Multi-terminal edge access, semantic acceleration and service orchestration system and method based on stationary orbit satellite internet

Abstract

The application provides a multi-terminal edge access, semantic acceleration and service arrangement system based on a stationary orbit satellite internet, relates to the technical field of satellite communication, and comprises the following steps: performing semantic extraction and semantic compression on original data, generating a target information package, acquiring real-time characteristic parameters of each link and combining a historical link database to generate a comprehensive evaluation vector of each link, dynamically determining whether a single-link transmission mode or a multi-link fusion transmission mode is adopted based on the target information package and the comprehensive evaluation vector and in combination with a service priority, transmitting the target information package to a gateway station edge node, a regional edge node or a cloud service platform through a target transmission path, and finally calculating a stability coefficient of the target transmission path and storing the stability coefficient and the target transmission path in a historical link database. The application can reduce satellite link bandwidth occupation, improve the reliability of high-priority services and the utilization rate of link resources.

Description

Technical Field

[0001] This invention relates to the technical field of geostationary satellite internet, specifically to a multi-terminal edge access, semantic acceleration, and service orchestration system and method based on geostationary satellite internet. Background Technology

[0002] Geostationary orbit communication satellites have advantages such as wide coverage, large area coverage per satellite, fast deployment speed, and suitability for areas with weak terrestrial communication infrastructure. With the development of high-throughput satellites, broadband transponders, and multi-beam technology, geostationary orbit satellite links are now able to undertake services such as Internet access, private network communication, emergency command, video backhaul, and industry data transmission.

[0003] In existing technologies, various types of service data are typically collected from field acquisition terminals via wired or short-range wireless methods and then aggregated to satellite terminal equipment. This satellite terminal encapsulates all upper-layer service data into standard IP data packets, which are then transparently transmitted to the ground gateway station via a satellite link. Finally, the ground service platform completes all data processing and decision-making. In scenarios where multiple links are available, existing systems typically employ shortest path or static routing strategies based on IP routing to select a single transmission link.

[0004] However, geostationary satellites are far from the Earth's surface, resulting in significant propagation delays and susceptibility to factors such as rain attenuation, obstruction, link congestion, high bandwidth costs, and limited backhaul resources. This leads to marked instability and time-varying link quality. In emergency response, drone inspections, and remote sensing operations, field terminals generate a large amount of images, videos, audio, target identification results, and sensor data. If all of this data were transmitted back via geostationary satellite links in its raw form, it would consume a significant amount of bandwidth, thereby reducing the reliability of high-priority services such as critical commands, voice messages, SOS signals, drone return-to-home commands, and work orders. Existing link management typically relies on fixed thresholds for judgment. When link resources are strained, existing technologies generally reduce the overall transmission experience of all services or passively switch links, thus failing to ensure that the transmission of all service types meets the transmission requirements. In multi-link scenarios, existing systems usually adopt a single routing strategy. After selecting a link, a static bandwidth allocation mechanism is usually used for data transmission. When there are large data packets, each link cannot adaptively allocate bandwidth, resulting in a large backlog of data packets or even packet loss when a single link is congested or its quality deteriorates, while other available links are idle or under low load, resulting in low overall link resource utilization. Summary of the Invention

[0005] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a multi-terminal edge access, semantic acceleration, and service orchestration system and method based on geostationary satellite internet. It solves the technical problems of excessive bandwidth consumption, insufficient reliability of service priority transmission, inability to adaptively allocate bandwidth among links when transmitting large-capacity data packets, and low utilization of each link.

[0006] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: A multi-terminal edge access, semantic acceleration, and service orchestration system based on geostationary satellite internet includes: The information packet generation module extracts semantics from the raw data and performs semantic compression to generate the target information packet; The evaluation vector generation module obtains the real-time feature parameters of each current link of the satellite, and combines them with the historical feature parameters of the corresponding links in the historical link database to generate a comprehensive evaluation vector for each link. The transmission path determination module determines the transmission strategy based on the target information packet and the comprehensive evaluation vector of each link, combined with the preset service priority. The transmission strategy includes the following two methods: In single-link transmission mode, one link is selected from all links as the target transmission path. Multi-link fusion transmission method, which merges at least two links from each link to obtain the target transmission path; The information packet transmission module transmits the target information packet to the corresponding target node in the gateway station edge node, regional edge node, or cloud service platform through the target transmission path, and distributes the data returned by the target node to the field terminal. The path stability assessment module obtains the transmission quality parameters of the target information packet on the target transmission path, calculates the stability coefficient of the target transmission path, and stores the target transmission path and the corresponding stability coefficient together in the historical link database.

