments according to the fragment granularity parameter and equivalent bandwidth in the resource allocation strategy; encapsulating each data fragment with an independent metadata header to form an independently transmittable fragment unit, including: The resource allocation strategy from the low-Earth orbit satellite node is sent to the Internet of Things terminal through the downlink control channel. The resource allocation strategy is parsed to extract the allocated equivalent bandwidth, fragment granularity parameters, and modulation and coding scheme parameters required for transmission. Based on the parsed equivalent bandwidth and fragment granularity parameters, the IoT terminal runs an adaptive fragmentation algorithm, combined with the total amount of data to be transmitted and the business lifetime requirements, to calculate an optimized fragment size under the premise of meeting the business time limit, so that a single fragment can be efficiently transmitted within one allocated transmission opportunity and the transmission overhead is minimized. Based on the optimized fragment size calculated by the IoT terminal, the data to be transmitted is logically cut below the application layer and above the transmission layer to generate one or more data blocks, each of which constitutes a data fragment body. An independent metadata header is generated for each data fragment body, and the metadata header is encapsulated with the data fragment body to form a complete fragment unit that can be transmitted independently. The metadata header includes at least: a globally unique identifier for uniquely identifying the fragment in the network, a fragment priority label inherited from the transmission priority, a sequence number and fragment information for identifying the order of the fragment in the original data and the total number of fragments, and a lifetime indicating the effective transmission duration of the fragment. The control IoT terminal will temporarily store the encapsulated, independently transmittable fragment units in its local transmission buffer, waiting for specific time and frequency resources allocated by the low-Earth orbit satellite node for uplink transmission.

[0012] Preferably, the IoT terminal, based on the parsed equivalent bandwidth and fragment granularity parameters, combined with the total data volume of its data to be transmitted and the service lifetime requirements, runs an adaptive fragmentation algorithm to calculate an optimized fragment size while meeting the service time limit. This ensures that a single fragment can be efficiently transmitted within one allocated transmission opportunity, minimizing transmission overhead. This includes: The equivalent bandwidth and fragmentation granularity parameters obtained from the resource allocation strategy are obtained through the Internet of Things terminal; the total data volume of the data to be transmitted and the service lifetime requirement are obtained; and the estimated bit error rate or historical transmission success rate statistics of the current channel are obtained. Based on the input parameters, a fragment size calculation model is constructed with the total transmission latency and the proportion of protocol overhead as joint optimization objectives; where the total transmission latency consists of transmission time and protocol processing time, and the proportion of protocol overhead is determined by the ratio of metadata header length to fragment size. Run the fragment size calculation model and solve it under the constraint of meeting the service lifetime requirements; the goal of the solution is to minimize the total transmission latency while keeping the protocol overhead ratio below a preset threshold; the solution output is the optimized fragment size. The IoT terminal compares the optimized fragment size obtained with the fragment granularity parameter in the resource allocation strategy. If the difference between the two is within the preset tolerance range, the optimized fragment size is adopted; if it exceeds the tolerance range, the fragment granularity parameter is used as the final fragment size for subsequent cutting operations.

[0013] Preferably, the IoT terminal logically segments the data to be transmitted below the application layer and above the transport layer according to the calculated optimized fragment size, generating one or more data blocks. Each data block constitutes a data fragment body, including: The raw data packet queue is formed by acquiring the data to be transmitted through the Internet of Things terminal. Each raw data packet contains data content, data packet identifier and arrival timestamp. The raw data packet queue follows the first-in-first-out transmission principle. By checking the lifetime status of each data packet in the original data packet queue through the Internet of Things terminal, data packets that exceed the preset lifetime threshold are removed from the original data packet queue. Based on the optimized fragment size, a logical segmentation unit is formed. Following the order of the original data packet queue, the data content of one or more complete original data packets is combined and filled into the logical segmentation unit to generate a data fragment body. The combination and filling follows these rules: a single data fragment body should carry an integer number of original data packets as much as possible; if the size of an original data packet exceeds the remaining capacity of the logical segmentation unit, the corresponding original data packet is not segmented but is placed as a whole into the next logical segmentation unit. For each generated data fragment body, record the list of original data packet identifiers contained therein and the sequence number of the data fragment body.

[0014] Secondly, this application discloses a fragmented transmission generation device for IoT low-orbit satellite terminal data, which adopts the following technical solution, including: The data acquisition module is used to monitor the local data queue through the Internet of Things terminal, obtain the queuing delay and data volume of the data to be transmitted; calculate the transmission priority and local resource demand of the data to be transmitted based on the queuing delay and data volume; and send the terminal status information containing the transmission priority and the local resource demand to the corresponding low-orbit satellite node. The resource allocation module is used to receive terminal status information sent by multiple IoT terminals within the coverage area through low-orbit satellite nodes; based on all the received terminal status information, it executes a resource scheduling algorithm to generate a resource allocation strategy for each IoT terminal with the goal of jointly optimizing system throughput, transmission latency and user fairness. The resource allocation strategy includes at least the equivalent bandwidth and fragmentation granularity parameters allocated to the corresponding terminal, and the resource allocation strategy is sent to the corresponding IoT terminal. The data segmentation module is used to receive the resource allocation strategy from the low-Earth orbit satellite node through the Internet of Things terminal, and segment the data to be transmitted into one or more independent data fragments according to the fragment granularity parameter and equivalent bandwidth in the resource allocation strategy; and encapsulate each data fragment with an independent metadata header to form a fragment unit that can be transmitted independently, wherein the metadata header contains at least a fragment priority tag derived from the transmission priority. The data transmission module is used to send independently transmittable fragment units on the time and frequency resources allocated to it by the IoT terminal on the low-Earth orbit satellite node; receive the fragment units through the low-Earth orbit satellite node, and execute differentiated forwarding, caching or packet loss strategies according to the fragment priority tag in the metadata header.

[0015] 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 fragmented transmission generation method for IoT low-orbit satellite terminal data described above.

[0016] 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 method for generating fragmented transmission data of IoT low-orbit satellite terminals.

