Optical switching network system

By utilizing multi-stage wavelength switching devices and relay nodes in the optical switching network system, the problems of high energy consumption and difficult control of traditional electrical switching networks are solved, and low-energy consumption, high-bandwidth data transmission and highly scalable optical switching networks are realized.

CN120614540APending Publication Date: 2025-09-09HUAWEI TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410262342.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-03-06
Publication Date
2025-09-09

AI Technical Summary

Technical Problem

Frequent photoelectric conversion in traditional electrical switching networks increases energy consumption, and centralized control planes find it difficult to implement real-time, accurate, and efficient control strategies in large-scale networks.

Method used

An optical switching network system is adopted, and multi-level wavelength switching devices and relay nodes are used to realize optical signal transmission. The relay path is determined by the topological address and wavelength switching relationship. In combination with the transmission control information of the electrical switching device, a highly scalable and easy-to-deploy optical switching network is constructed.

Benefits of technology

It achieves low-energy, high-bandwidth data transmission, reduces network infrastructure deployment costs and energy consumption, and improves network scalability and deployment efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120614540A_ABST
    Figure CN120614540A_ABST
Patent Text Reader

Abstract

The invention provides an optical switching network system, and belongs to the technical field of optical communication. The optical switching network system comprises a plurality of nodes and at least one stage of wavelength switching device connected with each node, and the first stage of wavelength switching device connected with each node is connected with each first stage of wavelength switching device located in the same stage of group. The ith-level wavelength switching device connected with each node is connected with at least one ith-level wavelength switching device in other groups in the same-level group, the nodes connected with the different-level wavelength switching devices connected with each node are the same, and under the condition that each node is not in direct communication with a destination node, the ith-level wavelength switching device is connected with at least one ith-level wavelength switching device in the other groups in the same-level group. The method comprises the following steps: determining a communication direct relay node in a plurality of nodes, sending a first optical signal to the relay node through a wavelength switching device on a communication direct link so as to send data to a destination node through the relay node, and outputting the received optical signal by each wavelength switching device through a wavelength switching function. The optical switching network system is easy to expand.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the field of optical communication technology, and in particular to an optical switching network system. Background Art

[0002] With the development of cloud computing, big data, and the Internet of Things, network traffic demand has increased significantly. Real-time, large-scale data transmission requires communication networks to provide higher bandwidth and lower latency. In traditional electrical switching networks, optical fiber between electrical switches can provide higher transmission bandwidth and lower latency. However, the communication process in traditional electrical switching networks requires frequent optical-to-electrical conversion, which increases energy consumption and poses serious energy challenges to electrical switching networks. To address this challenge, optical switching networks have emerged, as light can provide higher transmission bandwidth, lower energy consumption, lower network overhead, and higher density parallelism. The transmitted optical signals have the characteristics of low interference, low crosstalk, and low parasitic effects.

[0003] Currently, in optical switching networks, a centralized control plane collects communication requests from each node, performs global arbitration for communication requests, and configures the status of optical components such as optical switches based on the arbitration results. However, the overall latency of the centralized control plane, including traffic sensing, arbitrating communication requests, issuing configurations, and completing optical component configuration, is high. This makes it difficult to implement real-time, accurate, and efficient control strategies in large-scale networks. Summary of the Invention

[0004] The present disclosure provides an optical switching network system with high scalability and easy deployment. The technical solution is as follows:

[0005] The present disclosure provides an optical switching network system, which includes multiple nodes and at least one level wavelength switching device connected to each node; the first-level wavelength switching device connected to each node is connected to each first-level wavelength switching device located in the same-level group; in the case of multiple levels of wavelength switching devices, the i-th level wavelength switching device connected to each node is connected to at least one i-th level wavelength switching device in each other group located in the same-level group, the nodes connected to different-level wavelength switching devices connected to each node are the same, the i-1-th level group belongs to the i-th level group, and i is greater than or equal to 2; each node is used to, when communication with the destination node is not direct, determine a relay node with direct communication among the multiple nodes, and send a first optical signal to the relay node through the wavelength switching device on the direct communication link, so as to send the first data carried on the first optical signal to the destination node through the relay node; each wavelength switching device is used to output the received optical signal through the wavelength switching function.

[0006] In the solution disclosed herein, in an optical switching network system, nodes are connected to at least one level of wavelength switching devices. The first-level wavelength switching device is connected to each first-level wavelength switching device within the same level group, enabling data transmission between nodes within the first-level group. Furthermore, data transmission between nodes in other level groups and other levels of wavelength switching devices is also enabled. When direct communication is unavailable, data can be relayed through relay nodes. This allows any two nodes to transmit data. Furthermore, when adding new nodes to the optical switching network system, simply adding wavelength switching devices of the corresponding level and connecting the nodes to them can create a more advanced optical switching network system with high scalability and ease of deployment.

[0007] In an optional manner, each level of wavelength switching device is an arrayed waveguide grating router (AWGR); two wavelength switching devices having a connection relationship are connected through a first port and a second port, the first port and the second port correspond to the same wavelength switching relationship, the first port belongs to one of the two wavelength switching devices, and the second port belongs to the other wavelength switching device.

[0008] In the solution shown in the present disclosure, the wavelength switching device is implemented using AWGR. When two wavelength switching devices are connected, the ports connected to the two devices correspond to the same wavelength switching relationship, which facilitates determining the transmission path of the optical signal.

[0009] In an optional manner, for each node, the topological address of the node includes the numbers of the groups at each level to which the node belongs and the number of the node, and the number of the node is the number of the node under the wavelength switching device to which it is connected; each node is used to determine the relay node based on the topological address of the node and the topological address of the destination node.

[0010] In the solution disclosed herein, the topological address of a node records the location of the node. By using the topological address of the node to determine the relay node, the relay node can be found quickly.