[0007] Preferably, the generation process of the target information packet includes: extracting semantic features and compressing semantic features from the original data based on the modal features of the original data, and generating a compressed semantic feature vector; The compressed semantic feature vector is combined with one or more of the following: task number, geographic coordinates, timestamp, terminal type, business type, event description, confidence level, risk level, original data index, keyframe thumbnail, and handling suggestions to form the target information package.

[0008] Preferably, the process of generating the comprehensive evaluation vector includes: obtaining the standard values ​​of the characteristic parameters of the satellite links, calculating the relative deviation between the real-time characteristic parameters of each link and the standard values ​​of the characteristic parameters, and forming the corresponding first deviation vector; Calculate the relative deviation between the real-time characteristic parameters of each link and the historical characteristic parameters of the corresponding link, and construct the corresponding second deviation vector; The first deviation vector is corrected based on the second deviation vector to obtain the corrected deviation vector for the link. The real-time characteristic parameters of each link and the corrected deviation vector of the corresponding link are fused to obtain the fused feature vector, which is used as the comprehensive evaluation vector of the link.

[0009] Preferably, the preset service priorities, in descending order, include: life safety, command and communication, production operations, general internet, and background synchronization.

[0010] Preferably, the process of determining the transmission strategy includes: obtaining the service type and data volume of the target information packet, and generating a transmission requirement vector; Calculate the cosine similarity between the comprehensive evaluation vector and the transmission demand vector of each link, and denote it as the service matching degree of the corresponding link; Links with a business matching degree lower than a preset similarity threshold are removed, and the remaining links are recorded as a candidate link set. Calculate the product of the maximum available bandwidth of each link in the candidate link set and the corresponding service matching degree, and use it as the carrying index of the corresponding link. The ratio of the data size of the target packet to the allowable transmission delay is used as the minimum transmission throughput. If there is a link with a carrying capacity greater than or equal to the minimum transmission throughput, then single-link transmission mode is triggered. Otherwise, multi-link transmission mode will be triggered.

[0011] Preferably, the single-link transmission method includes: From the candidate link set, select the link with the highest service matching degree as the target transmission path; If there are multiple links with the highest service matching degree, the link with the largest available bandwidth is selected as the target transmission path.

[0012] Preferably, the multi-link fusion transmission method includes: selecting several links with the highest available bandwidth from the candidate link set, and the absolute value of the difference between the working frequency bands of any two links in the selected links is greater than a preset frequency diversity threshold, which is denoted as the aggregated link set; The target information packet is encoded into multiple redundant data blocks through forward erasure coding, and the encoded data blocks are distributed to several links in the aggregated link set for parallel transmission.

[0013] A method for multi-terminal edge access, semantic acceleration, and service orchestration based on geostationary satellite internet includes: semantic extraction of raw data and semantic compression to generate target information packets; The real-time characteristic parameters of each current link of the satellite are obtained, and combined with the historical characteristic parameters of the corresponding links in the historical link database, a comprehensive evaluation vector for the corresponding link is generated. Based on the target information packet and the comprehensive evaluation vector of each link, the transmission strategy is determined in combination with the preset service priority. The transmission strategy includes the following two methods: In single-link transmission mode, one link is selected from all links as the target transmission path. Multi-link fusion transmission method, which merges at least two links from each link to obtain the target transmission path; The target information packet is transmitted to the corresponding target node in the gateway station edge node, regional edge node or cloud service platform through the target transmission path, and the data returned by the target node is distributed to the field terminal. Obtain the transmission quality parameters of the target information packet on the target transmission path, calculate the stability coefficient of the target transmission path, and store the target transmission path and the corresponding stability coefficient together in the historical link database.

[0014] A storage medium storing a computer program, wherein the computer program causes a computer to execute the multi-terminal edge access, semantic acceleration, and service orchestration method based on geostationary satellite internet as described above.

