A multi-client interlayer transparent intercommunication method and system

Abstract

This application discloses a method and system for transparent Layer 2 interoperability among multiple clients. The method includes: multiple clients sending registration messages to a controller; the controller establishing a data forwarding table; the source client encapsulating and copying the original Layer 2 data multiple times, and sending it concurrently through multiple network interfaces; the controller deduplicating and decapsulating the first arriving message copy, querying the forwarding table to determine the target client, and then re-encapsulating and forwarding it. The system includes the controller and clients for executing the method. This application achieves highly reliable Layer 2 interoperability through multi-path concurrent transmission and centralized reception, combined with adaptive redundancy control and a high-performance deduplication mechanism, effectively saving network bandwidth resources and solving the performance bottleneck of data processing.

Description

Technical Field

[0001] This application relates to the field of network communication technology, and more specifically, to a method and system for constructing a virtual Layer 2 network over multiple wide area networks to enable transparent data communication between client-attached devices located in different geographical locations, and in particular, to a highly reliable communication method and system that supports multi-path redundant forwarding and adaptive redundancy optimization. Background Technology

[0002] In modern industrial automation, the Internet of Things (IoT), and distributed system applications, it is often necessary to interconnect service devices (such as industrial cameras, programmable logic controllers (PLCs), and sensors) deployed in different geographical locations and belonging to different local area networks (LANs), enabling them to communicate transparently at Layer 2 (i.e., based on MAC addresses) as if they were within the same LAN. However, traditional network architectures struggle to directly extend Layer 2 networks across wide area networks (WANs) or the public internet, which introduces significant complexity to the deployment of upper-layer services.

[0003] To address this issue, existing technologies typically employ VPN tunnels, VPLS, and other techniques to construct virtual Layer 2 networks. However, these solutions struggle to guarantee communication reliability in unstable network environments, such as 5G, radio stations, or the public internet. A single network link can experience jitter, congestion, or even outages, leading to the loss or delay of critical data, severely impacting real-time demanding scenarios such as industrial control.

[0004] To improve reliability, some solutions introduce static multipath redundancy mechanisms, which simply copy data multiple times and send it simultaneously through multiple network paths. While this "multiple send, multiple receive" mode can mitigate single-point network failures to some extent, it also introduces new problems. Maintaining full redundancy regardless of network conditions leads to significant bandwidth waste. The receiving end needs to process massive amounts of duplicate data packets, posing a huge challenge to the CPU processing capabilities of the central controller or receiving client. This can easily create performance bottlenecks under high concurrency traffic, limiting the overall capacity and scalability of the system.

[0005] Therefore, how to intelligently manage redundant resources and efficiently process redundant data while ensuring high reliability, in order to solve the reliability, bandwidth efficiency and performance bottlenecks of Layer 2 cross-domain interconnection in existing technologies, is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] This application provides a method and system for transparent Layer 2 communication between multiple clients, which solves the problems of low reliability, wasted bandwidth resources, and data processing performance bottlenecks in the prior art when performing Layer 2 communication across wide area networks.

[0007] To achieve the above objectives, this application provides a method for transparent Layer 2 interoperability between multiple clients, including: Multiple multi-transmitter selective receiver clients send registration messages to a multi-transmitter selective receiver controller. The registration message includes the client identifier DN of the multi-transmitter selective receiver client and the MAC address information of the service device connected to the southbound interface of the multi-transmitter selective receiver client. The multi-transmitter selective receiver controller receives and parses the registration message, establishes and maintains a data forwarding table, and the data forwarding table records the mapping relationship between the MAC address of each service device and the DN of the multi-transmitter selective receiver client to which the service device belongs and the currently active northbound network interface information of the multi-transmitter selective receiver client. The source multi-transmission selective reception client receives the raw Layer 2 data sent by the source service device, encapsulates the raw Layer 2 data, and the encapsulated message contains the DN and frame sequence number FN of the source multi-transmission selective reception client; the source multi-transmission selective reception client copies the encapsulated message multiple times and sends it concurrently to the multi-transmission selective reception controller through at least two northbound network interfaces; The multi-transmission selective receiver decapsulates the first arriving encapsulated packet copy and discards subsequent duplicate packet copies based on the DN and FN. The multi-transmission selective receiver parses the target MAC address from the original Layer 2 data, queries the data forwarding table, and determines the DN and network interface information of the target multi-transmission selective receiver client corresponding to the target MAC address. The multi-transmission selective receiver uses the network interface information of the target multi-transmission selective receiver client to recapsulate the original Layer 2 data and sends the recapsulated packet to the target multi-transmission selective receiver client. The target multi-transmitter selective receiver client receives the re-encapsulated message, decapsulates it, and sends the restored original Layer 2 data to the target service device through its southbound interface.

