Method, system and medium for link backup and composite load based on service message identification

Abstract

The application discloses a link backup and composite load method and system based on service message identification and a medium, and relates to the technical field of wireless communication; message forwarding backup or composite load strategy is generated in combination with internal marks of various service messages and link state results of various links; deep coupling of deep service perception (message internal mark) and multi-link state perception (link state result) is realized, a predictive and service policy-based conditional trigger type backup mechanism is realized; and the utilization rate of overall link resources is maximized while the transmission reliability and real-time performance of key service messages are improved.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and more specifically to a method, system, and medium for link backup and composite load based on service message identification. Background Technology

[0002] In critical mission scenarios such as emergency communications, industrial IoT, and remote control, a single 5G network may face risks such as coverage blind spots, network congestion, or core network failures. Wireless ad hoc networks (such as mesh networks) have advantages such as self-organization, self-healing, and decentralization, and can serve as an effective supplement to 5G, but they themselves suffer from limited bandwidth and increased latency due to increased hop count. Existing technologies typically employ simple dual-transmission selective reception or failover mechanisms, lacking intelligence.

[0003] In critical mission scenarios such as emergency communications, industrial IoT, and remote control, a single 5G network may face risks such as coverage blind spots, network congestion, or core network failures. Wireless ad hoc networks (such as mesh networks) have advantages such as self-organization, self-healing, and decentralization, and can serve as an effective supplement to 5G, but they themselves suffer from limited bandwidth and increased latency due to increased hop count. Existing technologies typically employ simple dual-transmission selective reception or failover mechanisms, lacking intelligence.

[0004] Traditional backup mechanisms often operate on an "all or nothing" model, backing up the entire business flow. This not only results in a significant waste of link resources but can also lead to a deterioration in the quality of the main link due to backup flows preempting resources. Furthermore, switching decisions are often based on simple link connectivity, making them slow to react to dynamic quality indicators such as latency, jitter, and packet loss rate, failing to meet the stringent stability requirements of high-security services. Summary of the Invention

[0005] To address the shortcomings of traditional backup mechanisms, such as high resource consumption, limited main link transmission quality, and slow response to dynamic quality indicators like latency, jitter, and packet loss rate, which fail to meet the stability requirements of high-security services, this invention aims to provide a method, system, and medium for link backup and composite load balancing based on service packet identification. It combines the internal markings of each service packet with the link status results of each link to generate packet forwarding backup and composite load balancing strategies. This achieves deep coupling between deep service awareness (internal packet marking) and multi-link status awareness, enabling predictive and service-policy-based condition-triggered backup mechanisms. This improves the reliability and real-time performance of critical service packet transmission while maximizing the utilization of overall link resources.

[0006] The above-mentioned technical objective of the present invention is achieved through the following technical solution: This solution provides a method for link backup and composite load balancing based on service message identification, the method including: At the sending end: Receive raw service data packets, perform message identification and classification on the raw service data packets, and mark each service message in the raw service data packets with an internal tag according to the message identification and classification results; Monitor the real-time performance parameters of 5G links and self-organizing network links, and input the real-time performance parameters into the link comprehensive quality assessment model to calculate the link status result of each link; Based on the internal tags of each service message and the link status results of each link, a forwarding or backup transmission strategy is generated for each service type message. Based on the backup and composite load transmission and forwarding strategy, the original business data packets are adjusted before being used as transmission data packets for backup and composite load forwarding. At the receiving end: The system receives and parses the transmitted data packets, performs reception restoration processing on the parsed data packets, and then submits them to the upper-layer protocol stack or application.

[0007] A further optimization is that the marking method for the internal markers includes: Differential service code points and 5-tuple information are identified from the current service message; the 5-tuple information includes: source IP, destination IP, source port, destination port, and transport layer protocol; wherein, the source IP and destination IP represent the source and destination of the current service message; the transport layer protocol represents the importance of the current service message; Based on the quintuple information and differential service code points, the transmission requirements of the current service message are marked; the transmission requirements include latency requirements, bandwidth requirements, and packet loss rate requirements.

[0008] A further optimization scheme is that the comprehensive quality assessment model for both the 5G link and the self-organizing network link is related to the following parameters: One-way delay from sender to receiver, round-trip delay between sender and receiver, packet loss rate, delay jitter, and signal status.