[0011] In an optional manner, each node is also used to determine the topological address corresponding to the physical communication address of the destination node in the correspondence between the physical communication address and the topological address, and obtain the topological address of the destination node. The physical communication address is the address of the node in the upper layer encapsulation of the network used by the system.

[0012] In the solution shown in the present disclosure, when determining the topological address of the destination node, the physical communication address can be used.

[0013] In an optional manner, for each node, the i-th level wavelength switching device connected to the node is connected to the target i-th level wavelength switching device in each other group within the same level group, and the target i-th level wavelength switching device is the i-th level wavelength switching device connected to the node with the same level 1 to level i-1 group number as the node.

[0014] In the solution disclosed herein, for each node, the i-th level wavelength switching device connected to the node is connected to an i-th level wavelength switching device located symmetrically in each other group within the same level group, facilitating data transmission between nodes in the groups.

[0015] In an optional manner, when the sending node that sends the first optical signal and the destination node are located in the first-level group, the relay node and the sending node are connected to the same first-level wavelength switching device, and the node number in the topological address of the relay node is the same as the node number in the topological address of the destination node, or, the relay node and the destination node are connected to the same first-level wavelength switching device, and the node number in the topological address of the relay node is the same as the node number in the topological address of the sending node.

[0016] In the scheme shown in the present disclosure, when the sending node and the destination node are located in the first-level group, the relay node can be connected to the same first-level wavelength switching device as the sending node, or the relay node can be connected to the same first-level wavelength switching device as the receiving node, so that data can be sent to the destination node through one relay.

[0017] In an optional manner, each node is configured to determine a first wavelength corresponding to the topological address of the relay node in a correspondence between topological addresses and wavelengths, and modulate the first data onto an optical signal of the first wavelength to obtain the first optical signal.

[0018] In the solution disclosed herein, a node can use the correspondence between a topological address and a wavelength to quickly determine the wavelength being used.

[0019] In an optional manner, each node is also used to, when communication with the destination node is direct, determine the second wavelength corresponding to the topological address of the destination node in the correspondence between the topological address and the wavelength, modulate the first data onto the optical signal of the second wavelength to obtain a second optical signal, and send the second optical signal to the wavelength switching device connected to the direct communication link to send the second optical signal to the destination node.

[0020] In an optional manner, each node includes at least one level of optical module, and the system also includes at least one level of wavelength division multiplexer. For each node, the j-th level optical module is connected to the j-th level wavelength division multiplexer, and the j-th level wavelength division multiplexer is connected to the j-th level wavelength switching device, j is greater than or equal to 1, and each level of optical module includes multiple optical modules; each node is used to determine the first optical module used to communicate with the relay node before sending the first optical signal to the relay node; generate the first optical signal through the first optical module; and send the first optical signal to the connected wavelength switching device through the first wavelength division multiplexer, and the first optical module is connected to the first wavelength division multiplexer.

[0021] In the solution shown in the present disclosure, when a node includes multiple optical modules, in order to connect the multiple optical modules to one port of each wavelength switching device, the multiple optical modules are connected to each wavelength switching device through each wavelength division multiplexer.

[0022] In an optional manner, each optical module is a wavelength-tunable optical module; each node is used to, before sending the first optical signal to the relay node, tune the first optical module to the target wavelength used for communication with the relay node to establish a binding relationship between the first optical module and the target wavelength; after communication with the relay node is completed, release the binding relationship between the first optical module and the target wavelength.

[0023] In the solution shown in the present disclosure, the optical module is a wavelength-tunable optical module. Before sending an optical signal, the wavelength of the wavelength-tunable optical module is tuned to the target wavelength used to establish a correspondence between the optical module and the target wavelength, so that the optical module serves the target wavelength. After the communication is completed, the optical module is released so that the optical module can also serve other wavelengths, thereby achieving flexible scheduling.

[0024] In an optional manner, the system further includes an electrical switching device; each node is connected to the electrical switching device; and each node is further configured to transmit control information with other nodes via the electrical switching device.

[0025] In the solution shown in the present disclosure, the optical switching network system also includes an electrical switching device, which is used to transmit control information between nodes. In this way, before the optical link is established, the control information can be transmitted through a low-speed and low-capacity electrical packet switching network.

[0026] In an optional manner, each node is also used to, before sending the first optical signal to the relay node, if the communication authorization of the relay node has not been obtained, send a communication request to the relay node through the electrical switching device, and receive a communication consent message sent by the relay node through the electrical switching device.

[0027] In the solution shown in the present disclosure, before sending an optical signal to a relay node, a communication request is sent through an electrical switching device, and a communication approval message is received, so that the relay node can be prepared to receive the optical signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a schematic diagram of a node provided by an exemplary embodiment of the present disclosure;

[0029] Figure 2 is a schematic diagram of a primary switching architecture provided by an exemplary embodiment of the present disclosure;

[0030] Figure 3 is a schematic diagram of a two-level switching architecture provided by an exemplary embodiment of the present disclosure;

[0031] Figure 4 is a logical schematic diagram of an optical switching network system provided by an exemplary embodiment of the present disclosure;

[0032] Figure 5 FIG. 1 is a schematic diagram of a three-level switching architecture provided by an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0033] In order to make the objectives, technical solutions and advantages of the present disclosure more clear, the embodiments of the present disclosure will be further described in detail below with reference to the accompanying drawings.

[0034] The following is an explanation of some terminology concepts involved in the embodiments of the present disclosure.

[0035] 1. Electrical switching refers to the process of segmenting data into small packets in an electrical network. Each packet carries the destination address and other control information. Based on the network status and routing policy, the appropriate path is dynamically selected to deliver the packet to the destination node. Electrical switching offers advantages such as high transmission efficiency, high resource utilization, and strong network flexibility, making it suitable for randomly changing traffic demands. Electrical switching utilizes electrical components such as circuit switches, routers, and caches, as well as communication protocols for other network elements or terminals. Electrical switching is also known as electrical packet switching.