[0008] Another aspect of this application provides a Layer 2 transparent interoperability system between multiple clients, including a multi-transmitter selective receiver controller and at least two multi-transmitter selective receiver clients; Each of the aforementioned multi-transmitter / receiver clients includes: Southbound interface, used for connecting to service devices; At least two northbound network interfaces are provided for communicating with the multi-transmit / receive controller; The registration module is configured to generate and send a registration message containing its own client identifier DN and the MAC address information of the service device. The encapsulation and transmission module is configured to encapsulate the raw Layer 2 data received from the southbound interface, and copy the encapsulated message multiple times and send it concurrently to the multi-transmission and selective reception controller through the at least two northbound network interfaces. The multi-transmission selective receiver controller includes: The registration receiving module is configured to receive and parse the registration message. The forwarding table management module is configured to create and maintain a data forwarding table based on the registration message; The relay forwarding module is configured to deduplicatize and decapsulate the received encapsulated messages, query the data forwarding table to determine the target multi-send / receive client, recapsulate the original Layer 2 data, and forward it.

[0009] Compared with the prior art, the technical solution provided in this application has the following beneficial effects: 1. By employing a dual encapsulation / decapsulation mechanism between the multi-transmitter selective receiver client and the multi-transmitter selective receiver controller, a virtual, unified Layer 2 network was successfully constructed over multiple heterogeneous wide area networks. This allows service devices in different physical locations to communicate directly via MAC addresses as if they were in the same local area network, simplifying the deployment and migration of upper-layer services.

[0010] 2. By utilizing multi-path concurrent transmission and selective reception mechanisms, it effectively combats jitter, congestion, and even interruptions in a single network path, ensuring zero packet loss and high availability of data transmission. It is particularly suitable for industrial control and other scenarios with stringent requirements for communication stability.

[0011] 3. By introducing an adaptive redundancy control algorithm, the system can dynamically adjust the number of redundant replicas based on real-time network quality (arrival dispersion), significantly saving network bandwidth and system processing resources while ensuring reliability. Simultaneously, the adoption of a sliding window deduplication engine based on FPGA and bitmap indexing increases data processing throughput by several orders of magnitude, significantly reduces CPU load, and enables the system to support larger-scale concurrent service flows, thus resolving the performance bottleneck of static redundancy schemes. Attached Figure Description

[0012] 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.

[0013] Figure 1 This is a structural block diagram of a multi-client two-layer transparent interconnection system provided in an embodiment of the present invention.

[0014] Figure 2 This is a flowchart of a multi-client transparent interoperability method provided in an embodiment of the present invention.

[0015] Figure 3 This is a block diagram of the internal structure of the multi-transmitter selective receiver controller provided in an embodiment of the present invention.

[0016] Figure 4 This is an overall system block diagram provided in another embodiment of the present invention.

[0017] Figure 5 This is a schematic diagram of data transmission between two client service devices provided in another embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0019] Example 1

[0020] This embodiment provides a method for transparent Layer 2 interoperability between multiple clients. This method constructs a star topology centered on a multi-transmitter selector controller, logically simulating the function of a "cloud switch," and achieving transparent Layer 2 data forwarding across a wide area network. For example... Figure 2 The diagram shown is a flowchart of a method provided in an embodiment of the present invention. The method may specifically include the following steps: Step S201: Client registration.

[0021] In a multi-client interconnection system, there are multiple multi-transmitter selective receiver clients and a centralized multi-transmitter selective receiver controller. When the system initializes or a new client joins, each multi-transmitter selective receiver client proactively sends a registration message to the controller. This registration message is a specially formatted data packet containing at least two key pieces of information: 1. Client Identifier (DN): This is a globally unique identifier used to uniquely identify a multi-transmitter / receiver client throughout the system. In this embodiment, the DN can be a 32-bit unsigned integer, serving as a data stream ID.