[0009] A further optimization scheme is that the method for generating the forwarding backup and composite load balancing strategy includes: If the latency requirement of the current service message is lower than the preset latency threshold, a single-link forwarding strategy is generated. The single-link forwarding strategy includes selecting a low-latency link from 5G links and self-organizing network links as the transmission link. If the bandwidth requirement of the current service message is higher than the preset bandwidth threshold, a link aggregation forwarding strategy is generated. The link aggregation forwarding strategy is the aggregation transmission of 5G links and self-organizing network links. If the packet loss rate requirement for the current business messages is lower than the preset packet loss rate threshold, a backup forwarding strategy will be generated.

[0010] A further optimized solution is that the method for sending adjustment processing includes: Based on the standard structure of the original business data packet, a custom field area is added between the transport layer protocol header and the data payload; the custom field area is used to implement specific transmission enhancement functions. The custom field area includes at least the following fields arranged in sequence: link ID field, compression identifier field, message seq field, and data length field.

[0011] A further optimized solution is that the standard structure includes an Ethernet header, an IP header, a transport layer protocol header, and a data payload arranged in sequence.

[0012] A further optimization is that the link ID field is used to identify a specific physical transmission link or logical transmission link; The compression identifier field is used to indicate the compression algorithm under test used in the data payload, the compression state, or the length information before and after compression. The message seq field is used at the receiving end to perform message reassembly in order, detect packet loss, deduplication, and data length field; The data length field is used to identify the length of the data payload.

[0013] A further optimized solution is that the method for receiving and restoring data includes: Parse and execute custom field regions in the transmitted data packets; Remove custom fields from the transmitted data packets.

[0014] This solution also provides a system for link backup and composite load balancing based on service packet identification, used to implement the above-mentioned method for link backup and composite load balancing based on service packet identification. The system includes: The sending end is used to receive raw service data packets, perform message identification and classification on the raw service data packets, and mark each service message in the raw service data packets with an internal tag according to the message identification and classification results. The transmitting end is also used to monitor the real-time performance parameters of 5G links and self-organizing network links, and input the real-time performance parameters into the link comprehensive quality assessment model to calculate the link status result of each link. The sending end is also used to generate forwarding backup and composite load balancing strategies for each message based on the internal tags of each service message and the link status results of each link. The sending end is also used to perform sending adjustment processing on the original service data packets based on forwarding backup and composite load strategies, and then use them as transmission data packets for backup forwarding. The receiving end is used to receive and parse the transmitted data packets, and after performing reception restoration processing on the parsed transmitted data packets, submit them to the upper-layer protocol stack or application.

[0015] This solution also provides a computer-readable medium having a computer program stored thereon, which, when executed by a processor, can implement the link backup and composite load balancing method based on service message identification as described above.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention provides a method, system, and medium for link backup and composite load balancing based on service message identification. It generates message forwarding backup and composite load balancing strategies by combining the internal tags of each service message and the link status results of each link. It achieves deep coupling between deep service awareness (internal tagging of messages) and multi-link status awareness (link status results), realizing predictive and service policy-based condition-triggered backup mechanisms. It improves the reliability and real-time performance of critical service message transmission while maximizing the utilization of overall link resources.

[0017] 2. This invention provides a method, system, and medium for link backup and composite load based on service message identification; 5G links and self-organizing network links, two networks with very different characteristics, are not simply parallel relationships, but form an organic whole under intelligent scheduling, covering the complete chain from identification and scheduling at the sending end to processing at the receiving end, emphasizing end-to-end collaboration, and achieving seamless transmission of highly guaranteed services; 3. This invention provides a method, system, and medium for link backup and composite load balancing based on service message identification; it configures different transmission backup strategies for different instructions, link status results, and service message types; through the backup forwarding strategy, even in the event of a complete interruption of the main link, it can still deliver messages 100% through the backup link, meeting the safety requirements of power control and other applications; in scenarios where bandwidth is scarce in self-organizing networks such as emergency communication, it maximizes the use of limited bandwidth to carry more services (such as multi-channel video preview) while ensuring core services (such as command voice and location information); the backup strategy is no longer triggered waiting for a complete link interruption, but is based on dynamic strategy rules to achieve proactive and predictive triggering. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 This is a schematic diagram illustrating the principle of link backup and composite load balancing based on business message identification. Figure 2 This is a diagram illustrating the process of identifying business messages. Figure 3A schematic diagram of the link state identification and forwarding process; Figure 4 This is a schematic diagram illustrating the principle of sending adjustment processing. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0020] Traditional backup mechanisms are resource-intensive, have limited main link transmission quality, and are slow to respond to dynamic quality indicators such as latency, jitter, and packet loss rate, failing to meet the stability requirements of high-security services. In view of this, this solution provides the following embodiments to solve the above-mentioned technical problems.

