Satellite communication terminal intelligent cache system based on edge computing

Abstract

The application provides a satellite communication terminal intelligent cache system based on edge computing, which is arranged in a satellite communication terminal and comprises a data classifier, a structured hierarchical queue, a link state monitoring and uploading scheduler, a breakpoint resume controller, an edge preprocessor and a secure storage interface. The data classifier is used for receiving uplink data streams from user equipment, analyzing the uplink data streams, and generating structured data packets with metadata headers. The metadata headers contain a globally unique sequence number, a data offset, a data length and a check value. The application sets metadata headers containing a globally unique sequence number, a data offset, a data length and a check value for uplink data, so that the cache data has a structured description, different data packets can be managed in detail, and lost data fragments can be accurately located according to the data offset.

Description

Technical Field

[0001] This invention relates to the field of satellite communication technology, and in particular to an intelligent caching system for satellite communication terminals based on edge computing. Background Technology

[0002] In existing satellite communication networks, data generated by user equipment typically needs to be transmitted back in real time via satellite links.

[0003] For example, Chinese patent CN106603553A discloses a ship-to-shore video snapshot transmission system and method, which encodes and saves image frames captured by the ship according to set parameters, and then transmits the saved images to the shore via satellite communication; when satellite communication is interrupted, the transmission status of the image is saved, and after the communication is reconnected, the saved transmission status is read and the unsent parts of the image are transmitted.

[0004] Chinese patent CN114268357A discloses a method, system, device and application for task offloading based on low-Earth orbit satellite edge computing. It introduces edge computing technology into low-Earth orbit satellite networks and allocates tasks by minimizing latency or minimizing energy depending on whether the task is time-sensitive. The receiving satellite then distributes the task to surrounding satellites for execution.

[0005] Chinese patent CN115587266A discloses an intelligent edge caching method for an integrated air-space-ground internet. In this method, relay nodes are used to cache satellite broadcast content in the integrated air-space-ground internet. Clustering is performed based on the characteristics of the relay nodes, and imitation learning is used to optimize caching decisions to improve the cache hit rate.

[0006] Although the aforementioned existing technologies have proposed solutions for satellite communication breakpoint resume, satellite edge computing, and edge caching, they still have the following significant drawbacks: 1. Insufficient reliability mechanisms: Although existing technologies can save the transmission status of specific files or images when satellite communication is interrupted and continue to transmit the unsent parts after communication is restored, they lack fragmentation acknowledgment (ACK / NACK) and unacknowledged data packet recording mechanisms for the high bit error rate of satellite links. It is difficult to accurately retransmit lost data fragments, which may still cause bandwidth waste.

[0007] 2. Lack of refined management: Existing technologies typically store and transmit data based only on data type, data size, or simple task attributes, without uniquely numbering, offsetting, or verifying uplink data packets from user devices. This makes it difficult to accurately locate the lost position when restoring transmission.

[0008] 3. Low resource utilization: Existing satellite edge computing solutions are mainly used for offloading on-board computing tasks, and existing edge caching solutions are mainly used for caching hot content broadcast by satellite. During periods of link idleness or low tariffs, satellite communication terminals often fail to fully utilize edge computing power to aggregate, compress, or deduplicate cached uplink data, resulting in low transmission efficiency.

[0009] 4. Insufficient security and durability: Cached data is mostly stored locally in plaintext or unverified form, lacking local encrypted storage and power failure protection mechanisms. Furthermore, when the cache is full, a simple discard strategy is usually adopted, without considering the priority, timeliness, and value differences of different data.

[0010] Therefore, it is necessary to build an uplink data "reservoir" with self-healing and self-optimization capabilities on the satellite communication terminal side to improve the reliability and efficiency of uplink data transmission in the satellite communication network, and enhance the security and persistence of cached data. Summary of the Invention

[0011] This invention aims to address the problems of unreliable data transmission, inefficient cache management, low efficiency of non-urgent data transmission, and insufficient security and persistence of cached data in the uplink of existing satellite communication networks. It provides an intelligent caching system for satellite communication terminals based on edge computing. By constructing a structured hierarchical queue and combining it with an offset-based breakpoint resume mechanism, link status awareness, and edge preprocessing, it achieves highly reliable and efficient data uploading.