[0022] 2. Service Device Information: This section records the network identifiers, primarily MAC addresses, of all service devices directly connected to the southbound interface of this multi-transmitter / receiver client. Service devices can be any equipment requiring Layer 2 communication, such as industrial cameras, PLCs, and sensors.

[0023] Step S202: Establish and maintain the data forwarding table.

[0024] The multi-transmitter / selective receiver controller continuously monitors its northbound network interface, receiving registration messages from various clients. The controller's internal registration and receiving module parses the received registration messages, extracting the DN and the MAC address list of the service device. Subsequently, the forwarding table management module establishes and dynamically maintains a global data forwarding table based on the parsed information. This data forwarding table is the core basis for implementing Layer 2 forwarding decisions. In a specific implementation, the structure of this data forwarding table is shown in the table below:

[0025]

[0026] It records the following mapping relationships: • Map the MAC address of each service device to the DN of the multi-transmitter selector client to which it belongs.

[0027] • Map each DN to the network information of all currently active northbound network interfaces of the client, such as the client's IP address and UDP port number in different networks, as well as the physical northbound port number on the controller side that received the client's registration message.

[0028] Using this table, the controller can quickly locate the target client and all available network paths to reach that client based on any target MAC address.

[0029] Step S203: Source-end encapsulation and multi-path concurrent transmission.

[0030] When a source service device (e.g., service device 1-1) needs to send data to a target service device (e.g., service device n-1) located under a different client, it generates a standard Layer 2 Ethernet data frame (i.e., raw Layer 2 data) and sends it to the source multi-transmitter selective receiver client directly connected to it. Upon receiving the raw Layer 2 data, the encapsulation and transmission module of the source multi-transmitter selective receiver client performs encapsulation and copy transmission operations: 1. In order to be transmitted over a wide area network and correctly processed by the controller, the client encapsulates the raw Layer 2 data. Specifically, the encapsulation process includes adding an inner header and an outer header to the raw Layer 2 data.

[0031] • Internal header: Contains information used for stream identification and deduplication within the multiplexer system. The core fields are the source client's DN (e.g., 32-bit) and frame sequence number FN (e.g., 16-bit). FN is a sequence number that monotonically increases with the data stream of that DN.

[0032] • External header: Contains information used for routing in the underlying wide area network, such as standard IP and UDP headers, with the destination address pointing to the network address of the multiplexer controller.

[0033] 2. After encapsulation, the source client copies the encapsulated message multiple times. The number of copies depends on the number of available northbound network interfaces. For example, if the client is simultaneously connected to four different networks (which could be 5G, Wi-Fi, wired Ethernet, leased lines, or any network capable of transmitting IP packets), it will generate four identical copies of the encapsulated message and concurrently send them to the multi-transmission and selection controller through these four northbound network interfaces. This "multi-transmission" mechanism is the foundation for achieving high reliability.

[0034] Step S204, controller relay forwarding.

[0035] The multi-transmitter / selective receiver controller receives encapsulated packet copies from different network paths through its multiple northbound network interfaces. The relay forwarding module within the controller performs the following series of processes: 1. High-performance deduplication: Due to multipath transmission by the source client, the controller receives multiple copy of messages with the same DN and FN from different networks at different times. To avoid duplicate processing, the controller must efficiently discard redundant copies and process only the first one to arrive. To achieve this, this embodiment employs a bitmap-indexed sliding window deduplication engine.

[0036] • Structure: The engine maintains an independent bitmap sliding window for each data stream identified by a DN. In a preferred embodiment, the size of the bitmap is 2. 16 The bits correspond exactly to all possible values ​​of the 16-bit FN.

[0037] • Operation: When a message with a DN and FN is received, the engine calculates the index by performing a modulo operation on the frame sequence number FN. For example, index = FN mod W, where W is the window size, and here W = 2. 16 Then, the state of the bit pointed to by the index in the bitmap is checked. If the state is 0, it means that this is the first arriving copy of the FN message, and the engine immediately sets the bit to 1 and passes the message to the subsequent module for processing; if the state is 1, it means that the FN message has been processed, the current message is a redundant copy, and the engine directly discards it. The time complexity of this operation is O(1), which can meet the requirements of extremely high line-speed processing. In order to prevent the aging problem caused by FN flipping, the structure can also maintain a window header pointer. When the FN grows beyond a preset threshold (e.g., 1024), the bitmap state that lags behind the pointer is automatically cleared.