[0021] Example 1

[0022] This embodiment provides a method for link backup and composite load balancing based on service packet identification, such as... Figure 1 As shown, the method includes: At the sending end: After the service host sends out the original service data packet, the sending end receives the original service data packet, performs message identification and message classification on the original service data packet, and marks each service message in the original service data packet with an internal mark according to the message identification and message classification results. In this embodiment, message identification and classification mainly involve Deep Packet Inspection (DPI) and Flow Analysis (DFI). Based on a built-in service feature library, service messages are identified and classified (e.g., key feature values ​​DSCP, destination port, destination IP), and each service message is marked with an internal tag containing information such as service priority and urgency. Specifically, for example... Figure 2 As shown, the marking method for the internal markers includes: S11, Identify the differential service code point and 5-tuple information from the current service message; the 5-tuple information includes: source IP, destination IP, source port, destination port, and transport layer protocol; The Differential Service Code Point (DSCP) is a 6-bit field in the IP header used to identify the service class and priority of a data packet. It is the foundation of QoS (Quality of Service) and can directly and efficiently demonstrate the importance of a service. Network devices (such as routers and switches) can directly prioritize and schedule data packets based on the DSCP value. The forwarding module can first read this field to quickly distinguish between high-guarantee services (such as EF accelerated forwarding, corresponding to voice and control commands) and ordinary services (such as BE best-effort, corresponding to web browsing and file downloads).

[0023] A 5-tuple uniquely identifies a network session or data stream. The source and destination ports are the most classic criteria for service identification, with many applications using well-known ports (such as HTTP-80, HTTPS-443, DNS-53, SSH-22, and RTP / RTCP for audio and video). The destination port allows for quick deduction of the likely service type.

[0024] Source and destination IP addresses can identify the origin and destination of a service. For example, traffic from a specific control server's IP (such as 10.1.1.100) can be identified as a control command stream; traffic destined for a specific video server's IP can be identified as a video return stream. This is more accurate than relying solely on ports and can prevent misjudgments caused by uncommon ports. For transport layer protocols, the choice between UDP and TCP inherently reflects the characteristics of the service. UDP is often used for services with high real-time requirements and a tolerance for minor data loss (such as voice, video, and telemetry); TCP is used for services with high reliability requirements but low latency sensitivity (such as file transfer and signaling). This provides a basis for subsequent backup strategies.

[0025] S12, based on the five-tuple information and differential service code points, mark the transmission requirements of the current service message; the transmission requirements include latency requirements, bandwidth requirements, and packet loss rate requirements; The specific differential service code point can distinguish the service type of the current service message. If the current service message belongs to the high-guarantee service, the first packet loss rate requirement is marked for the current service message; if the current service message belongs to the ordinary service, the second packet loss rate requirement is marked for the current service message; among them, the first packet loss rate requirement is higher than the second packet loss rate requirement. The importance of the current service message is determined based on the transport layer protocol. If the transport layer protocol of the current service message is UDP, then the current service message is marked with the first packet loss rate requirement; if the transport layer protocol of the current service message is TCP, then the current service message is marked with the second packet loss rate requirement.

[0026] The transmitting end also monitors the real-time performance parameters of 5G links and self-organizing network links, and inputs the real-time performance parameters into the link comprehensive quality assessment model to calculate the link status result of each link; The comprehensive quality assessment model for both 5G links and ad hoc network links is related to the following parameters: One-way delay from sender to receiver, round-trip delay between sender and receiver, packet loss rate, delay jitter, and signal status.

[0027] The sending end also generates forwarding backup and composite load balancing strategies for each message based on the internal tags of each service message and the link status results of each link. Specifically, the forwarding backup and composite load balancing strategy includes: If the latency requirement of the current service message is lower than the preset latency threshold, a single-link forwarding strategy is generated. The single-link forwarding strategy includes selecting a low-latency link from 5G links and self-organizing network links as the transmission link. If the bandwidth requirement of the current service message is higher than the preset bandwidth threshold, a link aggregation forwarding strategy is generated. The link aggregation forwarding strategy is the aggregation transmission of 5G links and self-organizing network links. If the packet loss rate requirement for the current business messages is lower than the preset packet loss rate threshold, a backup forwarding strategy will be generated.