[0012] To address the aforementioned technical problems, this invention provides an intelligent caching system for satellite communication terminals based on edge computing. The system is installed in the satellite communication terminal and includes a data classifier, a structured hierarchical queue, a link status monitoring and upload scheduler, a breakpoint resume controller, an edge preprocessor, and a secure storage interface.

[0013] A data classifier is used to receive uplink data streams from user equipment, parse the uplink data streams, and generate structured data packets with metadata headers, the metadata headers containing globally unique sequence numbers, data offsets, data lengths, and checksums. A structured hierarchical queue is used to write the structured data packets into a cache queue of corresponding priority according to the quality of service requirements of the structured data packets. The link status monitoring and upload scheduler is used to monitor the link status of the satellite link, calculate the link quality score based on the link status, and determine whether to enter the upload window based on the link quality score. After entering the upload window, the structured data packets in the cache queue are uploaded through the satellite link. The breakpoint resume controller is used to receive the confirmation feedback of the structured data packets. When the satellite link is interrupted or switched, it records the file identifier of the currently transmitted file, the confirmed offset, and the list of unconfirmed data packets. After the satellite link is restored, it retransmits the unconfirmed structured data packets according to the confirmed offset. An edge preprocessor is used to perform data aggregation, data compression, and data deduplication on the non-real-time data in the cache queue using the edge computing power of the satellite communication terminal before uploading. A secure storage interface is used to write the structured data packets and corresponding cached metadata into the physical storage medium of the satellite communication terminal.

[0014] Furthermore, the structured hierarchical queue includes a high-priority queue for storing real-time business data, a medium-priority queue for storing interactive business data, and a low-priority queue for storing background batch business data; the verification value is a CRC verification value or a HASH verification value.

[0015] Furthermore, the link status includes satellite elevation angle E, signal-to-noise ratio SNR, tariff rate C, and estimated remaining transit time T, and the link quality score LQS is calculated according to the following formula: LQS=w1×(E / 90)+w2×(SNR / SNRmax)-w3×(C / Cmax)+w4×(T / Tmax); Where w1=0.3, w2=0.3, w3=0.2, w4=0.2, when the link quality score LQS is greater than the preset threshold LQSth, it is determined to enter the upload window, and the preset threshold LQSth is 0.6.

[0016] Furthermore, the confirmation feedback includes ACK and NACK; when the NACK is received or a confirmation timeout occurs, the breakpoint resume controller determines the lost data fragments based on the recorded confirmed offsets, and controls the link status monitoring and upload scheduler to retransmit the lost data fragments.

[0017] Furthermore, the edge preprocessor is used to aggregate multiple small packets of data in the low-priority queue into a single data packet, compress log data and / or image data in the low-priority queue, and identify and remove redundant data that is repeatedly reported.

[0018] Furthermore, the secure storage interface is used to encrypt the cached data using a terminal key and store the verification value of the cached data before writing the cached data to the eMMC or SSD; the cached data is encrypted using AES-256.

[0019] Furthermore, the secure storage interface uses a log-structured writing method to write the cache metadata, so as to restore the consistency state of the structured hierarchical queue according to the operation log after the satellite communication terminal experiences an abnormal power failure. When the cache space is full, expired or low-value data in the low-priority queue will be evicted first. When there are structured data packets to be uploaded in the high-priority queue, the transmission of data in the low-priority queue is paused and the current breakpoint state is saved. The system then switches to the high-priority queue to upload the data. After the data in the high-priority queue is uploaded, the transmission of data in the low-priority queue is resumed according to the current breakpoint state.

[0020] Compared with the prior art, the present invention has at least the following beneficial effects: 1. This invention provides a structured description of cached data by setting a metadata header for uplink data that includes a globally unique sequence number, data offset, data length, and checksum. This enables fine-grained management of different data packets and allows for precise location of lost data fragments based on the data offset.

[0021] 2. This invention introduces an offset-based breakpoint resumption mechanism, combined with ACK / NACK confirmation, to record file identifiers, confirmed offsets, and unconfirmed data packet lists when the satellite link is interrupted or switched. After the link is restored, only unconfirmed data fragments are retransmitted, thus avoiding full retransmission. The amount of retransmitted data after link interruption can be reduced by more than 90%, and the data transmission success rate can be improved from approximately 75% in traditional solutions to over 99%.