[0038] Furthermore, the bitmap-indexed sliding window deduplication engine employs a two-layer storage structure. The first layer is the bitmap, used to quickly determine whether a message is a duplicate copy; the second layer is a timestamped circular buffer. When the bitmap determines that the current message is the first arriving copy, the system writes the message's metadata and the current arrival timestamp into the corresponding position in the circular buffer; when the bitmap determines that the current message is a duplicate copy, the system only extracts the current timestamp and updates it to the record corresponding to the message in the circular buffer for subsequent calculation of the arrival dispersion. Since there is no need to store the data body of the duplicate message, storage space is saved.

[0039] 2. Decapsulation and Table Lookup: For the first arriving packet that passes the deduplication check, the controller decapsulates it to restore the original Layer 2 data frame. Then, the controller parses the destination MAC address from this original data frame.

[0040] 3. Forwarding Decision: The controller uses the target MAC address to query the data forwarding table established in step S202. The query result will clearly indicate the DN of the target multi-transmitter client to which the target MAC address belongs, as well as information on all active network interfaces reaching the target client.

[0041] 4. Re-encapsulation and Forwarding: The controller uses the network interface information of the target client obtained from the query to perform a second encapsulation of the original Layer 2 data (the target address in the outer header is changed to the target client's address), and then sends the re-encapsulated packet to the target multi-sender / receiver client. The controller can also utilize multiple active links of the target client for redundant transmission to ensure the reliability of the link between the controller and the target client.

[0042] Step S205: Decapsulate the target end.

[0043] After receiving the message forwarded by the controller, the target multi-transmitter / receiver client performs a decapsulation operation to restore the original Layer 2 data frame. Finally, it sends the data frame to the final target service device through its southbound interface.

[0044] Thus, a transparent Layer 2 data transmission across multiple wide area networks is completed. From the perspective of the source and destination service devices, they communicate as if they were on a local area network, completely unaware of the complex multi-path forwarding and encapsulation / decapsulation processes involved. The entire process can be found in [reference needed]. Figure 5 The diagram shows a data transmission diagram between two client service devices.

[0045] Furthermore, to further optimize system performance and bandwidth utilization, this method introduces an adaptive redundancy control mechanism. This mechanism is implemented in the relay forwarding module of the controller. The controller not only processes packets but also records a precise arrival timestamp for each replica with the same DN and FN. The controller calculates the arrival dispersion of the packet replica group based on these timestamps. For a packet with frame sequence number FN=k, its arrival dispersion is represented as δ. k The value is calculated by obtaining the arrival timestamps T1, T2, ..., Tm of all m replicas of the message, and using the difference between the maximum and minimum arrival timestamps as the arrival dispersion. In a preferred embodiment, the above calculation logic is implemented using the following formula: δ k =max(T1,T2,...,Tm)−min(T1,T2,...,Tm) Where, δ k Ti represents the arrival dispersion of the message with frame number k, and Ti represents the arrival timestamp of the i-th replica.

[0046] The controller maintains a moving average dispersion δˉ and compares it with a preset stability threshold Tstable (e.g., a critical value reflecting good network synchronization, which can be set to 2ms).

[0047] • Reduced Redundancy: When the moving average dispersion δˉ is below the stability threshold Tstable, and no packet loss is detected within a certain period, it indicates that all network paths are very stable and synchronized. At this point, the controller determines that full redundant transmission is unnecessary and sends a control command (e.g., a redundancy control message) to the source client, instructing it to reduce the number of concurrently transmitted packet replicas R (e.g., from R=4 to R=2), selecting only the two paths with the lowest historical latency for transmission. This effectively saves network bandwidth.

[0048] • Add / Restore Redundancy: When the moving average dispersion δˉ exceeds the stability threshold Tstable, or when the controller fails to receive a desired FN message within a preset time window (determined as packet loss), it indicates network jitter, congestion, or a fault. In this case, the controller immediately sends a command instructing the source client to add or restore full redundancy transmission mode (e.g., R=4) to ensure the highest reliability of data transmission.