[0017] In summary, the IoT terminal in this application perceives its local data queue by calculating the queuing delay and data volume of the data to be transmitted, autonomously determining its transmission priority and local resource demand, and reporting this status information to the service system or satellite. The service system or satellite acts as a decision-making center, aggregating the dynamic demands of all terminals within the coverage area, running scheduling algorithms such as reinforcement learning, and generating and issuing a customized resource allocation strategy for each terminal with the goal of optimizing overall system efficiency. This strategy includes specific guidance on the size of fragmented data for the terminal. After receiving the strategy, the terminal does not simply execute fixed fragmentation, but adaptively calculates based on its own business time constraints to obtain the optimal fragment size. Then, it cuts the original data according to this size and encapsulates it into data fragment units with independent priority tags that can be individually identified and processed by the network. Finally, the terminal transmits these fragments on the precise time-frequency resources allocated by the satellite, and the satellite performs differentiated fast forwarding or buffering based on their priorities. This changes the disconnect between data generation and resource scheduling in traditional space-ground IoT. Through uplink and downlink bidirectional information interaction and policy coordination, it achieves precise matching and dynamic adaptation of network resources and terminal data characteristics, significantly improving the transmission success rate of massive IoT services in highly dynamic satellite channels, reducing end-to-end latency, and optimizing the overall spectrum and power resource utilization efficiency. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a method for generating fragmented transmission data from low-orbit satellite terminals used in the Internet of Things (IoT).

[0019] Figure 2 This is a structural block diagram of a device for generating fragmented data transmission from low-orbit satellite terminals used in the Internet of Things (IoT). Detailed Implementation

[0020] The following combination Figures 1-2 This application will be described in further detail.

[0021] With the large-scale deployment of low-Earth orbit satellite constellations and the deep integration of space-ground networks, building a communication system capable of supporting massive global real-time access and efficient data transmission for IoT terminals has become a core vision for future network evolution. However, under this vision, a fundamental contradiction is becoming increasingly prominent: in resource-constrained and topologically dynamic satellite channels, how can precious air interface resources be dynamically and efficiently matched with fragmented data streams generated by ground terminals that are highly bursty, heterogeneous, and time-sensitive? Currently, mainstream technological approaches seek breakthroughs from two dimensions: intelligent resource scheduling on the network side and data transmission protocols on the terminal side. While these studies have made significant progress in their respective fields, a key issue remains unresolved: a lack of deep, closed-loop coordination between network scheduling strategies and data generation mechanisms. Satellites cannot accurately perceive the real-time value and urgency of terminal data, and their optimized beam and resource allocation may be misguided; terminals also cannot predict the network's immediate load and channel status, and their generated data packets or fragments may be untimely. This fragmentation at the strategy level leads to bottlenecks in overall system performance, making it difficult to simultaneously achieve efficient resource utilization, guaranteed service latency, and fairness in massive access.

[0022] Based on this, this application proposes a fragmented transmission generation method and system for IoT low-orbit satellite terminal data, and designs a cross-layer interaction mechanism and collaborative algorithm to enable the intelligent resource scheduling of satellites and the intelligent data generation of terminals to form a feedback, thereby matching dynamic resource supply and dynamic data demand in the time domain, frequency domain, spatial domain and power domain.

[0023] Reference Figure 1 The embodiments of this application include at least steps S10 to S40.

[0024] S10 monitors the local data queue through the IoT terminal to obtain the queuing delay and data volume of the data to be transmitted; based on the queuing delay and data volume, it calculates the transmission priority and local resource demand of the data to be transmitted; and sends the terminal status information containing the transmission priority and local resource demand to the corresponding low-orbit satellite node.

[0025] S20: Receive terminal status information from multiple IoT terminals within the coverage area via low-orbit satellite nodes; Based on all received terminal status information, execute a resource scheduling algorithm to jointly optimize system throughput, transmission latency, and user fairness, generating a resource allocation strategy for each IoT terminal. The resource allocation strategy includes at least the equivalent bandwidth and fragmentation granularity parameters allocated to the corresponding terminal, and then distributes the resource allocation strategy to the corresponding IoT terminal.

[0026] S30 receives the resource allocation strategy from the low-Earth orbit satellite node through the IoT terminal. Based on the fragment granularity parameter and equivalent bandwidth in the resource allocation strategy, it cuts the data to be transmitted into one or more independent data fragments. Each data fragment is encapsulated with an independent metadata header to form a fragment unit that can be transmitted independently. The metadata header contains at least a fragment priority tag derived from the transmission priority.

[0027] S40, based on the time and frequency resources allocated to it by the IoT terminal on the low-Earth orbit satellite node, sends independently transmittable fragment units; receives fragment units through the low-Earth orbit satellite node, and executes differentiated forwarding, caching or packet loss strategies according to the fragment priority label in the metadata header.

[0028] Specifically, IoT terminals proactively monitor their own data queues, calculate transmission priorities representing data urgency and local resource demand representing resource needs, and report this status information. Low-Earth orbit (LEO) satellite nodes, acting as intelligent hubs, aggregate information from all terminals, run resource scheduling algorithms aimed at overall system efficiency, and generate and distribute resource allocation strategies with specific fragment granularity guidance for each terminal. Based on these strategies and their own business constraints, terminals adaptively segment and encapsulate the data to be transmitted into data fragment units with priority tags and independent routing capabilities. Finally, terminals transmit fragments on the precisely allocated satellite resources, and the satellites process these fragments differently based on their priorities to complete the transmission. The core function is to bridge the information and decision-making gaps between terminal data generation and network resource scheduling. Through real-time uplink and downlink interaction and policy coordination, the satellite's valuable resources can adapt to the immediate characteristics and needs of the terminal's fragmented data, thereby significantly improving transmission reliability and resource utilization efficiency at the system level.

[0029] In some embodiments, step S10 specifically includes the following steps: monitoring the local first-in-first-out data queue of the IoT terminal and obtaining the queuing time of each data packet to be transmitted in the queue; calculating the comprehensive queuing delay of the data packet to be transmitted based on the queuing time of all data packets to be transmitted; calculating the transmission priority of the data packet to be transmitted based on the comprehensive queuing delay using a preset priority mapping function, wherein the priority mapping function is configured such that the transmission priority is positively correlated with the comprehensive queuing delay, and its growth rate decreases as the comprehensive queuing delay increases; obtaining its own geographical location information based on the IoT terminal, and performing geographical clustering with neighboring terminals within the signal coverage area to obtain the geographical coverage area of ​​the cluster, and obtaining the highest transmission priority among all terminals in the corresponding cluster; calculating the local resource demand based on the product of the geographical coverage area and the highest transmission priority; encapsulating the calculated transmission priority and local resource demand into structured terminal status information through the IoT terminal; and sending the terminal status information to the low-orbit satellite node providing coverage service through the uplink channel.