[0036] 2. Optical switching refers to the direct switching of input optical signals to any optical output without any optoelectronic conversion. Optical switching is a key technology in all-optical networks, improving network bandwidth, flexibility, and reliability. Optical switching is generally considered to include optical circuit switching, optical burst switching, and optical packet switching. The optical switching used in the embodiments of this disclosure falls under the category of optical burst switching, which lies between these two technologies. Specifically, during communication, a group of related data packets is combined into an optical burst, which is then rapidly cached and forwarded at the switching nodes.

[0037] 3. Optical network refers to a network built based on optical switching technology.

[0038] 4. Electric network refers to a network built based on electric switching technology.

[0039] When it comes to scenarios with high demands on communication networks, such as high-performance computing (HPC) and artificial intelligence (AI), traditional electrical networks face challenges in bandwidth and latency. HPC clusters and AI service nodes achieve faster transmission speeds and longer transmission distances by interconnecting electrical switches using optical fibers. However, this requires the installation of additional optical modules for photoelectric conversion in each switching node, which increases the cost and energy consumption of network infrastructure deployment, resulting in severe energy consumption challenges for electrical networks.

[0040] Based on these factors, people have begun exploring ways to build networks centered around optical interconnects, and the use of all-optical interconnect architectures to replace traditional electrical interconnects is gradually becoming a trend. Optical interconnects offer higher transmission bandwidth, lower energy consumption, lower network overhead, and higher-density parallelism. The transmitted optical signals exhibit low interference, low crosstalk, and low parasitic effects. They also effectively avoid the attenuation and distortion of electrical signals along the link, resulting in optical interconnect networks with transparent transmission, wavelength division multiplexing, and flexible reconfiguration. Furthermore, the rapid development of optical device technology has significantly boosted the advancement of optical interconnect technology, providing a viable solution to the limitations of electrical interconnect development.

[0041] Currently, in optical switching networks, a centralized control plane collects communication requests from each node, performs global arbitration for communication requests, and configures the status of optical components such as optical switches based on the arbitration results. However, the overall latency of the centralized control plane, including traffic sensing, arbitrating communication requests, issuing configurations, and completing optical component configuration, is high. This makes it difficult to implement real-time, accurate, and efficient control strategies in large-scale networks.

[0042] Based on this, an embodiment of the present disclosure provides an optical switching network system, in which an all-optical switching network architecture is adopted and interconnection is achieved through optical relays rather than centralized control plane-type interconnection, which is easier to expand.

[0043] The optical switching network system can be applied in a variety of scenarios. For example, in scenarios such as data centers, cloud computing, edge computing and the Internet of Things. The optical switching network system includes multiple wavelength nodes and at least one level of wavelength switching device corresponding to each node. Generally speaking, each level of wavelength switching device is responsible for the optical signal communication between nodes in each level of grouping. The wavelength switching device can be an AWGR or other device with wavelength switching function, and the node can be a computer with passive optical switching, a dedicated device that integrates passive optical switching and computing functions, and serves as the infrastructure of a data center or cloud computing platform, capable of providing high-speed, low-latency and low-power data processing and transmission services. The node can also be a server, router, switch and graphics processing unit (GPU), etc., embedded with a computing module of passive optical switching. The node can also be a computing terminal of passive optical switching, which is a terminal device with passive optical switching and computing functions, which can be connected to a passive optical network to achieve efficient data acquisition, processing and transmission, and is suitable for scenarios such as the Internet of Things and edge computing. In addition, the node can also be a general computing device, which can be a general terminal or server. The computing device has a computing interface for passive optical switching, which can connect the computer and the passive optical network, thereby realizing the coordination of data exchange and computing.

[0044] It should be noted that when the node is a GPU, it is equivalent to all the nodes in the optical switching network system belonging to one server, or multiple GPUs belong to multiple servers, and each server includes at least one GPU, which is equivalent to the GPUs between servers in the optical switching network system communicating through optical signals, or the GPUs within a server communicating through optical signals.

[0045] It should also be noted that if the current node already has the hardware requirements for passive optical switching and computing, such as optical fiber interfaces, wavelength division multiplexers, optical modulators, and optical processors, then the functions required by the embodiments of the present disclosure can be achieved by simply flashing the corresponding passive optical switching computing software into the node. The flashed software will affect the data exchange and computing mode of the network element, the algorithms used for optical signal encoding, decoding, routing, scheduling, and processing, and the protocols used for communication with other nodes or terminals.

[0046] When the current node does not have the hardware conditions for passive optical switching and computing, such as only having electrical interfaces, electrical splitters, electrical detectors, electrical modulators and electrical processors, then corresponding hardware needs to be added to realize the functions required by the embodiments of the present disclosure.

[0047] For details, see Figure 1,The node includes a data processing module, a storage resource pool, an optical module resource pool and a control module. All ,its functions are integrated into the optical network card in the form of hardware, or in the ,form of software that is not part of the framework of the data plane ,development kit.

[0048] The data processing module is used to encapsulate a new optical frame header based on the original frame, or to decapsulate it according to the network protocol stack, so that the protocol layer can provide services to the upper layer and enable optical interconnection communication at the physical layer to the lower layer. For example, when the optical switching network system is used in Ethernet, the original frame is a media access control (MAC) frame.

[0049] The storage resource pool consists of queue pairs divided by multiple wavelengths. The wavelength queue pair corresponding to each wavelength is composed of a pair of sending queues and receiving queues. The frames encapsulated by the data processing module are placed in the corresponding sending queues in sequence according to the instructions of the wavelength field, and the data received by the optical module is stored in the receiving queue of the corresponding wavelength by the data path according to the wavelength of the optical signal. The data path is a bus.