[0049] Through this closed-loop adaptive control, this method achieves intelligent and refined management of network resources while ensuring zero packet loss.

[0050] This application achieves synergistic optimization through adaptive redundancy control and bitmap deduplication mechanism: when the network conditions are good, the controller instructs the client to reduce the transmission redundancy, which not only saves network bandwidth, but more importantly, directly reduces the probability of hash collisions and the bitmap writing frequency at the controller receiver, enabling the deduplication engine to support more concurrent service flows (number of DNs), thereby significantly improving the overall throughput capacity of the system.

[0051] Example 2

[0052] This embodiment provides a Layer 2 transparent interoperability system between multiple clients, which is the physical carrier of the method described in Embodiment 1. Please refer to... Figure 1 , Figure 3 and Figure 4 . Figure 4 The system's overall deployment architecture was demonstrated, including a multi-transmitter selective receiver controller, multiple (up to 4) heterogeneous networks, and multiple multi-transmitter selective receiver clients, with each client connected to multiple service devices. Figure 1 It is a more abstract structural diagram of the system, showing the core components and their connections.

[0053] like Figure 1 As shown, the system includes a multi-transmitter selective receiver controller 101 and at least two multi-transmitter selective receiver clients 102.

[0054] Each multi-transmitter / receiver client 102 is physically an embedded device, and its internal structure includes: • Southbound interface: Used for physical connection with local service devices, usually a standard Ethernet interface.

[0055] • Northbound network interfaces: at least two, used to connect to different wide area networks and communicate with the multi-transmitter selector controller 101.

[0056] • Processor and memory: Run the client software logic.

[0057] • Registration module: This software module is responsible for generating a registration message containing its own client identifier (DN) and the MAC address information of the connected service devices at startup, and sending it to the controller 101 through the northbound network interface.

[0058] • Encapsulation and transmission module: A software or hardware acceleration module responsible for encapsulating the raw Layer 2 data received from the southbound interface, adding internal and external headers, and then copying the encapsulated message multiple times and sending it concurrently through all available northbound network interfaces.

[0059] The multi-transmitter / receiver controller 101 is the core of the system, and physically it can be a high-performance server or a dedicated network device. Its internal structure includes: • Multiple northbound network interfaces: used to receive data sent by clients from different networks.

[0060] • Southbound interface (optional): Used to connect to the service equipment or local network on the central side.

[0061] • Processor and memory: Run the core logic of the controller.

[0062] • Registration Receiving Module: A software module responsible for receiving and parsing registration messages sent by the client.

[0063] Forwarding Table Management Module: This software module is responsible for creating and maintaining a global forwarding table based on registration information. This module requires memory (such as DRAM) to store the forwarding table.

[0064] • Relay forwarding module: This is the core of the controller's data plane, responsible for deduplicating, decapsulating, looking up tables, recapsulating, and forwarding received encapsulated messages.

[0065] To handle high-concurrency data streams, the multi-transmission selective receiver 101 in this embodiment preferably employs a collaborative processing architecture of System-on-Chip (SoC) and Field-Programmable Gate Array (FPGA). For example... Figure 3 The diagram shown is a block diagram of the internal structure of the controller 101.

[0066] • FPGA 301: As a hardware coprocessor, it is configured to perform high-speed data plane tasks. The controller's northbound network interface is directly connected to the FPGA 301 via the PHY chip. The FPGA 301 internally implements a parallel parsing engine, capable of performing CRC checks, header parsing (extracting DN and FN), high-precision timestamp acquisition, and most critically, bitmap-based deduplication operations at line speed.

[0067] • SoC 302: As the main control CPU, it is configured to perform complex control plane tasks. It receives metadata (such as non-duplicate messages and timestamp information) processed by FPGA 301, runs adaptive redundancy algorithms, makes forwarding decisions, and manages the data forwarding table.

[0068] In this hardware-software co-working architecture, repetitive and high-intensity tasks in the data plane are efficiently completed by the hardware FPGA 301, while complex decision-making and management tasks in the control plane are flexibly handled by the software SoC 302, achieving the best balance between performance and flexibility.