[0030] IoT terminals monitor their local first-in-first-out (FIFO) data queues. Let the terminal identifier be n, and the current time be t. Its queue of data packets to be transmitted can be represented as a vector. The queuing delay of each data packet in the queue constitutes a delay factor vector, and the terminal calculates the urgency of its data transmission, i.e., the transmission priority. This calculation is based on the inner product of the packet queue and the delay factor, and the formula is as follows: ; Here, ⟨> represents the inner product operation of vectors. This represents the queue of data packets to be transmitted by the nth terminal at time t. This represents the corresponding queuing delay factor vector. The larger the amount of data to be transmitted and the longer the queuing time, the higher its transmission priority.

[0031] Terminals obtain geographic location information and form user clusters with neighboring terminals. Assume the terminal belongs to the m-th cluster. What is the local resource requirement of this terminal? The priority is determined by both the geographical coverage area of ​​the cluster and the highest priority within the cluster, calculated using the following formula: ; in, This represents the geographical coverage area of ​​the m-th user cluster at time t. This represents the maximum transmission priority among all terminals within the cluster at time t. It indicates that the larger the coverage area of ​​a cluster and the higher the urgency of the services provided by the included terminals, the stronger the cluster's demand for satellite radio resources.

[0032] Specifically, by monitoring its local first-in-first-out (FIFO) data queue, the terminal calculates the comprehensive queuing delay and maps it to transmission priority via a non-linear function with a decreasing growth rate, thus precisely characterizing the timeliness and urgency of the data. Furthermore, the terminal uses geographic location information for collaborative clustering, combining the geographical coverage area of ​​the cluster with the highest priority within the cluster to calculate the local resource demand level, which simultaneously reflects the density and urgency of the service. Finally, the terminal encapsulates the priority and demand level into structured status information and proactively reports it. This transforms the terminal from a passive data sender into an intelligent node capable of accurately expressing its multi-dimensional needs (including timeliness and resource demand intensity) to the network side, providing a quantitative decision-making basis for subsequent on-demand resource scheduling and collaborative optimization on the satellite side, and laying the data foundation for cross-layer collaboration.

[0033] In some embodiments, step S20 specifically includes the following steps: receiving terminal status information sent by each IoT terminal within the coverage area through a low-Earth orbit satellite node; aggregating the terminal status information from all terminals to form a regional terminal status set at the current moment; executing a resource scheduling algorithm based on the regional terminal status set, wherein the resource scheduling algorithm is configured to make joint decisions on beam shape, transmit power, and bandwidth through at least one intelligent agent, and its decision objective is to maximize a comprehensive utility function weighted by the total system throughput, the negative value of total queuing delay, and the fairness index among users; based on the decision result of the resource scheduling algorithm, generating a resource allocation strategy containing specific parameters for each IoT terminal through a low-Earth orbit satellite node, the resource allocation strategy including at least: a scheduling time slot indicating transmission opportunities, a beam power allocation ratio indicating the degree of power concentration, and a normalized bandwidth allocation scheme indicating the spectrum resource share; and distributing the resource allocation strategy to the corresponding IoT terminal through a downlink control channel.

[0034] In practice, when executing resource scheduling algorithms, low-Earth orbit satellite nodes need to evaluate system performance under different resource allocation actions. The quality of the signal received by users is determined by the signal-to-interference-plus-noise ratio (SINR). When a satellite generates multiple beams, the SINR received by user n under the coverage of beam m is... The formula is as follows: ; in, This indicates the satellite's total transmission power. This represents the proportion of power allocated to beam m at time t (0 ≤ t). ≤1). This represents the equivalent channel gain magnitude (including path loss and antenna gain) from beam m to user n at time t. This represents the noise power spectral density. This represents the total bandwidth of beam m. This represents the proportion of normalized bandwidth allocated to the user within beam m at time t. This indicates the total number of users covered by beam m. This represents the sum of interference from other co-frequency beams i.

[0035] Given a signal-to-interference-plus-noise ratio (SINNR), what is the instantaneous throughput achievable by user n in beam m? The formula is as follows: ; This throughput is the direct basis for optimizing the total system throughput in resource scheduling algorithms.

[0036] Specifically, satellite nodes aggregate status information reported by all IoT terminals within their coverage area, forming a global situational awareness. Subsequently, based on this information set, a resource scheduling algorithm with an embedded multi-agent decision engine is executed. This algorithm aims to maximize a comprehensive utility function that integrates system throughput, negative total latency, and user fairness, performing integrated joint optimization of beamform, transmit power, and spectrum bandwidth. Based on this optimization decision, the satellite customizes a resource allocation strategy for each terminal, including precise time slots, power ratios, and bandwidth shares, and distributes this strategy through the control channel. The core function of this series of steps is to transform the dispersed and heterogeneous terminal demands into a globally optimal system-level resource allocation scheme, transforming the satellite network from a passive transmission pipeline into a scheduling center capable of proactively sensing and accurately deploying communication resources. This provides an executable physical layer transmission guarantee for the fragmented generation of data on the terminal side.

[0037] Furthermore, considering the specific issues of resource allocation, the corresponding processing steps are as follows: Each communication beam or user cluster within the coverage area is mapped to an independent sub-agent; a shared parameter policy network is configured for all sub-agents, and a macro-agent responsible for centralized training and coordination is set up; each sub-agent uses the normalized resource demand of the beam or user cluster it manages as its own state input; each sub-agent inputs its own state to the shared policy network, calculates and outputs the preliminary resource allocation action for the corresponding beam or user cluster, the preliminary resource allocation action including at least power allocation ratio and bandwidth allocation ratio; each sub-agent reports its calculated preliminary resource allocation action to the macro-agent, which, based on preset resource constraints and global optimization objectives, performs the preliminary resource allocation for each sub-agent. Actions undergo conflict detection and coordination arbitration to generate executable joint resource allocation schemes. These schemes are then executed via low-Earth orbit satellite nodes, involving beam illumination and data transmission according to predetermined power and bandwidth allocations. Each sub-agent observes changes in the environmental state after scheme execution and calculates its immediate reward based on the joint optimization objective. Each sub-agent constructs its decision-making and interaction process into empirical data, including state, action, reward, and the state at the next moment. Based on the macro-agent, the empirical data generated by all sub-agents is categorized according to the number of beams from which it originates and stored in corresponding distributed empirical replay pools. Batch empirical data is periodically sampled from each replay pool to update the parameters of the shared policy network and the value evaluation network within the macro-agent, enabling continuous optimization and learning.