[0022] 3. By setting up high-priority queues, medium-priority queues, and low-priority queues, and by setting up a real-time service preemption mechanism, this invention can prioritize the uploading of real-time services and alarm data, with the preemption response time for high-priority real-time services being less than 100ms.

[0023] 4. This invention calculates the link quality score by comprehensively considering satellite elevation angle, signal-to-noise ratio, tariff rate, and estimated remaining transit time. It can selectively upload cached data based on satellite link status and tariff conditions, and move non-urgent data to the low-rate window for transmission, thereby reducing satellite bandwidth operating costs by approximately 40%.

[0024] 5. This invention utilizes the edge computing power of satellite communication terminals to aggregate, compress, and deduplicatize cached data, which can merge multiple small data packets into one data packet and eliminate redundant data that is repeatedly reported, thereby reducing the effective payload volume by about 60% to 80% and improving the utilization rate of satellite links.

[0025] 6. This invention uses a terminal key to locally encrypt and store cached data, and stores a verification value, which can improve the security of cached data; it also implements power failure protection through a structured log writing method, which can restore the consistency of the cache queue after an abnormal power outage.

[0026] 7. When the cache space is full, this invention prioritizes the removal of expired or low-value-density data from the low-priority queue, which can avoid the loss of high-value data caused by simple FIFO discarding, improve the cache space utilization rate by about 35%, and ensure that the system has 99.9% availability in extreme environments. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the system architecture of an intelligent caching system for satellite communication terminals based on edge computing, according to an embodiment of the present invention.

[0028] Figure 2 This is a flowchart illustrating the workflow of an edge computing-based intelligent caching system for satellite communication terminals according to an embodiment of the present invention. Detailed Implementation

[0029] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. It should be noted that the following content is used to illustrate the technical solution of the present invention and is not intended to limit the scope of protection of the present invention.

[0030] like Figure 1 As shown, this embodiment provides a satellite communication terminal intelligent caching system based on edge computing, including a user equipment set 101, a satellite communication terminal 102, an intelligent caching management module 103, and a satellite link 104.

[0031] User equipment set 101 includes one or more of sensors, mobile phones and cameras, used to generate uplink data and send the uplink data to satellite communication terminal 102. Figure 1 The dashed arrows in the diagram indicate the flow of uplink data sent from user equipment set 101 to satellite communication terminal 102, while the solid arrows indicate the flow of data uploaded by satellite communication terminal 102 through satellite link 104.

[0032] The satellite communication terminal 102 has a built-in edge computing unit, a cache queue, and a scheduling engine. It is used to receive uplink data from the user equipment set 101, cache and preprocess the uplink data, and upload the cached data through the satellite link 104 according to the link status of the satellite link 104 and the set scheduling policy.

[0033] The intelligent cache management module 103 is located within the satellite communication terminal 102 and includes a data classifier, a structured hierarchical queue, a breakpoint resume controller, an edge preprocessor, and a secure storage interface. The satellite communication terminal 102 also includes a link status monitor and upload scheduler.

[0034] The data classifier receives uplink data streams from user equipment set 101, parses the uplink data streams, and generates structured data packets with metadata headers. The metadata headers include a globally unique sequence number (SeqID), a data offset (Offset), a data length (Length), and a checksum (CRC / HASH).

[0035] A globally unique sequence number (SeqID) is used to uniquely identify structured data packets. A data offset (Offset) indicates the position of the corresponding data within the current transmission file. A data length (Length) indicates the length of the corresponding data. A checksum (CRC / HASH) is used to verify data integrity. By using the globally unique sequence number (SeqID), data offset (Offset), data length (Length), and checksum (CRC / HASH) to provide a structured description of uplink data, the satellite communication terminal 102 can identify, locate, and verify the integrity of different structured data packets.

[0036] The data classifier writes structured data packets into the corresponding priority buffer queues in the structured hierarchical queue based on the quality of service requirements of the structured data packets. The structured hierarchical queues include high-priority queues, medium-priority queues, and low-priority queues.

[0037] High-priority queues are used to store real-time business data, such as alarm data; medium-priority queues are used to store interactive business data; and low-priority queues are used to store background batch business data, such as data periodically uploaded by sensors, log data, and image data. By setting different priority cache queues, different types of data can be cached and uploaded according to business priority.