[0015] An electronic device includes: a processor and a memory; the memory stores program instructions; the processor is configured to run the program instructions to execute the multi-terminal edge access, semantic acceleration, and service orchestration method based on geostationary satellite internet as described above.

[0016] (III) Beneficial Effects This invention provides a multi-terminal edge access, semantic acceleration, and service orchestration system and method based on geostationary satellite internet. Compared with existing technologies, it has the following advantages: This invention transforms raw data into semantic feature vectors through semantic compression and generates target information packets containing these vectors, thereby reducing the bandwidth burden on the links. By evaluating the quality of each link and combining service priorities with the target information packets, the invention analyzes and dynamically determines transmission strategies, adaptively allocating single-link transmission methods and multi-link fusion transmission methods. This ensures the complete transmission of the target information packets while improving the utilization efficiency of the links. Furthermore, the invention evaluates the stability of the transmitted path and feeds it back to the historical link database, providing a basis for subsequent link quality assessments. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a module connection diagram of the present invention.

[0019] Figure 2 This is a schematic diagram of the overall implementation process of the present invention.

[0020] Figure 3 A schematic diagram illustrating the implementation process for determining the transmission strategy. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are described clearly and completely. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] This application provides a multi-terminal edge access, semantic acceleration, and service orchestration system and method based on geostationary satellite internet. It solves the technical problems of excessive raw data content, easy occupation of large bandwidth, insufficient transmission reliability of service priorities, inability of each link to adaptively allocate when transmitting large-capacity data packets, and low utilization of each link. It achieves the goals of reducing satellite link bandwidth occupation, improving transmission efficiency, improving the reliability of high-priority services, and improving the overall link resource utilization.

[0023] The technical solution in this application is to solve the above-mentioned technical problems, and the general idea is as follows: First, the information packet generation module performs semantic extraction and semantic compression on the original data, transforming the original data into target information packets containing core semantic feature vectors. The semantic features replace the original large-capacity data such as images, videos, and audios for transmission, thereby reducing the amount of data to be transmitted and alleviating the bandwidth pressure on the satellite link.

[0024] Secondly, the real-time characteristic parameters of each link are obtained through the evaluation vector generation module, and combined with the historical characteristic parameters in the historical link database to generate a comprehensive evaluation vector for each link. This vector not only reflects the current real-time status of the link, but also corrects for real-time measurement deviations using historical data, making the link evaluation results more accurate and reliable.

[0025] Next, the transmission path determination module calculates the service matching degree between the service type of the target information packet and the comprehensive evaluation vector of each link, and eliminates links with matching degrees below a preset threshold to form a candidate link set. Based on this, by comparing the carrying index of each link with the minimum transmission throughput required by the target information packet, it adaptively determines whether to adopt a single-link transmission mode or a multi-link fusion transmission mode. When a single link is sufficient to meet the transmission requirements, the link with the highest service matching degree is selected for transmission; when a single link cannot meet the transmission requirements, multiple links with high available bandwidth and large frequency band differences are selected for fusion transmission, and the target information packet is encoded into multiple redundant data blocks for parallel transmission through forward erasure coding, ensuring that the receiving end can restore the complete information by receiving partial data blocks.

[0026] Finally, the target information packet is transmitted to the corresponding target node in the gateway edge node, regional edge node, or cloud service platform through the target transmission path. The data returned by the target node is distributed to the field terminal. At the same time, the stability coefficient of the target transmission path is calculated, and the target transmission path and the corresponding stability coefficient are stored together in the historical link database to provide historical basis for subsequent link quality assessment and transmission decisions, so that the system has continuous self-optimization capability in long-term operation.

[0027] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0028] See Figure 1 A multi-terminal edge access, semantic acceleration, and service orchestration system based on geostationary satellite internet includes: an information packet generation module that extracts semantics from raw data and generates target information packets after semantic compression; Specifically, the generation process of the target information packet includes: extracting semantic features and compressing semantic features from the original data based on the modal features of the original data, and generating a compressed semantic feature vector; The compressed semantic feature vector is combined with one or more of the following: task number, geographic coordinates, timestamp, terminal type, business type, event description, confidence level, risk level, original data index, keyframe thumbnail, and handling suggestions to form the target information package.