[0069] Furthermore, to support the advanced functions in Embodiment 1, the functional modules of controller 101 are specifically implemented as follows: The deduplication function in the relay forwarding module is specifically implemented by a bitmap-indexed sliding window deduplication engine 303 deployed in FPGA 301. This engine maintains an independent bitmap in the FPGA's on-chip memory (such as SRAM) for each data stream (identified by DN), and uses the message's frame sequence number FN as an index for status lookup and update, thereby determining whether a message is duplicated in O(1) time complexity.

[0070] The controller 101 also includes an adaptive redundancy control module 304, which typically runs on the SoC 302. It obtains the precise arrival timestamps of all copies of the same message (with the same DN and FN) from the FPGA 301, calculates the arrival dispersion, and compares it with a preset stability threshold. Based on the comparison result, the module generates corresponding control commands and sends them to the source client via the network, instructing it to dynamically adjust the number of copies sent.

[0071] To ensure the stable operation of the entire system, this embodiment also includes some necessary auxiliary modules. The system requires a power supply module to power the controller and all clients. The system also provides a management configuration interface (e.g., a web or command-line interface) for configuring the controller's IP address, port number, stability threshold, and other operating parameters. Specifically, to ensure the accuracy of the timing difference algorithm in adaptive redundancy control, a clock synchronization mechanism, such as Network Time Protocol (NTP), needs to be deployed between all system nodes (the controller and all clients) to ensure that the timestamp base of all devices is consistent.

[0072] In summary, this application solves the problem of Layer 2 transparent interoperability between geographically dispersed devices by constructing a centralized forwarding system with intelligent redundancy control and high-performance deduplication capabilities. It also significantly improves communication reliability, network efficiency, and system processing performance, making it suitable for various distributed application scenarios with stringent network quality requirements.

[0073] It should be noted that the specific parameters given in the above embodiments, such as DN being 32 bits, FN being 16 bits, and the stability threshold Tstable being 2ms, are merely exemplary preferred configurations. Those skilled in the art can adjust these parameters according to specific application scenarios, network environments, and performance requirements, and such adjustments will not depart from the scope disclosed in this application. Similarly, the term "module" can be a software unit, a hardware circuit (such as a logic unit in an FPGA or ASIC), or a combination of both, and its specific implementation should not be considered a limitation of this application. Those skilled in the art who employ equivalent structures or steps to implement this application should also fall within the protection scope of this application.

Claims

1. A method for transparent Layer 2 interoperability between multiple clients, characterized in that, include: Multiple multi-transmitter selective receiver clients send registration messages to a multi-transmitter selective receiver controller. The registration message contains the client identifier DN of the multi-transmitter selective receiver client and the MAC address information of the service device connected to the southbound interface. The multi-transmitter selective receiver controller receives the registration message and establishes a data forwarding table. The data forwarding table records the mapping relationship between the MAC address of the service device and the DN of the multi-transmitter selective receiver client and the currently active northbound network interface information. The source multi-transmitter selective receiver client receives the raw Layer 2 data, encapsulates the raw Layer 2 data, and encapsulates the encapsulated message so that the source multi-transmitter selective receiver client contains the DN and frame sequence number FN of the source multi-transmitter selective receiver client; the source multi-transmitter selective receiver client copies the encapsulated message multiple times and sends it concurrently to the multi-transmitter selective receiver controller through at least two northbound network interfaces. The multi-transmission selective receiver decapsulates the first arriving encapsulated packet copy and discards subsequent duplicate packet copies based on the DN and FN. The multi-transmission selective receiver parses the target MAC address in the original Layer 2 data, queries the data forwarding table, and determines the DN and network interface information of the target multi-transmission selective receiver client corresponding to the target MAC address. The multi-transmission selective receiver uses the network interface information to recapsulate the original Layer 2 data and sends the packet to the target multi-transmission selective receiver client, which decapsulates the data and sends the restored original Layer 2 data to the target service device through the southbound interface.