[0038] The link status monitoring and uploading scheduler monitors the link status of satellite link 104 in real time. The link status includes satellite elevation angle E, signal-to-noise ratio (SNR), tariff rate C, and estimated remaining transit time T. The unit of satellite elevation angle E is degrees, the unit of SNR is dB, the unit of tariff rate C is yuan / MB, and the unit of estimated remaining transit time T is minutes.

[0039] The link status monitoring and uploading scheduler calculates the link quality score (LQS) based on the satellite elevation angle (E), signal-to-noise ratio (SNR), tariff rate (C), and estimated remaining transit time (T). The link quality score (LQS) is calculated using the following formula: LQS=w1×(E / 90)+w2×(SNR / SNRmax)-w3×(C / Cmax)+w4×(T / Tmax)(1) Where w1=0.3, w2=0.3, w3=0.2, and w4=0.2. When the link quality score LQS is greater than the preset threshold LQSth, the satellite communication terminal 102 is determined to enter the upload window. The preset threshold LQSth is 0.6 by default.

[0040] By comprehensively considering satellite elevation angle, signal-to-noise ratio, tariff rate, and estimated remaining transit time, the satellite communication terminal 102 determines whether to enter the upload window, enabling it to upload data based on the communication quality, communication cost, and remaining available time of the satellite link 104. By using the tariff rate C as a parameter for calculating the Link Quality Score (LQS), the satellite communication terminal 102 can move non-urgent data to a lower-rate upload window for transmission and centrally upload cached data when the satellite signal is stable and the estimated remaining transit time meets the data upload requirements, thereby reducing satellite bandwidth operating costs by approximately 40%.

[0041] When satellite communication terminal 102 enters the upload window, the upload scheduler is activated, uploading the structured data packets in the buffer queue via satellite link 104 and checking the acknowledgment feedback returned by the receiving end. The acknowledgment feedback includes ACK and NACK.

[0042] The breakpoint resume controller records the transmission status of structured data packets based on acknowledgment feedback. When an acknowledgment feedback ACK is received, the acknowledgment offset ACK_Offset of the currently transmitted file is updated; when a negative acknowledgment feedback NACK is received or an acknowledgment timeout occurs, the lost data fragments are retransmitted based on the recorded acknowledgment offset ACK_Offset.

[0043] When satellite link 104 is interrupted or a link switch occurs, the breakpoint resumption controller records the file identifier (File_ID), the acknowledged offset (ACK_Offset), and a list of unacknowledged data packets for the currently transmitted file. After satellite link 104 is restored, the breakpoint resumption controller reads the breakpoint record and requests retransmission of unacknowledged data fragments based on the acknowledged offset (ACK_Offset), without retransmitting the entire currently transmitted file. By adopting the above-mentioned breakpoint resumption method based on data offsets, combined with acknowledgment feedback (ACK) and negative acknowledgment feedback (NACK), only unacknowledged data fragments need to be retransmitted after satellite link 104 is interrupted. This avoids repeatedly uploading data that has already been transmitted, reducing the amount of retransmitted data after link interruption by more than 90%, and increasing the data transmission success rate from approximately 75% in traditional schemes to over 99%. This improves the reliability of data transmission in satellite link 104 and reduces the bandwidth consumption of retransmitted data.

[0044] The edge preprocessor utilizes the edge computing power of the satellite communication terminal 102 to perform data aggregation, data compression, and data deduplication on the cached data.

[0045] For multiple small data packets in a low-priority queue, such as heartbeat packets periodically uploaded by sensors, the edge preprocessor merges multiple small data packets in the same low-priority queue into a single data packet to reduce the signaling overhead caused by uploading small data packets separately.

[0046] For log and image data in low-priority queues, the edge preprocessor compresses the data before uploading to reduce the amount of data that needs to be uploaded via satellite link 104. The edge preprocessor also identifies and removes duplicate reported redundant data. For example, when the sensor status remains unchanged, the edge preprocessor identifies and removes duplicate reported sensor data. By aggregating multiple small data packets, the signaling overhead caused by uploading multiple small data packets separately can be reduced; by compressing log and image data and removing duplicate reported redundant data, the payload volume uploaded via satellite link 104 can be reduced by approximately 60%–80%, thereby improving the transmission efficiency of satellite link 104.