[0028] Based on the above principles, a link selection strategy for service message types is determined, flexibly handling different service transmission requirements in the link. On the basis of the aforementioned principles, through real-time link feedback from 5G and ad hoc network links, the service forwarding and link selection strategies are dynamically adjusted, thereby realizing a flexible network transmission method for complex and dynamically changing wireless networks.

[0029] Specifically, for service packets with the first packet loss rate requirement, if the packet loss rate requirement is lower than the preset packet loss rate threshold, a backup forwarding strategy is generated; for service packets with the second packet loss rate requirement, a link aggregation forwarding strategy can be configured. For scenarios where the business type is remote control and the message command is an emergency operation, a backup forwarding strategy is generated; For cases where the main link latency is greater than a preset threshold (50ms in this embodiment) and the service packet is a video frame I-frame, a backup forwarding strategy is generated. For the highest-level commands such as "emergency stop" and "switch on / off", a backup forwarding strategy is generated. Even if the main link is completely interrupted, the command can still be delivered 100% through the backup link, meeting the safety requirements of power control and other applications.

[0030] The sending end also performs forwarding adjustment processing on the original business data packets based on forwarding backup and composite load strategies, and then forwards them as backup data packets. The backup strategy in this embodiment no longer waits for a complete link interruption before being triggered. Instead, it is based on dynamic policy rules to achieve proactive and predictive triggering. The granularity of backup decisions is refined from the link to the business message stream, and further focused on key messages. For example, backup is initiated only for I-frames in the video stream or emergency stop commands in control instructions, rather than the entire video stream or all instructions.

[0031] like Figure 3As shown in Table 1, the system simultaneously activates a link state detection mechanism to monitor the real-time performance of both the ad hoc network link and the 5G link. Based on the link state, a forwarding table is generated to obtain metrics such as latency, packet loss rate, and signal strength, and the optimal forwarding path is calculated. Traffic is forwarded according to the generated link forwarding table for Terminal A.

[0032]

[0033] On the path from A to B: the 5G link has lower latency (15ms vs 18ms), lower packet loss rate (2% vs 2.5%), and a relatively better RSSI value (-87); the self-organizing network link has slightly higher latency and packet loss rate, but the signal strength is also -87; therefore, for data transmission from A to B, the 5G link has a greater advantage in terms of latency and packet loss rate.

[0034] On the path from A to C: 5G link: lower latency (25ms vs 28ms), but RSSI value of -113, which usually means the 5G signal is very weak, possibly in an area with edge coverage or severe interference. Ad Hoc Network link: slightly higher latency, but RSSI value of -55, very good signal strength, indicating excellent physical connection quality of the ad hoc network on this path. Conclusion: Although 5G has lower latency, the signal is extremely poor; although the ad hoc network has slightly higher latency, the signal is excellent and the packet loss rate is the same. In actual deployment, extremely poor signal may lead to a decrease or instability in actual throughput, so the system may tend to choose the ad hoc network.

[0035] Forwarding table generation logic:

[0036] Input: The "weights" in the above statistics table (which may be weighted scores calculated based on latency, packet loss rate, etc.).

[0037] Output: Terminal Link Forwarding Table A, which defines how different traffic should travel through different links.

[0038] The execution rules, as shown in Table 2 below, determine the flow of data packets: This solution utilizes both self-organizing networks and 5G to achieve network redundancy and complementarity. It makes dynamic decisions based on real-time link quality indicators (latency, packet loss rate, RSSI) and supports multi-dimensional forwarding strategies based on protocols (TCP / UDP), ports, and IP addresses, enabling more refined QoS guarantees.

[0039] like Figure 4 As shown, the method for sending adjustment processing includes: Based on the standard structure of the original business data packet, a custom field area is added between the transport layer protocol header and the data payload; the custom field area is used to implement specific transmission enhancement functions. The custom field area includes at least the following fields arranged in sequence: Link ID field, Compression Identifier field, Message Seq field, and Data Length field. The Link ID field and Compression Identifier field each occupy 1 byte, while the Message Seq field and Data Length field each occupy 1 byte.

[0040] The standard structure includes, in sequence, an Ethernet header, an IP header, a transport layer protocol header, and a data payload.

[0041] The Ethernet header is responsible for MAC address addressing at the data link layer, the IP header is responsible for IP address addressing and routing at the network layer, and the transport layer protocol header is responsible for port addressing and connection control (TCP) or simple transport (UDP) at the transport layer.