[0029] It should be added that the semantic feature extraction includes: when the original data is text data, using the TF-IDF statistical method to extract semantic feature vectors.

[0030] When the original data is image data, the SIFT feature extraction method is used to extract semantic features.

[0031] When the original data is video data, video keyframes are first extracted at predetermined time intervals, and then the SIFT feature extraction method is used to process the data to obtain semantic feature vectors. For example, the predetermined time interval ranges from 3 seconds to 15 seconds.

[0032] This invention uses discrete wavelet transform to perform semantic compression on semantic feature vectors, with the compression ratio ranging from 8 to 16 times.

[0033] The TF-IDF statistical method, SIFT feature extraction method, and discrete wavelet transform mentioned above are all existing technologies and will not be described in detail in this invention.

[0034] The evaluation vector generation module obtains the real-time characteristic parameters of each link of the satellite and combines them with the historical characteristic parameters of the corresponding links in the historical link database to generate a comprehensive evaluation vector for each link.

[0035] It should be noted that the real-time characteristic parameters include link performance parameters and environmental and interference parameters. Specifically, the link performance parameters include one or more of link bandwidth, latency, jitter, packet loss rate, signal-to-noise ratio, and signal strength; the environmental and interference parameters include one or more of rainfall attenuation, atmospheric absorption, cloud and fog attenuation, and co-channel interference.

[0036] Among them, "multiple" refers to two or more.

[0037] It should be added that the aforementioned historical link database uses the link ID as the primary key, with each transmission on each link counted as a record. It stores characteristic parameters for each link, including timestamp, link bandwidth, latency, jitter, packet loss rate, signal-to-noise ratio, signal strength, rainfall attenuation, atmospheric absorption, cloud and fog attenuation, co-channel interference, and the corresponding stability coefficient. The database retains the most recent 100 transmission records or records from the last 24 hours; older data outside this range is automatically archived or deleted.

[0038] Specifically, the process of generating the comprehensive evaluation vector includes: obtaining the standard values ​​of the characteristic parameters of the satellite links, calculating the relative deviation between the real-time characteristic parameters of each link and the standard values ​​of the characteristic parameters, and constructing the corresponding first deviation vector. The calculation formula is: ,in, These are the standard values ​​of characteristic parameters of the satellite link. These are the real-time characteristic parameters of the link. The number of feature parameters. It is the transpose of the vector.

[0039] It should be added that the standard values ​​of the above characteristic parameters can be obtained from the link design specifications of satellite communication system equipment.

[0040] Calculate the relative deviation between the real-time characteristic parameters of each link and the historical characteristic parameters of the corresponding link, and construct the corresponding second deviation vector. The calculation formula is: ,in, These are historical characteristic parameters of the satellite link.

[0041] It should be noted that when a link does not yet have historical feature parameters, the second deviation vector is set to zero, and only the first deviation vector is used as the corrected deviation vector.

[0042] The first deviation vector is corrected based on the second deviation vector to obtain the corrected deviation vector for the link. It should be noted that the corrected deviation vector described above can be obtained in the following way: First, the first deviation vector Input the exponential decay function to obtain the correction factor vector. .

[0043] The exponential decay function mentioned above is: ,in, This represents taking the absolute value of each component of the first deviation vector. The same attenuation coefficient value is used for all characteristic parameters as a preset attenuation coefficient. The value ranges from 0.5 to 5. For example, The value is 1.

[0044] It should be noted that the above exponentiation calculation This is a component-wise operation, i.e.: , where i = 1, 2, ..., n.

[0045] Then the correction factor vector With the second deviation vector After performing component-by-component multiplication, a product vector is obtained. Finally, the product vector is added to the first deviation vector to obtain the corrected deviation vector for this link. , i = 1, 2, ..., n.

[0046] The real-time characteristic parameters of each link and the corrected deviation vector of the corresponding link are fused to obtain the fused feature vector, which is used as the comprehensive evaluation vector for that link. Feature fusion can be achieved through vector concatenation, which is an existing technology and will not be described in detail here. Those skilled in the art can choose according to actual needs, and this invention does not limit it.