[0047] The secure storage interface is connected to the physical storage medium of the satellite communication terminal 102, which is either eMMC or SSD. Before being written to the physical storage medium, cached data is encrypted using a terminal key, and a checksum of the cached data is stored. In this embodiment, the cached data is encrypted using AES-256.

[0048] The secure storage interface uses a log-based structured writing method to write cached metadata. In the event of an abnormal power outage in the satellite communication terminal 102, the write status of the cached metadata is saved through the operation log. After the satellite communication terminal 102 is powered on again, the operation log is read, and the consistency state of the structured hierarchical queue is restored based on the operation log, preventing the loss of cached metadata due to the abnormal power outage. By using AES-256 to encrypt the cached data written to the physical storage medium, user data can be protected. Even if the satellite communication terminal 102 is lost or the physical storage medium is damaged, the risk of cached data leakage is close to zero. By using a log-based structured writing method to save the write status of cached metadata, zero loss of cached metadata is guaranteed, and the system availability reaches 99.9% in extreme environments.

[0049] When the cache space in the physical storage medium is full, the intelligent cache management module 103 executes an intelligent eviction policy, prioritizing the eviction of expired or low-value-density data in the low-priority queue, rather than using a simple FIFO discard policy.

[0050] When data is being uploaded in a low-priority queue and a structured data packet to be uploaded appears in a high-priority queue, the upload scheduler pauses the transmission of data in the low-priority queue, the breakpoint resume controller saves the breakpoint state of the current low-priority data, and the upload scheduler switches to the high-priority queue, prioritizing the upload of data in the high-priority queue.

[0051] After data upload to the high-priority queue is complete, the breakpoint resume controller reads the saved current breakpoint state, and the upload scheduler resumes the transmission of data in the low-priority queue based on the current breakpoint state. This enables high-priority real-time services to preempt low-priority background batch services. By prioritizing the eviction of expired or low-value-density data in the low-priority queue when the cache space is full, the loss of high-value data caused by simple FIFO discarding can be avoided, improving cache space utilization by approximately 35%. By pausing the transmission of low-priority data and saving the current breakpoint state when data to be uploaded appears in the high-priority queue, the preemption response time for high-priority real-time services can be reduced to less than 100ms, while ensuring that low-priority data can continue to be transmitted after the high-priority data has been uploaded.

[0052] like Figure 2 As shown, the working process of the above-mentioned intelligent caching system for satellite communication terminals based on edge computing includes the following:

[0053] S1 receives uplink data sent by user equipment set 101, parses the uplink data, and generates a structured data packet with a metadata header. The metadata header includes a globally unique sequence number SeqID, a data offset, a data length, and a checksum CRC / HASH.

[0054] S2, the data classifier writes the structured data packets into the corresponding structured hierarchical queues according to the service quality requirements of the structured data packets. Specifically, real-time business data is written to the high-priority queue, interactive business data to the medium-priority queue, and background batch business data to the low-priority queue.

[0055] S3, the link status monitoring and upload scheduler monitors the satellite elevation angle E, signal-to-noise ratio SNR, tariff rate C, and estimated remaining transit time T in real time, and calculates the link quality score LQS according to formula (1). When the link quality score LQS is greater than the preset threshold LQSth, the satellite communication terminal 102 is determined to enter the upload window.

[0056] S4. When the satellite communication terminal 102 enters the upload window, the upload scheduler is started, the structured data packets in the buffer queue are uploaded through the satellite link 104, and the acknowledgment feedback ACK or negative acknowledgment feedback NACK returned by the receiving end is checked.

[0057] When a negative acknowledgment (NACK) is received or an acknowledgment timeout occurs, the breakpoint resume controller determines the data position that needs to be transmitted based on the acknowledged offset (ACK_Offset), and retransmits the lost data fragments to achieve precise fragment-level resume transmission.

[0058] When satellite link 104 is interrupted or switched, the breakpoint resume controller records the file identifier (File_ID), the acknowledged offset (ACK_Offset), and the list of unacknowledged data packets for the currently transmitted file. After satellite link 104 is restored, it continues to upload unacknowledged data fragments based on the acknowledged offset (ACK_Offset).