[0050] The optical module resource pool consists of at least one optical module, each of which is either a tunable optical module or a fixed-wavelength optical module. Alternatively, at least one optical module may include some fixed-wavelength optical modules and some tunable optical modules. In the case of tunable optical modules, the functions and roles of each optical module in the optical module resource pool are equivalent, with the control module responsible for dynamic scheduling and configuration. This allows for efficient utilization of network resources while reducing deployment costs and complexity and meeting node communication needs. For example, idle optical modules are assigned to the wavelength queue selected for scheduling and are tuned to the corresponding operating wavelength.

[0051] The optical module has three states: idle state, locked state, and transmission state. Since transmission is a control strategy driven by the sending node, the transmission state can be divided into master transmission state and slave transmission state according to the node type. The sending node is in the master transmission state because it holds the main control of the communication, and the receiving node is in the slave transmission state.

[0052] After the node is started, each optical module initially enters the idle state. The optical module in the idle state is in task waiting and does not need to perform any operation. The control module continuously monitors whether there is data accumulation in the sending queue of the wavelength queue. If there is data accumulation in a sending queue and the optical module is not bound to other wavelength queues, the wavelength queue is bound to an idle optical module, and the optical module enters the locked state. The optical module in the locked state is tuned to the specified wavelength according to the instructions of the control module and sends a communication request to the receiving node. If a communication consent message is received, the optical module enters the main transmission state, continues to send and receive data, and checks whether it holds a contract (the contract will be explained later). After the contract is exhausted and the end of the sending and receiving data sent by the receiving node is received, it directly enters the idle state. For the optical module of the receiving node, when a communication request is received and the authorization is granted, the optical module is transferred to the slave transmission state. After the sending node interrupts the transmission, the control optical module enters the idle state.

[0053] On a sending node, when the number of idle optical modules available for scheduling is less than the number of wavelength queues waiting for scheduling, competition among optical modules may occur within the sending node. The control module schedules the optical modules in the following ways, including but not limited to:

[0054] Method 1 is round-robin scheduling. The specific principle is: the order of wavelength queues is stored in the node. Each time the control module performs scheduling, it checks the communication needs of the next wavelength queue in sequence. If there is no data to be sent in the current wavelength queue, it continues to check the next wavelength queue in sequence until it polls a wavelength queue with data to be sent.

[0055] Method 2 is to give priority to scheduling the wavelength queue with the largest amount of data to be sent. The specific principle is: during each scheduling, the control module determines the wavelength queue with the largest amount of data to be sent among the wavelength queues, and gives priority to sending the wavelength queue with the largest amount of data. When multiple wavelength queues have the same amount of data to be sent, a wavelength queue is randomly selected.

[0056] Method 3: To prevent a wavelength queue from constantly occupying network resources due to a large amount of data to be sent, a weight is calculated based on the current amount of data to be sent and the last scheduling time (the calculation method can be to use the product of the amount of data to be sent and t as the weight, where t is the length of time from the last scheduling to the current time). The wavelength queue with the largest weight is selected for scheduling. If multiple wavelength queues have the same weight, a wavelength queue is randomly selected.

[0057] The control module is responsible for the management and coordination of each module in the node. The control module includes a routing and switching unit, a relay management unit, a contract management unit, and a transceiver management unit. Among them, the routing and switching unit records the correspondence between the physical communication address of each node in the optical switching network system and the topological address of the node, and also stores the correspondence between the topological address of the node and the wavelength. The physical communication address is the address of the node in the upper layer encapsulation of the network used by the optical switching network system. For example, the optical switching network system is applied to Ethernet, the upper layer encapsulation frame is the Ethernet MAC frame, and the physical communication address is the MAC address. For each node, the topological address includes the number of each level of grouping to which the node belongs and the number of the node. The number of the node is the number of the node under the connected wavelength switching device. When the node is connected to multiple levels of wavelength switching devices, if the ports connected under the multiple levels of wavelength switching devices are the same, then the number of the node under the multiple levels of wavelength switching devices is the same, so that the topological address of the node is unique. For example, when the relationship table between the input port, wavelength, and output port of the optical signal in each wavelength switching device at different levels is Table 1 below, each wavelength switching device includes ports 1 to 8, and node 1 is connected to port 8 of the wavelength switching device at different levels. The routing switching unit also provides a table entry query interface to the data processing module and the relay management unit, so that when the data processing module encapsulates the frame, it determines the topological address of the source and destination nodes based on the correspondence between the source and destination physical communication addresses and the topological addresses of the nodes, and then determines the wavelength required for data transmission based on the correspondence between the node's topological address and the wavelength, and sends it to the corresponding queue in the storage resource pool. When two nodes need to communicate through relay, the topological address of the relay node is obtained from the relay management unit, and the wavelength required for sending to the relay node is determined based on the correspondence between the node's topological address and the wavelength.

[0058] Optionally, the control module may be a part of a processor in the node, or the control module may be a software program.

[0059] The relay management unit is responsible for managing and calculating relay actions, and calculates the required topological address of the relay node based on the topological addresses of the source and destination nodes. The calculation process is as follows: the node stores a correspondence between the topological address of the source node, the topological address of the destination node, and the topological address of the relay node. In this correspondence, the relay node is the next hop node from the source node to the destination node. Alternatively, the topological address of the relay node is composed of the last-level group number in the topological address of the destination node and the number of the source node other than the last-level number. For example, an optical switching network system includes multiple secondary groups, and the relay node and the source node have the same two-dimensional coordinates within the secondary group, and the same secondary group number as the destination node.

[0060] The contract management unit manages contracts, recording authorization information for each wavelength queue and providing a query and modification interface to the transceiver management unit. A contract is an authorization from a receiving node to a sending node to send data frames. Only when a wavelength queue has received authorization can data be sent or received. A contract refers to the amount of data that the receiving node grants the sending node to continuously send, which can be either the number of frames or the amount of data to be sent.