2. The method according to claim 1, characterized in that, Also includes: The multi-transmitter selector controller records the arrival timestamp for each replica in a group of encapsulated message replicas with the same DN and FN; The multiple-transmission selective receiver calculates the arrival dispersion of the encapsulated message replica group based on the arrival timestamp; When the dispersion is lower than a preset stability threshold, the multiple-transmission selective receiver controller sends a control instruction to the source multiple-transmission selective receiver client, which instructs the source multiple-transmission selective receiver client to reduce the number of message copies it sends concurrently. When the arrival dispersion is higher than the stability threshold or no message with a specific FN is received within a preset time window, the multiple-transmission selective receiver controller instructs the source multiple-transmission selective receiver client to increase or restore the number of message copies it concurrently sends.

3. The method according to claim 2, characterized in that, The reach of discreteness δ k The calculation method is as follows: obtain the arrival timestamps T1, T2, ..., Tk of all k replicas in the encapsulated message replica group, and use the difference between the maximum and minimum arrival timestamps as the arrival dispersion, i.e., δ. k =max(T1,T2,...,Tk)−min(T1,T2,...,Tk); The control command is a redundant control message, which specifies the combination or number of northbound network interfaces that the source multi-transmitter selective receiver client should use when subsequently sending data.

4. The method according to claim 1, characterized in that, The step of the multiple-transmitter selector controller discarding subsequently received duplicate message copies is specifically executed through a bitmap index-based sliding window structure. An independent bitmap sliding window is maintained for each data stream identified by a DN. The FN in the received message is used as an index to query and mark in the bitmap sliding window to determine whether the message is a duplicate message copy.

5. The method according to claim 4, characterized in that, The size of the bitmap sliding window is a preset value W; the index is calculated by performing a modulo operation on the frame sequence number FN, i.e., index = FN mod W; the specific process of querying and marking is as follows: check the state of the bit pointed to by the index in the bitmap; if the state of the bit is 0, then set the state to 1 and process the current message, which is determined to be a non-duplicate message; if the state of the bit is 1, then directly discard the current message, which is determined to be a duplicate message.

6. A layer-2 transparent interoperability system between multiple clients, characterized in that, Includes a multi-transmitter selector controller and at least two multi-transmitter selector clients; Each of the aforementioned multi-transmitter / receiver clients includes: Southbound interface, used for connecting to service devices; At least two northbound network interfaces are provided for communicating with the multi-transmit / receive controller; The registration module is configured to generate and send a registration message containing its own client identifier DN and the MAC address information of the service device. The encapsulation and transmission module is configured to encapsulate the raw Layer 2 data received from the southbound interface, and copy the encapsulated message multiple times and send it concurrently to the multi-transmission and selective reception controller through the at least two northbound network interfaces. The multi-transmission selective receiver controller includes: The registration receiving module is configured to receive and parse the registration message. The forwarding table management module is configured to create and maintain a data forwarding table based on the registration message; The relay forwarding module is configured to deduplicatize and decapsulate the received encapsulated messages, query the data forwarding table to determine the target multi-send / receive client, recapsulate the original Layer 2 data, and forward it.

7. The system according to claim 6, characterized in that, The multi-transmitter selector controller adopts a collaborative architecture of system-on-chip (SoC) and field-programmable gate array (FPGA); the FPGA is configured to perform high-speed data plane tasks, including message parsing, bitmap-based deduplication, and arrival timestamp acquisition; the SoC is configured to perform control plane tasks, including running adaptive redundancy algorithms and managing the data forwarding table.

8. The system according to claim 6, characterized in that, The multi-transmission selective receiver controller also includes an adaptive redundancy control module, which is configured to: calculate the arrival dispersion of message replica groups with the same DN and FN, compare the arrival dispersion with a preset stability threshold, and generate a control command based on the comparison result to instruct the multi-transmission selective receiver client to adjust the number of transmitted replicas.

9. The system according to claim 6, characterized in that, The deduplication function in the relay forwarding module is implemented by a sliding window deduplication engine based on bitmap index. The sliding window deduplication engine maintains a bitmap for each data stream and uses the frame sequence number FN of the packet as an index to perform status queries and updates in the bitmap to determine whether the packet is duplicated.

10. The method according to claim 1, characterized in that, The steps for the source multi-transmission selective receiver client to encapsulate the original Layer 2 data specifically include: adding an internal header and an external header to the outside of the original Layer 2 data; the internal header contains the DN and the FN; the external header contains the network address information of the multi-transmission selective receiver controller.

Images