[0042] The Link ID field is used to identify a specific physical or logical transmission link, such as distinguishing between 5G Link 1, 5G Link 2, and self-organizing network links. The Link ID field is a key identifier in multi-link backup or load balancing scenarios.

[0043] The compression identifier field is used to indicate the compression algorithm under test used in the data payload, the compression state, or the length information before and after compression, so as to achieve bandwidth saving and adaptive decompression.

[0044] The message seq field is used at the receiving end to reassemble messages in order, detect packet loss, deduplicate, and includes a data length field; this is crucial for ensuring the orderliness of service flows in out-of-order, multi-path network environments.

[0045] The data length field is used to identify the length of the data payload, making it easier for the receiver to parse it quickly and accurately.

[0046] Before the original business data packets are sent, custom fields are inserted into the original data packets according to optimization strategies (such as multi-link scheduling, data compression, and sequence numbering) to form a new, enhanced data packet structure as the transmission data packet.

[0047] At the receiving end: The system receives and parses the transmitted data packets, performs reception restoration processing on the parsed data packets, and then submits them to the upper-layer protocol stack or application.

[0048] The method for receiving and restoring the data includes: Parse and execute custom field regions in the transmitted data packets; Remove custom fields from the transmitted data packets.

[0049] Upon receiving the transmitted data packet, the receiving end first parses the inserted custom fields to obtain information such as link, compression, and sequence number. Then, it performs corresponding processing based on this information (such as selecting the corresponding link processing module, decompression, and sorting the cache by sequence number). Finally, it strips these custom fields, restores the data payload, and submits it to the upper-layer standard network protocol stack (UDP / TCP) or application.

[0050] The deduplication, sorting, and reassembly mechanisms at the receiving end minimize disturbances to upper-layer applications (such as video decoders and control software), ensuring that network layer transmission failures are invisible to business users and meeting the continuity requirements of media live streaming, remote operation, and other applications.

[0051] In this solution, the two very different networks, 5G link and self-organizing network link, are not simply parallel, but form an organic whole under intelligent scheduling. It covers the entire chain from the identification and scheduling at the sending end to the processing at the receiving end, emphasizing end-to-end collaboration. This is the key to achieving seamless transmission of highly guaranteed services, and it is not just a single-point technology at the sending end, but a transmission and forwarding mechanism for the entire network.

[0052] Example 2 This embodiment provides a system for link backup and composite load balancing based on service packet identification, used to implement the method for link backup and composite load balancing based on service packet identification described in Embodiment 1. The system includes: The sending end is used to receive raw service data packets, perform message identification and classification on the raw service data packets, and mark each service message in the raw service data packets with an internal tag according to the message identification and classification results. The transmitting end is also used to monitor the real-time performance parameters of 5G links and self-organizing network links, and input the real-time performance parameters into the link comprehensive quality assessment model to calculate the link status result of each link. The sending end is also used to generate forwarding backup and composite load balancing strategies for each service message type based on the internal tags of each service message and the link status results of each link. The sending end is also used to perform sending adjustment processing on the original service data packets based on forwarding backup and composite load strategies, and then use them as transmission data packets for backup forwarding. The receiving end is used to receive and parse the transmitted data packets, and after performing reception restoration processing on the parsed transmitted data packets, submit them to the upper-layer protocol stack or application.

[0053] Example 3 This embodiment provides a computer-readable medium storing a computer program thereon. The computer program, when executed by a processor, can implement the link backup and composite load balancing method based on service message identification as described in Embodiment 1; specifically, it performs the following process: At the sending end: Receive raw service data packets, perform message identification and classification on the raw service data packets, and mark each service message in the raw service data packets with an internal tag according to the message identification and classification results; Monitor the real-time performance parameters of 5G links and self-organizing network links, and input the real-time performance parameters into the link comprehensive quality assessment model to calculate the link status result of each link; Based on the internal tags of each service message and the link status results of each link, a forwarding backup and composite load balancing strategy is generated for each message. Based on forwarding backup and composite load balancing strategies, the original business data packets are adjusted before being used as backup data packets for transmission. At the receiving end: The system receives and parses the transmitted data packets, performs reception restoration processing on the parsed data packets, and then submits them to the upper-layer protocol stack or application.