[0061] The transceiver management unit is responsible for scheduling and selecting signals for each wavelength queue in the storage resource pool, applying for communication with the scheduled destination node, and storing the contract information in response. It then configures an idle optical module and data path, assigning the optical module to the scheduled wavelength queue. The wavelength queue and optical module then start transmitting and receiving data. The transceiver management unit also monitors the contracts held by each wavelength queue, terminating communication or reapplying for a new contract when the contract is exhausted.

[0062] As the function performed by the data processing module above, the data processing module encapsulates the optical frame header outside the frame transmitted in the original network to form a new frame structure, which can be called an optical frame. The frame structure of the optical frame includes information such as version field, wavelength field, type field (also called primitive field), protocol field, destination address field, source address field and data field.

[0063] The version field indicates the protocol version. Nodes communicating with each other must use the same protocol version to communicate. This field can be a 1-byte unsigned integer.

[0064] Wavelength field: Used to indicate the wavelength used when data is transmitted on the optical link. The wavelength field can be a 1-byte unsigned integer.

[0065] Type field: Used to indicate the type of data field carried by the frame. The type field can be a 1-byte unsigned integer. For example, a type field of 0x1 indicates that the data field is normal data, and 0x2 indicates that the frame carries a control message.

[0066] The Protocol field indicates the protocol used by the previous encapsulation layer, allowing the received frame data to be passed to the protocol corresponding to the previous encapsulation layer for processing. For example, a value of 0x1 in the Protocol field indicates that the upper-layer protocol is Ethernet, and the data must be passed to the Ethernet protocol.

[0067] The destination address field and source address field each occupy a 4-byte unsigned integer. In an optical switching network system, nodes communicate via optical links. That is, nodes communicate with a specific destination node using a specified wavelength, eliminating the need to perform a lookup in a physical communication address forwarding table. Therefore, nodes only query the location of the final destination node when relaying packets, using the node's topological address as the address instead of the physical communication address. As mentioned above, the node's topological address uses a fixed block addressing method to correspond to the physical meaning of the actual topology.

[0068] Data field: Includes the frame and control message encapsulated in the previous layer. The optical frame header defined in the present embodiment is added to the frame and control message encapsulated in the previous layer to produce the optical frame transmitted over the optical link in the present embodiment. For example, in Ethernet applications, the frame encapsulated in the previous layer is an Ethernet MAC frame.

[0069] It should be noted that when applied to Ethernet, the frame encapsulation in the embodiment of the present disclosure is performed in the link layer, and the optical frame header of the embodiment of the present disclosure is encapsulated outside the MAC frame encapsulated by the Ethernet protocol, and then transmitted to the next node through the physical layer.

[0070] Next, we will describe the specific architecture of the optical switching network system. First, we will describe the schematic diagram of the architecture of the rack switching in the optical switching network system. Figure 2 The optical switching network system includes a wavelength switching device and N nodes. The wavelength switching device is a first-level wavelength switching device, where N is greater than 1. The topological address of each node is simply the node number. The wavelength switching device and the N nodes form a rack. The wavelength switching device is connected to the N nodes, and each node is capable of outputting optical signals of N-1 wavelengths, where N is the number of nodes. The optical signals of the N-1 wavelengths are respectively sent to the other N-1 nodes. Each node is equipped with one or more optical network cards as needed, and optical modules are installed on each port of the optical network card. If multiple optical modules are used, a wavelength division multiplexer is also required to multiplex the optical signals of different wavelengths emitted by the optical modules into a single optical fiber, which is then connected to the corresponding ports of the wavelength switching device. If only one optical module is used, a wavelength division multiplexer is not required. Here, a network card can be equipped with one or more optical modules. Here, each node is also capable of outputting optical signals of N wavelengths, but one of them is sent to itself and is generally not used.

[0071] It should be noted that in Figure 2 In the figure, the wavelength division multiplexer is connected to the wavelength switching device using two ports, which is actually one port, just for the convenience of illustrating the optical signal transmission process, and each of the ports 1 to 4 in the wavelength switching device is one port.

[0072] Optionally, the wavelength division multiplexer can also be replaced by an optical device such as a microring resonator to achieve wavelength multiplexing and demultiplexing.

[0073] Optionally, the wavelength switching device is an AWGR. The AWGR is a passive, cacheless wavelength router that requires no configuration. It features round-robin routing, forwarding optical signals to corresponding output ports based on their input ports and wavelengths. For example, if the AWGR includes eight ports, the relationship between the input ports, wavelengths, and output ports of optical signals is shown in Table 1.

[0074] Table 1

[0075]

[0076] As shown in Table 1, the wavelength at the intersection of two ports is the wavelength used for communication between them. For example, communication between port 1 and port 2 uses an optical signal of λ2. Thus, when nodes are connected to ports and communicate with each other, the wavelength used is the wavelength used for communication between the connected ports. For example, if node 1 is connected to port 1 and node 2 is connected to port 2, the optical signal of λ2 is used when nodes 1 and 2 communicate.

[0077] It should be noted that in Table 1, although the intersection positions of the same ports also correspond to wavelengths, there is generally no light signal from the port itself.