[0038] The macro-optimization objective of the resource scheduling algorithm, namely maximizing the overall utility function, is consistent with the cumulative reward objective pursued by each sub-agent (corresponding to a beam or user cluster). The reward obtained by sub-agent m in time slot t... It can be designed as follows: ; in, , , and These are preset weighting coefficients used to balance the contributions of different optimization objectives. This represents the sum of instantaneous throughput for all users i covered by this sub-agent. This represents the set of users covered by beam m. This represents the sum of the queuing delays for these users. It is a term designed to promote fairness, for example, it can be proportional to The negative value of encourages resource demand to converge toward the average. The long-term goal of the agent is to maximize the discounted cumulative reward expectation. , where γ is the discount factor (0 < γ < 1), and [·]E[·] represents the expectation operator. The update direction of the policy network parameters is along the direction that increases this cumulative reward.

[0039] Specifically, the global optimization task is decomposed as follows: each beam or user cluster acts as a sub-agent, making preliminary power and bandwidth allocation proposals based on its local state; a central macro-agent coordinates all proposals, arbitrating and fusing them based on global resource constraints to form a final executable joint resource allocation scheme. After the system executes this scheme, each sub-agent collects feedback and calculates rewards, forming experience data stored in a categorized experience replay pool for continuous updating of shared network parameters. The core function of this scheme is that it reduces the difficulty of solving high-dimensional action spaces through distributed decision-making, while ensuring the fairness and efficiency of global resource allocation through centralized coordination. This enables the satellite resource scheduling system to have the ability to learn online and adaptively optimize, dynamically approximating the globally optimal strategy in complex environments.

[0040] In some embodiments, step S30 specifically includes the following steps: sending the resource allocation strategy from the low-Earth orbit satellite node to the IoT terminal through the downlink control channel; parsing the resource allocation strategy to extract the allocated equivalent bandwidth, fragment granularity parameters, and modulation and coding scheme parameters required for transmission; and running an adaptive fragmentation algorithm by the IoT terminal based on the parsed equivalent bandwidth and fragment granularity parameters, combined with the total amount of data to be transmitted and the service lifetime requirements, to calculate an optimized fragment size under the premise of meeting the service time limit, so that a single fragment can be efficiently transmitted within one allocated transmission opportunity and the transmission overhead is minimized. Based on the optimized fragment size calculated by the IoT terminal, the data to be transmitted is logically segmented below the application layer and above the transport layer to generate one or more data blocks, each data block constituting a data fragment body. An independent metadata header is generated for each data fragment body, and the metadata header is encapsulated with the data fragment body to form a complete fragment unit that can be transmitted independently. The metadata header includes at least: a globally unique identifier for uniquely identifying the fragment in the network, a fragment priority tag inherited from the transmission priority, a sequence number and fragment information for identifying the order of the fragment in the original data and the total number of fragments, and a lifetime indicating the effective transmission duration of the fragment. The control IoT terminal temporarily stores the encapsulated, independently transmittable fragment unit in the local transmission buffer, waiting for specific time and frequency resources allocated by the low-Earth orbit satellite node for uplink transmission.

[0041] Specifically, the terminal first parses the scheduling instructions from the satellite to obtain key parameters such as equivalent bandwidth and fragment granularity. Then, combining its total data volume and service time constraints, it runs a local optimization algorithm to calculate the optimal fragment size that balances transmission efficiency and protocol overhead under current network resource constraints. Next, the terminal logically segments the original data according to this size, generating data fragment bodies. Each fragment is then encapsulated with a header containing complete metadata such as a unique identifier, priority, sequence number, and time-to-live, forming a transmission unit that can be independently identified, routed, and processed by the network. Finally, the encapsulated fragments are temporarily stored in a buffer, awaiting the designated uplink time slot. This series of steps transforms the macro-level resource scheduling strategy on the satellite side into micro-level, executable data encapsulation actions on the terminal side. Through the terminal's local intelligent calculations, it ensures that each generated data fragment is precisely matched in size and attributes to currently available wireless resources, thereby maximizing the success rate of single transmissions, reducing retransmissions, and enabling high-priority data to be quickly identified and prioritized by the network.

[0042] Furthermore, considering the multi-objective optimization decision-making process running locally on the IoT terminal, the corresponding processing steps are as follows: Obtain the equivalent bandwidth and fragment granularity parameters parsed from the resource allocation strategy through the IoT terminal; obtain the total data volume of the data to be transmitted and the service lifetime requirement; obtain the estimated bit error rate or historical transmission success rate statistics of the current channel; based on the input parameters, construct a fragment size calculation model with the total transmission delay and protocol overhead ratio as joint optimization objectives; where the total transmission delay consists of transmission time and protocol processing time, and the protocol overhead ratio is determined by the ratio of metadata header length to fragment size; run the fragment size calculation model and solve it under the constraint of meeting the service lifetime requirement; the solution objective is to minimize the total transmission delay while controlling the protocol overhead ratio below a preset threshold; the solution output is the optimized fragment size; the IoT terminal compares the obtained optimized fragment size with the fragment granularity parameters in the resource allocation strategy. If the difference is within a preset tolerance range, the optimized fragment size is adopted; if it exceeds the tolerance range, the fragment granularity parameters are used as the final fragment size for subsequent segmentation operations.

[0043] In practice, when determining the fragment size, the terminal needs to estimate the time required to transmit one fragment under the allocated resources. Its transmission rate... Estimates can be made based on the allocated bandwidth and signal-to-noise ratio: ; in, This refers to the sub-channel bandwidth. The number of sub-channels allocated to this terminal (from the resource allocation strategy). This represents the estimated signal-to-noise ratio.

[0044] Let the total size of the data to be transmitted be... Business lifecycle is The fragment header cost is H. Then the terminal solves for the optimized fragment size. The optimization problem can be simplified to: under constraints Next, we search for a protocol that minimizes the total transmission time and protocol overhead. Below the threshold This is a constrained nonlinear optimization problem, which can be solved numerically.