[0047] By using the above-mentioned correction method, this invention ensures that when the first deviation vector is small, i.e., the real-time feature parameter is close to the standard value, the correction factor vector approaches 1, and the second deviation vector is completely preserved, so that historical degradation information can correct the current first deviation vector.

[0048] When the first deviation vector is large, meaning the real-time feature parameters have deviated significantly from the standard value, the correction factor vector approaches 0, and the contribution of the second deviation is automatically suppressed to avoid over-correction.

[0049] Throughout the correction process, the correction factor vector changes continuously with the first deviation vector, thus ensuring a smooth correction process.

[0050] This invention integrates real-time and historical feature parameters of the link to obtain a multi-dimensional comprehensive evaluation vector that can fully and dynamically reflect the true status of the link, making the link quality assessment more accurate and avoiding misjudgments caused by evaluation based on a single indicator.

[0051] The transmission path determination module determines the transmission strategy based on the target information packet and the comprehensive evaluation vector of each link, combined with the preset service priority. Specifically, the preset business priorities, in descending order, include: life safety, command and communication, production operations, general internet, and background synchronization.

[0052] Furthermore, the aforementioned life safety services include SOS, rescue coordinates, drone return instructions, and emergency command instructions; the aforementioned command and communication services include voice, SMS, low-bitrate video, and task work orders; the aforementioned production operation services include drone inspection data, equipment status data, and sensor data; the aforementioned general internet services include web browsing, general chat, and file downloads; and the aforementioned background synchronization services include logs and model updates.

[0053] The transmission strategy includes the following two methods: In single-link transmission mode, one link is selected from all links as the target transmission path. The multi-link fusion transmission method combines at least two links from each link to obtain the target transmission path; See Figure 3 Specifically, the process of determining the transmission strategy includes: obtaining the service type and data volume of the target information packet, and generating a transmission demand vector; Calculate the cosine similarity between the comprehensive evaluation vector and the transmission demand vector of each link, and denote it as the service matching degree of the corresponding link; The cosine similarity mentioned above is existing technology and will not be shown or described in detail in this invention.

[0054] Links with a business matching degree lower than a preset similarity threshold are removed, and the remaining links are recorded as a candidate link set. For example, the preset similarity threshold is set to 0.7.

[0055] Calculate the product of the maximum available bandwidth of each link in the candidate link set and the corresponding service matching degree, and use it as the carrying index of the corresponding link. The ratio of the data size of the target packet to the allowable transmission delay is used as the minimum transmission throughput. It should be added that the above-mentioned allowable transmission delay can be determined comprehensively based on the service type of the target information packet and the preset service priority. For example, for life safety services, the allowable transmission delay is set to less than or equal to 1 second, and for production operation services, the allowable transmission delay is set to less than or equal to 30 seconds.

[0056] If there is a link with a carrying capacity greater than or equal to the minimum transmission throughput, then single-link transmission mode is triggered. Otherwise, multi-link transmission mode will be triggered.

[0057] Specifically, the single-link transmission method includes: selecting the link with the highest service matching degree from the candidate link set as the target transmission path; If there are multiple links with the highest service matching degree, the link with the largest available bandwidth will be selected as the target transmission path. "Multiple links" refers to two or more links.

[0058] Specifically, the multi-link fusion transmission method includes: selecting several links with the highest available bandwidth from the candidate link set, and the absolute value of the difference between the working frequency bands of any two selected links is greater than a preset frequency diversity threshold, which is denoted as an aggregated link set; wherein, the aforementioned several links are two or more.

[0059] It should be noted that the above-mentioned preset frequency diversity threshold can be determined according to the operating frequency band of satellite communication. It is set to 2 GHz in the Ku band and 5 GHz in the Ka band. When the absolute value of the difference between the center frequencies of the two links is greater than the preset frequency diversity threshold, it is determined that the channel fading of the two links is unrelated under the same rainfall environment, thereby ensuring that frequency diversity gain can be obtained when multiple links are transmitted in parallel.

[0060] By using forward erasure coding, the target information packet is encoded into multiple redundant data blocks, and the encoded data blocks are distributed to several links in the aggregated link set for parallel transmission. This allows the receiving end to fully restore the target information packet through erasure decoding without receiving all the redundant data blocks. Even if a link is temporarily interrupted due to rain attenuation or other reasons, it will not affect the overall transmission, greatly improving the robustness of the system.