[0078] As can be seen from Table 1, AWGR itself has a loop routing feature and does not have the phenomenon of contention conflict. Contention conflict will occur on the node side due to the lack of optical modules in the node. Figure 1In the topology shown, when the number of optical modules at each node is fully configured, that is, in an interconnected network consisting of n nodes, if each node is equipped with n-1 optical modules, it is considered fully configured. The optical modules can be wavelength-tunable optical modules or fixed-wavelength optical modules. Nodes can communicate with each other arbitrarily. When a node sends data to a destination node, it obtains a routing relationship table. The routing relationship table records the correspondence between wavelengths and the topological addresses of nodes. By querying the routing relationship table, the wavelength required to reach the destination node is obtained. The optical module corresponding to the wavelength is used to send data, so that the data is sent to the destination node. For example, the example of node (1, m) sending data to node (1, k) is used here. Node (1, m) encapsulates the MAC frame header for the data to be sent according to the Ethernet protocol. In the MAC frame header, the destination address is the MAC address of node (1, k), and the source address is the MAC address of node (1, m). Then, based on the symmetry of the AWGR wavelength routing table and the regularity of the interconnection topology, it is determined that node (1, m) can directly communicate with node (1, k). Node (1, m) encapsulates an optical frame header for a MAC frame to obtain an optical frame. In the optical frame header, the destination address is the topological address of node (1, k), the source address is the topological address of node (1, m), and the wavelength is wavelength a used by node (1, m) to node (1, k). The optical frame is inserted into a sending queue corresponding to wavelength a. When an idle optical module exists, the idle optical module is bound to a wavelength queue corresponding to wavelength a. The optical module emits an optical signal of wavelength a. The optical module modulates the optical frame onto an optical signal of wavelength a to obtain optical signal 1. The optical signal 1 is sent to a connected first-level wavelength switching device so that the optical signal 1 reaches node (1, k). After receiving the optical frame, node (1, k) inserts the optical frame into a receiving queue corresponding to wavelength a based on the wavelength in the optical frame header. The destination address in the optical frame header or the destination address in the MAC frame header is used to determine whether optical signal 1 is sent to itself, and the data is obtained by parsing based on various protocols.

[0079] When the optical modules of each node are not fully equipped, that is, in an interconnected network composed of N nodes, if the number of optical modules equipped on each node is less than N-1, it is said to be not fully equipped. When the receiving node receives the optical signal, a contention conflict occurs due to the limitation of the number of optical modules. For example, if a receiving node is equipped with k optical modules, k is less than N-1, but at a certain moment more than k nodes send data to the receiving node at the same time, and the receiving node can only support receiving data sent by k nodes at most, it will cause a contention conflict problem. In this case, the nodes are also connected through an electrical switching device, and negotiation is carried out through the electrical switching device to resolve the contention conflict problem. For example, before node 1 sends data to node 2, it sends a communication request to node 2 through the electrical switching device. Node 2 receives the communication request through the electrical switching device. If there is an idle optical module to receive the data sent by node 1, a communication consent message is sent to node 1 through the electrical switching device, and an idle first optical module is configured to be ready to receive. Node 1 receives the communication consent message and sends data to the connected AWGR using an optical signal with a wavelength corresponding to that of node 1 and node 2. The AWGR transmits the optical signal to node 2, and the first optical module obtains the data sent by node 1 from the optical signal. After sending the data to node 2, node 1 stops sending data to node 2.

[0080] Optionally, when node 1 sends a communication request to node 2, the communication request also carries the amount of data to be sent and the wavelength used to send the data. The amount of data to be sent is the amount of data sent by node 1 to node 2. After receiving the communication request, node 2 obtains the amount of data to be sent from the communication request, determines the amount of data that the wavelength queue corresponding to the wavelength used can receive, determines the minimum of the two, and determines it as the contract. When node 2 sends a communication consent message, it also carries the communication contract. After node 1 sends the amount of data specified in the contract to node 2, it stops sending data to node 2. Here, after node 1 stops sending data to node 2, node 1 sends a stop sending message to node 2 via an optical signal. The first optical module receives the stop sending message. Node 2 controls the first optical module to send a confirmation message to node 1 and releases the first optical module to an idle state. After node 1 receives the confirmation message, it releases the optical module that sends data to node 2 to an idle state.

[0081] Furthermore, the receiving node authorizes communication requests using a non-preemptive, first-come, first-served mechanism. The receiving node determines whether there are idle optical modules in the optical module resource pool. If no idle optical modules are available, the receiving node sends a rejection message to the node sending the communication request, informing it that communication is currently unavailable. Alternatively, the communication request message can include the topological address or physical communication address of the node sending the communication request, rather than the wavelength. The receiving node can then determine the wavelength to use based on the node's topological or physical communication address.

[0082] It should be noted that, in the case of full configuration, the sending node may also send a communication request to the receiving node before sending data to the receiving node, so that the receiving node is ready to receive the data sent by the sending node. Of course, whether the sending node sends a communication request before sending data to the receiving node can be executed based on the configuration.

[0083] Optionally, after node 1 and node 2 start sending data, since an optical link has been established between the two, node 2 can also send data to node 1 without requesting communication from node 1.

[0084] Optionally, the electrical switching device may be an electrical switch, or an electrical switching network composed of electrical switches. For example, each electrical switching device may be an electrical switch, and all nodes may be connected to the electrical switch. In another example, different nodes may be connected to different electrical switches, and the different electrical switches may be connected through a central electrical switch or directly.

[0085] Next, a two-level switching architecture of the optical switching network system is described. The two-level switching architecture includes switching of nodes connected to the same first-level wavelength switching device and switching of nodes connected to different first-level wavelength switching devices.

[0086] Figure 3 Provides a schematic diagram of the secondary switching architecture of the optical switching network system, see Figure 3 , the optical switching network system includes multiple first-level wavelength switching devices and multiple nodes, each first-level wavelength switching device includes p+u ports, wherein p ports are internal ports for connecting p nodes, u ports are external ports for connecting u other first-level wavelength switching devices, and each node is connected to only one first-level wavelength switching device. The multiple first-level wavelength switching devices and multiple nodes included in the optical switching network system constitute a first-level grouping, which can also be called a cluster. The number of first-level wavelength switching devices in a cluster can be set according to actual needs, and the number of nodes connected to each first-level wavelength switching device can also be set according to actual needs, which is not limited in the embodiments of the present disclosure. Figure 3 4 wavelength switching devices, namely AWGR1 to AWGR4, are shown in FIG. , and each wavelength switching device is connected to 5 nodes.