[0045] Specifically, the terminal first obtains resource allocation strategy parameters from the satellite, service constraints of its own data to be transmitted, and channel state information as input. Based on these inputs, the terminal constructs a mathematical model that jointly optimizes the total transmission delay (including actual transmission time and protocol processing delay) and the proportion of protocol overhead. The goal is to find the optimal solution that minimizes the total delay and keeps the overhead controllable while meeting the service deadline, i.e., the calculated optimized fragment size. Finally, the terminal performs tolerance verification with the suggested parameters issued by the satellite to ensure that local decisions are consistent with the global scheduling strategy, thereby determining the final fragment size execution standard. This empowers the terminal to perform fine-grained optimization under the guidance of satellite macro-scheduling. Through the mathematical model, theoretical bandwidth, service urgency, and channel reliability are transformed into a specific and optimal data encapsulation size, thereby achieving a balance between transmission efficiency and reliability within a single transmission opportunity and avoiding resource waste or transmission failure caused by fixed or arbitrary fragmentation.

[0046] Furthermore, considering the preprocessing and encapsulation process of the terminal-side data queue, the corresponding processing steps are as follows: Obtain the raw data packet queue formed by the data to be transmitted through the IoT terminal. Each raw data packet contains data content, a data packet identifier, and an arrival timestamp. The raw data packet queue follows a first-in, first-out (FIFO) transmission principle. Check the lifetime status of each data packet in the raw data packet queue through the IoT terminal, and remove data packets exceeding a preset lifetime threshold from the raw data packet queue. Based on the optimized fragment size, use it as a logical segmentation unit. According to the order of the raw data packet queue, combine the data content of one or more complete raw data packets into the logical segmentation unit to generate a data fragment body. The combination and filling follows these rules: a single data fragment body should carry an integer number of raw data packets as much as possible; if the size of a raw data packet exceeds the remaining capacity of the logical segmentation unit, the corresponding raw data packet is not segmented but placed as a whole into the next logical segmentation unit. For each generated data fragment body, record its list of raw data packet identifiers and the sequence number of the data fragment body.

[0047] Specifically, the terminal first maintains a timestamped, first-in-first-out (FIFO) queue of raw data packets and actively removes timed-out data to ensure information freshness. Then, strictly adhering to the calculated optimal fragment size as the logical unit, it sequentially extracts one or more complete raw data packets from the queue to fill and assemble them, generating data fragment bodies. The rule is to prioritize the integrity of the original data packets, only moving them to the next unit if a single raw data packet is too large, thus avoiding additional segmentation overhead. Simultaneously, the system records the list of original packets contained within each fragment body and its own sequence number. This process achieves efficient unification of business logic (data integrity, timeliness) and transmission physical constraints (fixed fragment size) at the data encapsulation source, reducing the complexity and latency of reassembly at the receiving end, and minimizing protocol processing overhead by avoiding secondary segmentation of the original data packets, ensuring efficient and reliable fragmented transmission.

[0048] The implementation principle of the fragmented transmission generation method for IoT low-Earth orbit satellite terminal data in this application embodiment is as follows: The IoT terminal perceives its local data queue. By calculating the queuing delay and data volume of the data to be transmitted, it autonomously determines its transmission priority and local resource demand, and reports this status information to the service system or satellite. The service system or satellite acts as a decision center. It gathers the dynamic demands of all terminals within the coverage area, runs scheduling algorithms such as reinforcement learning, and generates and issues a customized resource allocation strategy for each terminal with the goal of optimizing the overall system efficiency. This strategy includes specific guidance on the size of the terminal's fragmented data. After receiving the strategy, the terminal does not simply execute fixed fragmentation, but adaptively calculates based on its own business time constraints to obtain the optimal fragment size. Then, it cuts the original data according to this size and encapsulates it into data fragment units with independent priority tags that can be individually identified and processed by the network. Finally, the terminal transmits these fragments on the precise time-frequency resources allocated by the satellite, and the satellite performs differentiated fast forwarding or buffering according to their priorities. This changes the disconnect between data generation and resource scheduling in traditional space-ground IoT. Through uplink and downlink bidirectional information interaction and policy coordination, it achieves precise matching and dynamic adaptation of network resources and terminal data characteristics, significantly improving the transmission success rate of massive IoT services in highly dynamic satellite channels, reducing end-to-end latency, and optimizing the overall spectrum and power resource utilization efficiency.

[0049] Figure 1 This is a flowchart illustrating a method for generating fragmented transmission data from a low-Earth orbit satellite terminal in an embodiment of the Internet of Things (IoT). It should be understood that, although... Figure 1The 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.

[0050] Based on the same technical concept, referring to Figure 2 This application also provides a fragmented transmission generation device for IoT low-orbit satellite terminal data, which adopts the following technical solution: The device includes: The data acquisition module is used to monitor the local data queue through the IoT terminal, obtain the queuing delay and data volume of the data to be transmitted; calculate the transmission priority and local resource demand of the data to be transmitted based on the queuing delay and data volume; and send the terminal status information containing the transmission priority and local resource demand to the corresponding low-orbit satellite node. The resource allocation module is used to receive terminal status information sent by multiple IoT terminals within the coverage area through low-orbit satellite nodes; based on all the received terminal status information, it executes a resource scheduling algorithm to generate a resource allocation strategy for each IoT terminal with the goal of jointly optimizing system throughput, transmission latency and user fairness. The resource allocation strategy includes at least the equivalent bandwidth and fragmentation granularity parameters allocated to the corresponding terminal, and the resource allocation strategy is sent to the corresponding IoT terminal. The data segmentation module is used to receive resource allocation strategies from low-Earth orbit satellite nodes through IoT terminals, and to segment the data to be transmitted into one or more independent data fragments according to the fragment granularity parameters and equivalent bandwidth in the resource allocation strategy; and to encapsulate an independent metadata header for each data fragment to form an independently transmittable fragment unit. The metadata header contains at least a fragment priority tag derived from the transmission priority. The data transmission module is used to send independently transmittable fragment units on the time and frequency resources allocated to it by the IoT terminal on the low-Earth orbit satellite node; receive fragment units through the low-Earth orbit satellite node, and execute differentiated forwarding, caching or packet loss strategies according to the fragment priority label in the metadata header.