[0054] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for link backup and composite load balancing based on service message identification, characterized in that, The method includes: At the sending end: Receive raw service data packets, perform message identification and classification on the raw service data packets, and mark each service message in the raw service data packets with an internal tag according to the message identification and classification results; Monitor the real-time performance parameters of 5G links and self-organizing network links, and input the real-time performance parameters into the link comprehensive quality assessment model to calculate the link status result of each link; Based on the internal tags of each service message and the link status results of each link, a forwarding backup and composite load balancing strategy is generated for each message. Based on forwarding backup and composite load balancing strategies, the original business data packets are adjusted before being used as transmission data packets for backup and composite load balancing forwarding. At the receiving end: The system receives and parses the transmitted data packets, performs reception restoration processing on the parsed data packets, and then submits them to the upper-layer protocol stack or application.

2. The method for link backup and composite load balancing based on service message identification according to claim 1, characterized in that, The marking method for the internal markers includes: Differential service code points and 5-tuple information are identified from the current service message; the 5-tuple information includes: source IP, destination IP, source port, destination port, and transport layer protocol; wherein, the source IP and destination IP represent the source and destination of the current service message; the transport layer protocol represents the importance of the current service message; Based on the five-tuple information and differential service code points, the transmission requirements of the current service message are marked; the transmission requirements include latency requirements, bandwidth requirements, and packet loss rate requirements.

3. The method for link backup and composite load balancing based on service message identification according to claim 2, characterized in that, The comprehensive link quality assessment model includes: The comprehensive quality assessment model for both 5G links and ad hoc network links is related to the following parameters: One-way delay from sender to receiver, round-trip delay between sender and receiver, packet loss rate, delay jitter, and signal status.

4. The method for link backup and composite load balancing based on service message identification according to claim 3, characterized in that, The method for generating the forwarding backup and composite load balancing strategy includes: If the latency requirement of the current service message is lower than the preset latency threshold, a single-link forwarding strategy is generated. The single-link forwarding strategy includes selecting a low-latency link from 5G links and self-organizing network links as the transmission link. If the bandwidth requirement of the current service message is higher than the preset bandwidth threshold, a link aggregation forwarding strategy is generated. The link aggregation forwarding strategy is the aggregation transmission of 5G links and self-organizing network links. If the current packet loss rate requirement is lower than the preset packet loss rate threshold, a backup and composite load balancing forwarding strategy will be generated.

5. The method for link backup and composite load balancing based on service message identification according to claim 1, characterized in that, The method for sending adjustment processing includes: Based on the standard structure of the original business data packet, a custom field area is added between the transport layer protocol header and the data payload; the custom field area is used to implement specific transmission enhancement functions. The custom field area includes at least the following fields arranged in sequence: link ID field, compression identifier field, message seq field, and data length field.

6. The method for link backup and composite load balancing based on service message identification according to claim 5, characterized in that, The standard structure includes, in sequence, an Ethernet header, an IP header, a transport layer protocol header, and a data payload.

7. The method for link backup and composite load balancing based on service message identification according to claim 5, characterized in that, The link ID field is used to identify a specific physical or logical transmission link; The compression identifier field is used to indicate the compression algorithm under test used in the data payload, the compression state, or the length information before and after compression. The message seq field is used at the receiving end to perform message reassembly in order, detect packet loss, deduplication, and data length field; The data length field is used to identify the length of the data payload.

8. The method for link backup and composite load balancing based on service message identification according to claim 5, characterized in that, The method for receiving and restoring the data includes: Parse and execute custom field regions in the transmitted data packets; Remove custom fields from the transmitted data packets.

9. A system for link backup and composite load balancing based on service message identification, characterized in that, The system is used to implement the link backup and composite load balancing method based on service message identification as described in any one of claims 1-8, the system comprising: The sending end is used to receive raw service data packets, perform message identification and classification on the raw service data packets, and mark each service message in the raw service data packets with an internal tag according to the message identification and classification results. The transmitting end is also used to monitor the real-time performance parameters of 5G links and self-organizing network links, and input the real-time performance parameters into the link comprehensive quality assessment model to calculate the link status result of each link. The sending end is also used to generate forwarding backup and composite load balancing strategies for each service message type based on the internal tags of each service message and the link status results of each link. The sending end is also used to perform sending adjustment processing on the original service data packets based on forwarding backup and composite load strategies, and then use them as transmission data packets for backup forwarding. The receiving end is used to receive and parse the transmitted data packets, and after performing reception restoration processing on the parsed transmitted data packets, submit them to the upper-layer protocol stack or application.

10. A computer-readable medium having a computer program stored thereon, characterized in that, The computer program, when executed by a processor, can implement the link backup and composite load balancing method based on service message identification as described in any one of claims 1-8.

Images