[0087] exist Figure 2In the optical switching network system shown, two nodes connected to the same first-level wavelength switching device communicate through the same first-level wavelength switching device. The communication method is similar to the process of node 1 communicating with node 2 described above. If two nodes connected to two first-level wavelength switching devices cannot communicate directly, they communicate through a relay node that is directly accessible. The relay node can be a node connected to one of the two first-level wavelength switching devices. For example, node A communicates with node B. Node A is connected to a first wavelength switching device, which is also connected to node C. Node B is connected to a second wavelength switching device. Node A determines that the relay node between itself and node B is node C. Node A searches the routing relationship table for the second wavelength corresponding to node C's address, modulates the data to be sent to node B onto the first wavelength optical signal, obtains and transmits the first optical signal, and then transmits the first optical signal to node C via the first wavelength switching device. Node C receives the first optical signal, determines that the data modulated on the first optical signal does not belong to itself, uses the destination address carried in the data, searches the routing relationship table for the third wavelength corresponding to the destination address, modulates the data sent to node B onto the optical signal of the third wavelength, and obtains the third optical signal. Node C then sends the third optical signal to the second wavelength switching device, which then sends the second optical signal to node B. For another example, if the second wavelength switching device is also connected to node D, and node D is used as a relay node, node A first sends an optical signal to node D, and node D sends an optical signal to node B, so that the data is received by node B.

[0088] Optionally, the first-level wavelength switching device is an AWGR, and the AWGR and the connected nodes form a rack. Each rack is numbered using a natural number, and the connection method within each rack is the same as that in Figure 2 , each node in a rack is defined by a topological address, which includes the number of the first-level group (i.e., the rack number) and the two-dimensional coordinates of the node number in the rack. For example, a node numbered (3, 1) indicates that it is the first node in the third rack. The connection relationship between racks can be: u ports of an AWGR are connected to u other AWGRs respectively, and when connected, the kth port is connected to the kth port of another AWGR, indicating that the AWGRs are connected using ports corresponding to the same wavelength exchange relationship. For example, Figure 3 Four racks are shown, denoted as R1 to R4. Each rack includes an AWGR and nodes connected to the AWGR. Each AWGR includes 8 ports, namely port 1 to port 8, where port 1 to port 5 connects to 5 nodes, and port 6 to port 8 connects to another 3 AWGRs. The connection relationship between the AWGRs is shown in Table 2.

[0089] Table 2

[0090]

[0091] In Table 2, AWGR1 provides a first port, and AWGR2 provides a second port. The first port is connected to the second port. The first port is port 6 of AWGR1, and the second port is port 6 of AWGR2. The two have the same wavelength exchange relationship. Here, the same wavelength exchange relationship means that the optical signal of the first wavelength input from port A in AWGR1 will be transmitted to port 6 of AWGR1, and the optical signal of the first wavelength input from port 6 in AWGR2 will also be transmitted to port A of AWGR2.

[0092] From the symmetry of the AWGR wavelength routing table, we can see that two nodes located in different racks but with the same rack number can communicate directly. For example, combining Table 1 and Table 2, Figure 2 The intermediate node (3, 1) uses wavelength λ4 to send the optical signal from port 5 of AWGR3 to port 8. Port 8 of AWGR3 is directly connected to port 8 of AWGR2. At this time, the wavelength of the optical signal is still λ4. According to the wavelength routing table of AWGR, the optical signal is transmitted from port 8 of AWGR2 to port 5, and finally output from port 5 to node (2, 1).

[0093] Thus, within a cluster, when cross-rack communication is performed, a relay node is required to perform intra-rack relay. Assume that node (i, m) (the mth node in the i-th rack) needs to send data to node (j, n) (the nth node in the j-th rack). There are multiple ways to send data, and two feasible methods are provided below:

[0094] Method 1: Node (i, m) first sends data to the relay node (j, m) in the rack where node (j, n) is located. The data is then sent to node (j, n) within the rack via the relay node. The process is as follows: Node (i, m) encapsulates the data to be sent with a MAC frame header according to the Ethernet protocol. In the MAC frame header, the destination address is the MAC address of node (j, n), and the source address is the MAC address of node (i, m). Then, based on the symmetry of the AWGR wavelength routing table and the regularity of the interconnection topology, the relay node is determined to be the mth node in the jth rack. Node (i, m) encapsulates the MAC frame with an optical frame header to obtain an optical frame. In the optical frame header, the destination address is the topological address of node (j, n), the source address is the topological address of node (i, m), and the wavelength is wavelength 1 used by node (i, m) to node (j, m). The optical frame is then inserted into the transmit queue corresponding to wavelength 1. When an idle optical module exists, it is bound to the wavelength queue corresponding to wavelength 1. The optical module emits an optical signal of wavelength 1. The optical module modulates the optical frame onto the optical signal of wavelength 1 to obtain a second optical signal, which is then sent to the connected first-level wavelength switching device. This second optical signal reaches node (j, m). After receiving the second optical signal, node (j, m) inserts the optical frame into the receive queue corresponding to wavelength 1 based on the wavelength in the optical frame header. Node m uses the destination address in the optical frame header to determine a topological address that is not its own. Node m uses the destination address in the optical frame header to determine the corresponding wavelength 2 based on the correspondence between the node's topological address and wavelength. The wavelength in the optical frame header is modified to wavelength 2, and the optical frame is inserted into the transmit queue corresponding to wavelength 2. If an idle optical module exists, it is bound to the wavelength queue corresponding to wavelength 2. The optical module then emits an optical signal at wavelength 2. The optical module modulates the optical frame onto the optical signal at wavelength 2 to obtain a third optical signal. This signal is then sent to the connected first-level wavelength switching device, allowing the third optical signal to reach node (j, n). Upon receiving the signal, node (j, n) inserts the optical frame into the receive queue corresponding to wavelength 2 based on the wavelength in the optical frame header. Using the destination address in the optical frame header as its own topological address, node (j, n) determines that the frame is being sent. Each layer of the protocol then processes the frame to obtain the final data. See Table 3 for the frame headers at nodes (i, m) and (j, m).