[0051] In some embodiments, the data acquisition module is specifically used to monitor its local first-in-first-out data queue through the Internet of Things terminal and obtain the queuing time of each data packet to be transmitted in the queue. Based on the queuing time of all pending data packets, the overall queuing delay of the pending data packets is calculated. Based on the overall queuing delay, the transmission priority of the data to be transmitted is calculated through a preset priority mapping function. The priority mapping function is configured such that the transmission priority is positively correlated with the overall queuing delay, and its growth rate decreases as the overall queuing delay increases. Based on the IoT terminal, it obtains its own geographical location information, performs geographical clustering with neighboring terminals within the signal coverage area, obtains the geographical coverage area of ​​the cluster, and obtains the highest transmission priority among all terminals in the corresponding cluster. The local resource demand is calculated based on the product of geographical coverage area and the highest transmission priority. The calculated transmission priority and local resource requirements are encapsulated into structured terminal status information through IoT terminals. The terminal status information is sent to the low-Earth orbit satellite node that provides coverage services via the uplink channel.

[0052] In some embodiments, the resource allocation module is specifically used to receive terminal status information sent by each IoT terminal within the coverage area through a low-orbit satellite node; and to aggregate the terminal status information from all terminals to form a set of regional terminal statuses at the current moment. Based on the set of regional terminal states, a resource scheduling algorithm is executed. The resource scheduling algorithm is configured to make joint decisions on beam shape, transmit power and bandwidth through at least one agent. Its decision objective is to maximize a comprehensive utility function weighted by the total system throughput, the negative value of total queuing delay and the fairness index among users. Based on the decision results of the resource scheduling algorithm, a resource allocation strategy containing specific parameters is generated for each IoT terminal through low-orbit satellite nodes. The resource allocation strategy includes at least: scheduling time slots indicating transmission opportunities, beam power allocation ratios indicating the degree of power concentration, and normalized bandwidth allocation schemes indicating spectrum resource shares. The resource allocation strategy is sent to the corresponding IoT terminals through the downlink control channel.

[0053] In some embodiments, the resource allocation module is specifically used to map each communication beam or user cluster within the coverage area as an independent sub-agent, configure a policy network with shared parameters for all sub-agents, and set up a macro-agent responsible for centralized training and coordination. Each sub-agent uses the normalized resource demand of the beam or user cluster it manages as its own state input; each sub-agent inputs its own state into the shared policy network, calculates and outputs the preliminary resource allocation action for the corresponding beam or user cluster, the preliminary resource allocation action includes at least the power allocation ratio and the bandwidth allocation ratio. Each sub-agent reports its preliminary resource allocation actions to the macro-agent. Based on preset resource constraints and global optimization objectives, the macro-agent performs conflict detection and coordination arbitration on the preliminary resource allocation actions of each sub-agent, and generates an executable joint resource allocation scheme. The joint resource allocation scheme is executed through low-Earth orbit satellite nodes, that is, beam illumination and data transmission are carried out according to the determined power and bandwidth allocation results; each sub-agent observes the changes in the environmental state after the scheme is executed, and calculates its immediate reward according to the joint optimization objective; each sub-agent constructs the process of this decision and interaction into empirical data including state, action, reward and the state at the next moment. Based on the macro-agent, the experience data generated by all sub-agents are classified according to the number of beams from which they originate and stored in the corresponding distributed experience replay pools. Batch experience data is periodically sampled from each experience replay pool to update the parameters of the shared policy network and the value evaluation network within the macro-agent, so as to achieve continuous optimization learning.

[0054] In some embodiments, the data segmentation module is specifically used to send the resource allocation strategy from the low-Earth orbit satellite node to the Internet of Things terminal through the downlink control channel, parse the resource allocation strategy, and extract the allocated equivalent bandwidth, fragment granularity parameters, and modulation and coding scheme parameters required for transmission. Based on the parsed equivalent bandwidth and fragment granularity parameters, the IoT terminal runs an adaptive fragmentation algorithm, combined with the total amount of data to be transmitted and the business lifetime requirements, to calculate an optimized fragment size under the premise of meeting the business time limit, so that a single fragment can be efficiently transmitted within one allocated transmission opportunity and the transmission overhead is minimized. Based on the optimized fragment size calculated by the IoT terminal, the data to be transmitted is logically cut below the application layer and above the transmission layer to generate one or more data blocks, each of which constitutes a data fragment body. An independent metadata header is generated for each data fragment body, and the metadata header is encapsulated with the data fragment body to form a complete fragment unit that can be transmitted independently. The metadata header includes at least: a globally unique identifier for uniquely identifying the fragment in the network, a fragment priority label inherited from the transmission priority, a sequence number and fragment information for identifying the order of the fragment in the original data and the total number of fragments, and a lifetime indicating the effective transmission duration of the fragment. The control IoT terminal will temporarily store the encapsulated, independently transmittable fragment units in its local transmission buffer, waiting for specific time and frequency resources allocated by the low-Earth orbit satellite node for uplink transmission.

[0055] In some embodiments, the data segmentation module is specifically used to obtain the equivalent bandwidth and fragmentation granularity parameters parsed from the resource allocation strategy through the Internet of Things terminal; obtain the total data volume of the data to be transmitted and the service lifetime requirement; and obtain the estimated bit error rate or historical transmission success rate statistics of the current channel. Based on the input parameters, a fragment size calculation model is constructed with the total transmission latency and the proportion of protocol overhead as joint optimization objectives; where the total transmission latency consists of transmission time and protocol processing time, and the proportion of protocol overhead is determined by the ratio of metadata header length to fragment size. Run the fragment size calculation model and solve it under the constraint of meeting the business lifetime requirements; the goal of the solution is to minimize the total transmission latency while keeping the protocol overhead ratio below a preset threshold; the output result is the optimized fragment size. The IoT terminal compares the optimized fragment size obtained with the fragment granularity parameter in the resource allocation strategy. If the difference between the two is within the preset tolerance range, the optimized fragment size is adopted; if it exceeds the tolerance range, the fragment granularity parameter is used as the final fragment size for subsequent cutting operations.

[0056] In some embodiments, the data segmentation module is specifically used to obtain the original data packet queue formed by the data to be transmitted through the Internet of Things terminal. Each original data packet contains data content, data packet identifier and arrival timestamp. The original data packet queue follows the first-in-first-out transmission principle. By checking the lifetime status of each data packet in the original data packet queue through the Internet of Things terminal, data packets that exceed the preset lifetime threshold are removed from the original data packet queue. Based on the optimized fragment size, a logical segmentation unit is formed. Following the order of the original data packet queue, the data content of one or more complete original data packets is combined and filled into the logical segmentation unit to generate a data fragment body. The combination and filling follows these rules: a single data fragment body should carry an integer number of original data packets as much as possible; if the size of an original data packet exceeds the remaining capacity of the logical segmentation unit, the corresponding original data packet is not segmented, but is placed into the next logical segmentation unit as a whole. For each generated data fragment body, record the list of original data packet identifiers contained therein and the sequence number of the data fragment body.