[0061] Among them, the aforementioned redundant data blocks are three or more, and the number of data entries are two or more.

[0062] As a preferred embodiment of the present invention, the aforementioned forward erasure coding can be fountain coding. The specific coding algorithm of fountain coding is existing technology and will not be described in detail here.

[0063] This invention proactively predicts based on the comprehensive evaluation vector of the target information packet and each link, as well as the preset service priority, thereby providing a basis for selecting the optimal transmission path for the target information packet. When a single link cannot meet the requirements, it triggers multi-link fusion transmission. Based on the bandwidth and operating frequency band of the link, it obtains an aggregated link set, thereby improving the throughput of data transmission through multiple links in parallel, so as to improve the link utilization.

[0064] The information packet transmission module transmits the target information packet to the corresponding target node in the gateway station edge node, regional edge node, or cloud service platform through the target transmission path, and distributes the data returned by the target node to the field terminal. In one embodiment, the gateway edge node is deployed on the satellite ground gateway side, and has low-latency parsing and local rapid processing capabilities. It is used to process target information packets of high-priority services such as life safety and command and communication, and return the processed results to the field terminal through the satellite link.

[0065] The regional edge nodes are deployed at the regional center and have the ability to aggregate and collaboratively process data from multiple gateway stations. They are used to process target information packages for production operations and can analyze data from multiple gateway station edge nodes within the region.

[0066] The cloud-based business platform is deployed in the central cloud and has the capabilities of global data aggregation, in-depth analysis, and long-term storage. It is used to process target information packages for general Internet-related and backend synchronous-related businesses.

[0067] The transmission path determination module selects the corresponding target node based on the service priority of the target information packet: life safety and command communication services are preferentially transmitted to the gateway edge node to utilize its latency response capability; production operation services are transmitted to the regional edge node; and ordinary Internet and background synchronization services are transmitted to the cloud service platform. If the target node corresponding to a high-priority service is unavailable, such as due to excessive load or failure of the gateway edge node, it is automatically downgraded to the next higher-level node, the regional edge node, or the cloud service platform for processing.

[0068] The path stability assessment module obtains the transmission quality parameters of the target information packet on the target transmission path, calculates the stability coefficient of the target transmission path, and stores the target transmission path and the corresponding stability coefficient together in the historical link database.

[0069] In one embodiment, when the service type of the target information packet belongs to the life safety category or the command and communication category, the stability coefficient is obtained by acquiring the transmission parameters of the target transmission path during the transmission of the target information packet. The transmission parameters include one or more of transmission delay, delay jitter, packet loss rate, and signal strength fluctuation. The transmission parameters are subjected to minimum-maximum linear normalization.

[0070] It should be added that the maximum and minimum values ​​required for the above-mentioned minimum-maximum linear normalization are obtained from the transmission parameters. That is, the maximum and minimum values ​​of the corresponding parameters are selected from all transmission parameters and used as the maximum and minimum values ​​required for the normalization of the corresponding parameters, respectively. The minimum-maximum linear normalization process is an existing technology and will not be described in detail here.

[0071] Based on the preset weight coefficients corresponding to each transmission parameter, the weighted sum of each transmission parameter after minimum-maximum linear normalization is performed to obtain the stability coefficient of the target transmission path; wherein, the sum of the preset weight coefficients corresponding to each transmission parameter is 1.

[0072] The preset weighting coefficients can be dynamically adjusted according to the service type. For example, for life safety services, in order to ensure real-time performance and reliability, the weights of transmission delay, delay jitter, packet loss rate and signal strength fluctuation are set to 0.45, 0.25, 0.2 and 0.1, respectively. For ordinary Internet services, in order to ensure throughput and stability, the weights of transmission delay, delay jitter, packet loss rate and signal strength fluctuation are set to 0.2, 0.15, 0.35 and 0.3, respectively.

[0073] In another embodiment, when the business type of the target information packet is production operation, general Internet, or background synchronization, the stability coefficient can be obtained in the following way: The ratio of the amount of data correctly received by the receiver and the total amount of data sent by the sender during the transmission of the target information packet is calculated as a stability coefficient.