[0059] S5, before data is uploaded from the low-priority queue, the edge preprocessor aggregates multiple small packets of data from the same low-priority queue into a single data packet, compresses log and image data, and identifies and removes redundant data that is reported repeatedly.

[0060] S6, when data to be uploaded appears in the high-priority queue, the upload scheduler pauses the transmission of data in the low-priority queue, the breakpoint resume controller saves the current breakpoint state of the low-priority data, the upload scheduler switches to the high-priority queue, and prioritizes the upload of data in the high-priority queue.

[0061] After the data in the high-priority queue is uploaded, the breakpoint resume controller reads the saved current breakpoint state, and the upload scheduler continues to upload the remaining data in the low-priority queue based on the current breakpoint state.

[0062] S7, when the cache space is full, the intelligent cache management module 103 executes an intelligent eviction policy, prioritizing the eviction of expired or low-value-density data in the low-priority queue. When the satellite communication terminal 102 experiences an abnormal power outage, it saves the write status of cache metadata through a structured log writing method; after the satellite communication terminal 102 is powered on again, it reads the operation log and restores the consistency state of the structured hierarchical queue.

[0063] The following example illustrates the specific working process of the above system using a marine monitoring buoy that periodically transmits data via satellite communication terminal 102. Satellite communication terminal 102 is equipped with 128GB of local solid-state storage and an edge computing unit.

[0064] The sensors in the ocean monitoring buoy report data in JSON format every 5 minutes. After receiving the JSON data, the satellite communication terminal 102 assigns a globally unique sequence number (SeqID=0x1234), sets the data offset to 0, calculates the CRC32 checksum, and stores the generated structured data packet in the low-priority archiving area. The low-priority archiving area corresponds to the low-priority queue in the structured hierarchical queue.

[0065] Since the sensor reports only 1KB of data at a time, the edge preprocessor uses the edge computing power of the satellite communication terminal 102 to aggregate 288 data packets received within 24 hours into a super packet, and uses the LZMA algorithm to compress the super packet, achieving a compression ratio of 10:1.

[0066] The link status monitoring and upload scheduler monitors the link status of satellite link 104 in real time. When the system detects that the current time is 2:00 AM, the current period is a low-rate period, and the satellite signal is stable, the link quality score (LQS) calculated by the link status monitoring and upload scheduler is greater than the preset threshold (LQSth). The system then determines that satellite communication terminal 102 has entered the upload window, and the upload scheduler begins uploading the compressed super packet.

[0067] When the super packet was uploading the 100th data block, satellite link 104 suddenly stopped. The breakpoint resume controller wrote the following breakpoint record to non-volatile memory: File_ID: 20260710_Log; ACK_Offset: 99; Status: Paused.

[0068] The above breakpoint record indicates that the file identifier of the currently transmitted file is 20260710_Log, the first 99 data blocks have been acknowledged, and the current transmission status is paused.

[0069] Thirty minutes later, satellite link 104 was restored. Simultaneously, the ocean monitoring buoy detected abnormal vibrations and generated high-priority alarm data. At this point, data awaiting upload appeared in the high-priority queue. The upload scheduler paused the transmission of super packets in the low-priority queue, the breakpoint resume controller saved the breakpoint state of the current low-priority transmission, and the upload scheduler switched channels and prioritized the upload of high-priority alarm data.

[0070] After the high-priority alarm data upload is complete, the breakpoint resume controller reads the file identifier (File_ID), the confirmed offset (ACK_Offset), and the current transmission status. Based on ACK_Offset=99, the first 99 data blocks have been uploaded. The upload scheduler continues uploading the remaining data in the super packet from Offset=100, without re-uploading the already confirmed data blocks.

[0071] During the aforementioned data caching and uploading process, all data written to the local solid-state storage is encrypted using AES-256, and a corresponding checksum is stored simultaneously. If an abnormal power outage occurs during data writing, the satellite communication terminal 102 restarts and reads the operation log (Journal) to restore the consistency state of the structured hierarchical queue based on the operation log.