[0057] This application also discloses a control device.

[0058] 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 fragmented transmission data of the IoT low-orbit satellite terminal described above.

[0059] This application also discloses a computer-readable storage medium.

[0060] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executed, such as the fragmented transmission generation method for IoT low-orbit satellite terminal data 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.

[0061] 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 method for generating fragmented transmission data from a low-Earth orbit satellite terminal for the Internet of Things, characterized in that, include: By monitoring the local data queue through IoT terminals, the queuing latency and data volume of data to be transmitted can be obtained. Based on the queuing delay and data volume, calculate the transmission priority and local resource requirement of the data to be transmitted; send the terminal status information containing the transmission priority and the local resource requirement to the corresponding low-orbit satellite node. Receive terminal status information sent by multiple IoT terminals within the coverage area through low-orbit satellite nodes; Based on all received terminal status information, a resource scheduling algorithm is executed to jointly optimize system throughput, transmission latency, and user fairness. A resource allocation strategy is generated for each IoT terminal. The resource allocation strategy includes at least the equivalent bandwidth and fragmentation granularity parameters allocated to the corresponding terminal. The resource allocation strategy is then sent to the corresponding IoT terminal. The resource allocation strategy received from the low-Earth orbit satellite node is obtained through the Internet of Things terminal. Based on the fragment granularity parameter and equivalent bandwidth in the resource allocation strategy, the data to be transmitted is cut into one or more independent data fragments. Each data fragment is encapsulated with an independent metadata header to form a fragment unit that can be transmitted independently. The metadata header contains at least a fragment priority tag derived from the transmission priority. Based on the time and frequency resources allocated to it by the IoT terminal on the low-Earth orbit satellite node, the fragment unit that can be transmitted independently is sent; the fragment unit is received by the low-Earth orbit satellite node, and a differentiated forwarding, caching or packet loss strategy is executed according to the fragment priority tag in the metadata header.

2. The method for generating fragmented transmission data from a low-orbit satellite terminal for the Internet of Things according to claim 1, characterized in that, The method involves monitoring the local data queue via an IoT terminal to obtain the queuing delay and data volume of the data to be transmitted. Based on the queuing delay and data volume, the transmission priority and local resource requirement of the data to be transmitted are calculated, including: By monitoring its local first-in-first-out data queue through IoT terminals, the queuing time of each data packet to be transmitted in the queue can be obtained. Based on the queuing time of all pending data packets, the overall queuing delay of the pending data packets is calculated. Based on the overall queuing delay, the transmission priority of the data to be transmitted is calculated through a preset priority mapping function. The priority mapping function is configured such that the transmission priority is positively correlated with the overall queuing delay, and its growth rate decreases as the overall queuing delay increases. Based on the IoT terminal, it obtains its own geographical location information, performs geographical clustering with neighboring terminals within the signal coverage area, obtains the geographical coverage area of ​​the cluster, and obtains the highest transmission priority among all terminals in the corresponding cluster. The local resource demand is calculated based on the product of geographical coverage area and the highest transmission priority. The calculated transmission priority and the local resource demand are encapsulated into structured terminal status information through the Internet of Things terminal. The terminal status information is sent to the low-Earth orbit satellite node that provides coverage services via the uplink channel.

3. The method for generating fragmented transmission data of IoT low-orbit satellite terminals according to claim 1, characterized in that, The process involves receiving terminal status information from multiple IoT terminals within the coverage area via low-Earth orbit satellite nodes; based on the received terminal status information, executing a resource scheduling algorithm to generate a resource allocation strategy for each IoT terminal, with the goal of jointly optimizing system throughput, transmission latency, and user fairness, including: The system receives terminal status information from various IoT terminals within the coverage area via low-orbit satellite nodes; it then aggregates the terminal status information from all terminals to form a set of regional terminal statuses at the current moment. Based on the set of terminal states in the region, a resource scheduling algorithm is executed. The resource scheduling algorithm is configured to make joint decisions on beam shape, transmit power and bandwidth through at least one agent. Its decision objective is to maximize a comprehensive utility function weighted by the total system throughput, the negative value of total queuing delay and the fairness index among users. Based on the decision results of the resource scheduling algorithm, a resource allocation strategy containing specific parameters is generated for each IoT terminal through low-orbit satellite nodes. The resource allocation strategy includes at least: a scheduling time slot indicating transmission opportunities, a beam power allocation ratio indicating the degree of power concentration, and a normalized bandwidth allocation scheme indicating the share of spectrum resources. The resource allocation strategy is sent to the corresponding IoT terminal via the downlink control channel.

4. The method for generating fragmented transmission data of IoT low-orbit satellite terminals according to claim 3, characterized in that, The step of executing the resource scheduling algorithm based on the regional terminal status set includes: Each communication beam or user cluster within the coverage area is mapped to an independent sub-agent, a policy network with shared parameters is configured for all sub-agents, and a macro-agent is set up to be responsible for centralized training and coordination. Each sub-agent uses the normalized resource demand of the beam or user cluster it manages as its own state input; each sub-agent inputs its own state into the shared policy network, calculates and outputs the preliminary resource allocation action for the corresponding beam or user cluster, the preliminary resource allocation action includes at least the power allocation ratio and the bandwidth allocation ratio. Each sub-agent reports its preliminary resource allocation actions to the macro-agent. Based on preset resource constraints and global optimization objectives, the macro-agent performs conflict detection and coordination arbitration on the preliminary resource allocation actions of each sub-agent, and generates an executable joint resource allocation scheme. The joint resource allocation scheme is executed through low-Earth orbit satellite nodes, i.e., beam illumination and data transmission are performed according to the determined power and bandwidth allocation results; each sub-agent observes the changes in the environmental state after the scheme is executed, and calculates its immediate reward according to the joint optimization objective; each sub-agent constructs the process of this decision and interaction into empirical data including state, action, reward and the state at the next moment. Based on the macro-agent, the experience data generated by all sub-agents are classified according to the number of beams from which they originate and stored in the corresponding distributed experience replay pools. Batch experience data is periodically sampled from each experience replay pool to update the parameters of the shared policy network and the value evaluation network within the macro-agent, so as to achieve continuous optimization learning.