[0074] See Figure 2 This invention provides a method for multi-terminal edge access, semantic acceleration, and service orchestration based on geostationary satellite internet, including: semantic extraction of raw data and semantic compression to generate target information packets; The real-time characteristic parameters of each current link of the satellite are obtained, and combined with the historical characteristic parameters of the corresponding links in the historical link database, a comprehensive evaluation vector for the corresponding link is generated. Based on the target information packet and the comprehensive evaluation vector of each link, the transmission strategy is determined in combination with the preset service priority. The transmission strategy includes the following two methods: In single-link transmission mode, one link is selected from all links as the target transmission path. The multi-link fusion transmission method combines at least two links from each link to obtain the target transmission path; The target information packet is transmitted to the corresponding target node in the gateway station edge node, regional edge node or cloud service platform through the target transmission path, and the data returned by the target node is distributed to the field terminal. Obtain the transmission quality parameters of the target information packet on the target transmission path, calculate the stability coefficient of the target transmission path, and store the target transmission path and the corresponding stability coefficient together in the historical link database.

[0075] This invention provides a storage medium storing a computer program, wherein the computer program enables a computer to execute the multi-terminal edge access, semantic acceleration, and service orchestration method based on geostationary satellite internet as described above.

[0076] This invention also provides an electronic device, the electronic device comprising: A processor and a memory; the memory stores program instructions; the processor is used to run the program instructions to execute the multi-terminal edge access, semantic acceleration and service orchestration method based on geostationary satellite internet as described above.

[0077] It is understood that the multi-terminal edge access, semantic acceleration and service orchestration system, storage medium and electronic device based on geostationary satellite internet provided in the embodiments of the present invention correspond to the multi-terminal edge access, semantic acceleration and service orchestration method based on geostationary satellite internet provided in the embodiments of the present invention. The explanation, examples and beneficial effects of the relevant content can be referred to the corresponding parts of the multi-terminal edge access, semantic acceleration and service orchestration method based on geostationary satellite internet, and will not be repeated here.

[0078] In summary, compared with existing technologies, it has the following beneficial effects: 1. This invention uses semantic features to replace the original data for transmission, reducing the amount of data transmitted while retaining key semantic information, thereby reducing the pressure on the bandwidth of geostationary satellite links.

[0079] 2. Based on the quality assessment of the target information packet and each link, this invention determines the single-link transmission mode and the multi-link fusion transmission mode by combining the service priority. At the same time, the multi-link fusion transmission mode realizes adaptive load balancing and collaborative transmission among multiple links.

[0080] 3. The present invention uses a path stability assessment module to feed back and store the stability coefficient of each transmission to a historical link database, so that the assessment vector generation module can correct real-time measurement deviations based on historical data in subsequent decisions.

[0081] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0082] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-terminal edge access, semantic acceleration, and service orchestration system based on geostationary satellite internet, characterized in that: include: The information packet generation module extracts semantics from the raw data and performs semantic compression to generate the target information packet; The evaluation vector generation module obtains the real-time feature parameters of each current link of the satellite, and combines them with the historical feature parameters of the corresponding links in the historical link database to generate a comprehensive evaluation vector for each link. The transmission path determination module determines the transmission strategy based on the target information packet and the comprehensive evaluation vector of each link, combined with the preset service priority. The transmission strategy includes single-link transmission mode and multi-link fusion transmission mode. The information packet transmission module transmits the target information packet to the corresponding target node in the gateway station edge node, regional edge node, or cloud service platform through the target transmission path, and distributes the data returned by the target node to the field terminal. The path stability assessment module obtains the transmission quality parameters of the target information packet on the target transmission path, calculates the stability coefficient of the target transmission path, and stores the target transmission path and the corresponding stability coefficient together in the historical link database.

2. The multi-terminal edge access, semantic acceleration, and service orchestration system as described in claim 1, characterized in that, The process of generating the target information packet includes: Based on the modal characteristics of the original data, semantic features are extracted and semantic compression is performed on the original data to generate compressed semantic feature vectors; The compressed semantic feature vector is combined with one or more of the following: task number, geographic coordinates, timestamp, terminal type, business type, event description, confidence level, risk level, original data index, keyframe thumbnail, and handling suggestions to form the target information package.