[0072] Through the aforementioned system, satellite communication terminal 102 can perform structured management of uplink data using globally unique sequence numbers (SeqID), data offsets (Offset), data lengths (Length), and checksums (CRC / HASH). It also enables fragment-level breakpoint resumption by combining acknowledgment feedback (ACK), negative acknowledgment feedback (NACK), and the acknowledged offset (ACK_Offset). Simultaneously, through structured hierarchical queuing, real-time service preemption, link availability window determination, edge data aggregation, compression and deduplication, local encrypted storage, power outage protection, and intelligent data eviction, it improves the reliability, efficiency, security, and buffer space utilization of uplink data transmission in the satellite communication network.

Claims

1. A satellite communication terminal intelligent caching system based on edge computing, characterized in that, Located in satellite communication terminals, including: A data classifier is used to receive uplink data streams from user equipment, parse the uplink data streams, and generate structured data packets with metadata headers, the metadata headers containing globally unique sequence numbers, data offsets, data lengths, and checksums. A structured hierarchical queue is used to write the structured data packets into a cache queue of corresponding priority according to the quality of service requirements of the structured data packets. The link status monitoring and upload scheduler is used to monitor the link status of the satellite link, calculate the link quality score based on the link status, and determine whether to enter the upload window based on the link quality score. After entering the upload window, the structured data packets in the cache queue are uploaded through the satellite link. The breakpoint resume controller is used to receive the confirmation feedback of the structured data packets. When the satellite link is interrupted or switched, it records the file identifier of the currently transmitted file, the confirmed offset, and the list of unconfirmed data packets. After the satellite link is restored, it retransmits the unconfirmed structured data packets according to the confirmed offset. An edge preprocessor is used to perform data aggregation, data compression, and data deduplication on the non-real-time data in the cache queue using the edge computing power of the satellite communication terminal before uploading. A secure storage interface is used to write the structured data packets and corresponding cached metadata into the physical storage medium of the satellite communication terminal.

2. The intelligent caching system for satellite communication terminals based on edge computing according to claim 1, characterized in that, The structured hierarchical queue includes a high-priority queue for storing real-time business data, a medium-priority queue for storing interactive business data, and a low-priority queue for storing background batch business data. The verification value is either a CRC check value or a HASH check value.

3. The intelligent caching system for satellite communication terminals based on edge computing according to claim 1, characterized in that, The link status includes satellite elevation angle E, signal-to-noise ratio SNR, tariff rate C, and estimated remaining transit time T. The link quality score (LQS) is calculated according to the following formula: LQS=w1×(E / 90)+w2×(SNR / SNRmax)-w3×(C / Cmax)+w4×(T / Tmax); Where w1=0.3, w2=0.3, w3=0.2, w4=0.2, when the link quality score LQS is greater than the preset threshold LQSth, it is determined to enter the upload window, and the preset threshold LQSth is 0.

6.

4. The intelligent caching system for satellite communication terminals based on edge computing according to claim 1, characterized in that, The confirmation feedback includes ACK and NACK. When the NACK is received or a confirmation timeout occurs, the breakpoint resume controller determines the lost data fragments based on the recorded confirmed offset and controls the link status monitoring and upload scheduler to retransmit the lost data fragments.

5. The intelligent caching system for satellite communication terminals based on edge computing according to claim 2, characterized in that, The edge preprocessor is used to aggregate multiple small packets of data in the low-priority queue into a single data packet, compress log data and / or image data in the low-priority queue, and identify and remove redundant data that is reported repeatedly.

6. The intelligent caching system for satellite communication terminals based on edge computing according to claim 1, characterized in that, The secure storage interface is used to encrypt the cached data using a terminal key and store the verification value of the cached data before writing the cached data to the eMMC or SSD; the cached data is encrypted using AES-256.

7. The intelligent caching system for satellite communication terminals based on edge computing according to claim 2, characterized in that, The secure storage interface uses a log-structured writing method to write the cache metadata, so as to restore the consistency state of the structured hierarchical queue according to the operation log after the satellite communication terminal experiences an abnormal power failure. When the cache space is full, expired or low-value data in the low-priority queue will be evicted first. When there are structured data packets to be uploaded in the high-priority queue, the transmission of data in the low-priority queue is paused and the current breakpoint state is saved. The system then switches to the high-priority queue to upload the data. After the data in the high-priority queue is uploaded, the transmission of data in the low-priority queue is resumed according to the current breakpoint state.

Images