5. The method for generating fragmented transmission data of IoT low-orbit satellite terminals according to claim 1, characterized in that, The method involves receiving the resource allocation strategy from low-Earth orbit satellite nodes via an IoT terminal, and then cutting the data to be transmitted into one or more independent data fragments based on the fragment granularity parameter and equivalent bandwidth in the resource allocation strategy. Each data fragment is encapsulated with an independent metadata header to form a fragment unit that can be transmitted independently, including: The resource allocation strategy from the low-Earth orbit satellite node is sent to the Internet of Things terminal through the downlink control channel. The resource allocation strategy is parsed to extract the allocated equivalent bandwidth, fragment granularity parameters, and modulation and coding scheme parameters required for transmission. Based on the parsed equivalent bandwidth and fragment granularity parameters, the IoT terminal runs an adaptive fragmentation algorithm, combined with the total amount of data to be transmitted and the business lifetime requirements, to calculate an optimized fragment size under the premise of meeting the business time limit, so that a single fragment can be efficiently transmitted within one allocated transmission opportunity and the transmission overhead is minimized. Based on the optimized fragment size calculated by the IoT terminal, the data to be transmitted is logically cut below the application layer and above the transmission layer to generate one or more data blocks, each of which constitutes a data fragment body. An independent metadata header is generated for each data fragment body, and the metadata header is encapsulated with the data fragment body to form a complete fragment unit that can be transmitted independently. The metadata header includes at least: a globally unique identifier for uniquely identifying the fragment in the network, a fragment priority label inherited from the transmission priority, a sequence number and fragment information for identifying the order of the fragment in the original data and the total number of fragments, and a lifetime indicating the effective transmission duration of the fragment. The control IoT terminal will temporarily store the encapsulated, independently transmittable fragment units in its local transmission buffer, waiting for specific time and frequency resources allocated by the low-Earth orbit satellite node for uplink transmission.

6. The method for generating fragmented transmission data of a low-orbit satellite terminal for the Internet of Things according to claim 5, characterized in that, The process involves using an IoT terminal to perform an adaptive fragmentation algorithm based on the parsed equivalent bandwidth and fragment granularity parameters, combined with the total data volume to be transmitted and the service lifetime requirements. This algorithm calculates an optimized fragment size while meeting service time constraints, ensuring that a single fragment can be efficiently transmitted within a single allocated transmission opportunity and minimizing transmission overhead. This includes: The equivalent bandwidth and fragmentation granularity parameters obtained from the resource allocation strategy are obtained through the Internet of Things terminal; the total data volume of the data to be transmitted and the service lifetime requirement are obtained; and the estimated bit error rate or historical transmission success rate statistics of the current channel are obtained. Based on the input parameters, a fragment size calculation model is constructed with the total transmission latency and the proportion of protocol overhead as joint optimization objectives; where the total transmission latency consists of transmission time and protocol processing time, and the proportion of protocol overhead is determined by the ratio of metadata header length to fragment size. Run the fragment size calculation model and solve it under the constraint of meeting the service lifetime requirements; the goal of the solution is to minimize the total transmission latency while keeping the protocol overhead ratio below a preset threshold; the solution output is the optimized fragment size. The IoT terminal compares the optimized fragment size obtained with the fragment granularity parameter in the resource allocation strategy. If the difference between the two is within the preset tolerance range, the optimized fragment size is adopted; if it exceeds the tolerance range, the fragment granularity parameter is used as the final fragment size for subsequent cutting operations.

7. The method for generating fragmented transmission data of a low-orbit satellite terminal for the Internet of Things according to claim 5, characterized in that, The IoT terminal, based on the calculated optimized fragment size, logically segments the data to be transmitted below the application layer and above the transport layer to generate one or more data blocks. Each data block constitutes a data fragment body, including: The raw data packet queue is formed by acquiring the data to be transmitted through the Internet of Things terminal. Each raw data packet contains data content, data packet identifier and arrival timestamp. The raw data packet queue follows the first-in-first-out transmission principle. By checking the lifetime status of each data packet in the original data packet queue through the Internet of Things terminal, data packets that exceed the preset lifetime threshold are removed from the original data packet queue. Based on the optimized fragment size, a logical segmentation unit is formed. Following the order of the original data packet queue, the data content of one or more complete original data packets is combined and filled into the logical segmentation unit to generate a data fragment body. The combination and filling follows these rules: a single data fragment body should carry an integer number of original data packets as much as possible; if the size of an original data packet exceeds the remaining capacity of the logical segmentation unit, the corresponding original data packet is not segmented but is placed as a whole into the next logical segmentation unit. For each generated data fragment body, record the list of original data packet identifiers contained therein and the sequence number of the data fragment body.

8. A fragmented transmission and generation device for IoT low-orbit satellite terminal data, characterized in that, The device includes: The data acquisition module is used to monitor the local data queue through the Internet of Things terminal, obtain the queuing delay and data volume of the data to be transmitted; calculate the transmission priority and local resource demand of the data to be transmitted based on the queuing delay and data volume; and send the terminal status information containing the transmission priority and the local resource demand to the corresponding low-orbit satellite node. The resource allocation module is used to receive terminal status information sent by multiple IoT terminals within the coverage area through low-orbit satellite nodes; based on all the received terminal status information, it executes a resource scheduling algorithm to generate a resource allocation strategy for each IoT terminal with the goal of jointly optimizing system throughput, transmission latency and user fairness. The resource allocation strategy includes at least the equivalent bandwidth and fragmentation granularity parameters allocated to the corresponding terminal, and the resource allocation strategy is sent to the corresponding IoT terminal. The data segmentation module is used to receive the resource allocation strategy from the low-Earth orbit satellite node through the Internet of Things terminal, and segment the data to be transmitted into one or more independent data fragments according to the fragment granularity parameter and equivalent bandwidth in the resource allocation strategy; and encapsulate each data fragment with an independent metadata header to form a fragment unit that can be transmitted independently, wherein the metadata header contains at least a fragment priority tag derived from the transmission priority. The data transmission module is used to send independently transmittable fragment units on the time and frequency resources allocated to it by the IoT terminal on the low-Earth orbit satellite node; receive the fragment units through the low-Earth orbit satellite node, and execute differentiated forwarding, caching or packet loss strategies according to the fragment priority tag in the metadata header.

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.

Images