3. The multi-terminal edge access, semantic acceleration, and service orchestration system as described in claim 1, characterized in that, The process of generating the comprehensive evaluation vector includes: Obtain the standard values ​​of the characteristic parameters of the satellite links, calculate the relative deviation between the real-time characteristic parameters of each link and the standard values ​​of the characteristic parameters, and construct the corresponding first deviation vector; Calculate the relative deviation between the real-time characteristic parameters of each link and the historical characteristic parameters of the corresponding link, and construct the corresponding second deviation vector; The first deviation vector is corrected based on the second deviation vector to obtain the corrected deviation vector for the link. The real-time characteristic parameters of each link and the corrected deviation vector of the corresponding link are fused to obtain the fused feature vector, which is used as the comprehensive evaluation vector of the link.

4. The multi-terminal edge access, semantic acceleration, and service orchestration system as described in claim 1, characterized in that, The preset business priorities, in descending order, include: life safety, command and communication, production operations, general internet, and background synchronization.

5. The multi-terminal edge access, semantic acceleration, and service orchestration system as described in claim 1, characterized in that, The process of determining the transmission strategy includes: Obtain the business type and data volume of the target information packet, and generate a transmission requirement vector; Calculate the cosine similarity between the comprehensive evaluation vector and the transmission demand vector of each link, and denote it as the service matching degree of the corresponding link; Links with a business matching degree lower than a preset similarity threshold are removed, and the remaining links are recorded as a candidate link set. Calculate the product of the maximum available bandwidth of each link in the candidate link set and the corresponding service matching degree, and use it as the carrying index of the corresponding link. The ratio of the data size of the target packet to the allowable transmission delay is used as the minimum transmission throughput. If there is a link with a carrying capacity greater than or equal to the minimum transmission throughput, then single-link transmission mode is triggered. Otherwise, multi-link transmission mode will be triggered.

6. The multi-terminal edge access, semantic acceleration, and service orchestration system as described in claim 1, characterized in that, The single-link transmission method includes: From the candidate link set, select the link with the highest service matching degree as the target transmission path; If there are multiple links with the highest service matching degree, the link with the largest available bandwidth is selected as the target transmission path.

7. The multi-terminal edge access, semantic acceleration, and service orchestration system as described in claim 1, characterized in that, The multi-link fusion transmission method includes: From the candidate link set, select several links with the highest available bandwidth, and the absolute value of the difference between the working frequency bands of any two selected links is greater than the preset frequency diversity threshold. This is called the aggregated link set. The target information packet is encoded into multiple redundant data blocks through forward erasure coding, and the encoded data blocks are distributed to several links in the aggregated link set for parallel transmission.

8. A method for multi-terminal edge access, semantic acceleration, and service orchestration based on geostationary satellite internet, characterized in that: include: Semantic extraction and semantic compression are performed on the original data to generate the target information package; The real-time characteristic parameters of each current link of the satellite are obtained, and combined with the historical characteristic parameters of the corresponding links in the historical link database, a comprehensive evaluation vector for the corresponding link is generated. Based on the target information packet and the comprehensive evaluation vector of each link, combined with the preset service priority, the transmission strategy is determined; The transmission strategy includes the following two methods: In single-link transmission mode, one link is selected from all links as the target transmission path. Multi-link fusion transmission method, which merges at least two links from each link to obtain the target transmission path; The target information packet is transmitted to the corresponding target node in the gateway station edge node, regional edge node or cloud service platform through the target transmission path, and the data returned by the target node is distributed to the field terminal. Obtain the transmission quality parameters of the target information packet on the target transmission path, calculate the stability coefficient of the target transmission path, and store the target transmission path and the corresponding stability coefficient together in the historical link database.

9. A storage medium, characterized in that, It stores a computer program, wherein the computer program causes the computer to execute the multi-terminal edge access, semantic acceleration and service orchestration method based on geostationary satellite internet as described in claim 8.

10. An electronic device, characterized in that, The electronic device includes: A processor and a memory; the memory stores program instructions; the processor is used to run the program instructions to execute the multi-terminal edge access, semantic acceleration and service orchestration method based on geostationary satellite internet as described in claim 8.

Images