[0095] Table 3

[0096]

[0097]

[0098] It should be noted that in Table 3, in the optical frame sent by node (i, m), the destination address in the optical frame header is the topological address of the destination node. In practice, the destination address can also be the topological address of a relay node. In this case, the relay node uses the destination address in the MAC frame header to determine whether the optical frame is sent to itself. Similarly, in Table 3, in the optical frame sent by node (j, m), the source address in the optical frame header is the topological address of the source node. In practice, the source address can also be the topological address of a relay node.

[0099] In method 2, node (i, m) first sends the data to the relay node (i, n) in the rack where it is located, and then communicates between racks through the relay node (i, n) to send the data to node (j, n). This process is described in method 1 and will not be repeated here.

[0100] In addition, within a cluster, for communication between nodes within a rack, since the nodes belong to the same AWGR downstream, the communication strategy between the two follows the example of node (i, m) sending data to node (i, k) in the previous article, which will not be repeated here.

[0101] From the above Figure 3 As can be seen from the description, for node (1, 1), port 7 to port 1 in the AWGR in R1 and the AWGR in R3 form a wavelength switching network. Nodes (1, 1) to (1, 5) are connected to this wavelength switching network. The rest of the AWGR in R3, except for port 7 to port 1, also forms a wavelength switching network. Node (3, 1) is a relay node. This is equivalent to connecting each wavelength switching network in the two wavelength switching networks. Different wavelength switching networks are connected using relay nodes. See Figure 4 .

[0102] Next, the three-level switching architecture of the optical switching network system is described. The three-level switching architecture includes switching of nodes connected to the same first-level wavelength switching device, switching of nodes connected to different first-level wavelength switching devices, and switching of nodes connected to different second-level wavelength switching devices.

[0103] See also Figure 5 The optical switching network system includes multiple secondary groups, each of which includes a first-level wavelength switching device, a second-level wavelength switching device, and a node. In each secondary group, the first-level wavelength switching device and the second-level wavelength switching device are both connected to the node, that is, each node is connected to a first-level wavelength switching device and a second-level wavelength switching device. The first-level wavelength switching device is connected to each first-level wavelength switching device in the same second-level group, see Figure 3The connection mode shown. Each node is connected to the second-level wavelength switching device and is connected to at least one second-level wavelength switching device in each other second-level group in the same level second-level group. In this way, it is equivalent to Figure 3 In the optical switching network system shown, each rack has an additional second-level wavelength switching device for communication between secondary groups. The p ports of the second-level wavelength switching device serve as internal ports, connecting to the p nodes in the rack. Therefore, each node needs to add an optical network card to connect to the internal ports of the second-level wavelength switching device. The u ports of the second-level wavelength switching device serve as external ports to connect to the second-level wavelength switching devices in u secondary groups outside the secondary group. Therefore, when communicating within the same secondary group, see the Figure 3 As described in , when communicating across secondary groups, nodes communicate through the second-level wavelength switching device connected to other secondary groups. Figure 5 , four secondary groups, ie, four clusters, are shown, but only the connection between the cluster in the upper left corner and the other clusters is shown, and the rest are similar.

[0104] In an optical switching network system, each secondary group can be numbered according to a natural number, and each node is numbered and defined by a three-dimensional coordinate system including the secondary group number, the rack number, and the node number within the rack, indicating the topological address. Therefore, an optical switching system including multiple secondary groups is also called a three-level optical switching network system.

[0105] The nodes connected to different levels of wavelength switching devices of each node are the same. For example, for node 1, the connected first-level wavelength switching device connects nodes 1 to 5, and the connected second-level wavelength switching device also connects nodes 1 to 5.

[0106] In an optional manner, the wavelength switching devices at each level are AWGRs, and the connection method within the secondary group is described above and will not be repeated here. When connecting between secondary groups, for a certain node, the i-th level wavelength switching device connected to the node is connected to the target second-level wavelength switching device in each other secondary group within the secondary group, and the target second-level wavelength switching device is the i-th level wavelength switching device connected to the node with the same first-level secondary group number as the node. For example, node 1 is the first node under the first rack in the first secondary group, and the topological address of node 1 is represented as (1, 1, 1). The second-level wavelength switching device connected to node 1 is connected to the target second-level wavelength switching device, and the topological address of the node connected to the target second-level wavelength switching device is (2, 1, 3), indicating the third node under the first rack in the second secondary group.

[0107] For each node, the port to which the node is connected in the first-level wavelength switching device is the third port, and the port to which the node is connected in the second-level wavelength switching device is the fourth port. The third port and the fourth port correspond to the same wavelength switching relationship.

[0108] In addition, the connection relationship between the second-level wavelength switching devices is similar to the connection relationship between racks. The connection relationship can be: u ports of the second-level wavelength switching device are respectively connected to u second-level wavelength switching devices, and its k-th port is connected to the k-th port of the second-level wavelength switching device of other second-level groups, that is, the second-level wavelength switching devices use the same port connection. For example, the first AWGR and the second AWGR are both second-level wavelength switching devices, belonging to two second-level groups. The first AWGR and the second AWGR are connected through the first port and the second port. The first port belongs to the first AWGR, and the second port belongs to the AWGR. The first port and the second port correspond to the same wavelength exchange relationship. For example, Figure 5 There are four secondary groups, namely, secondary group 1 to secondary group 4. Each secondary group includes 4 racks. Each AWGR includes 8 ports, namely, port 1 to port 8. For the connection mode of AWGR in the secondary group, see Figure 2 ,Port 1 to Port 5 of the AWGR between the second-level groups (i.e., AWGR-1 to AWGR-4) are connected to 5 nodes, and Port 6 to Port 8 are respectively connected to a second-level AWGR of the other three second-level groups. The connection relationship between the second-level AWGRs is shown in Table 4.

[0109] Table 4

[0110]

[0111] In Table 4, AWGR-1 provides a first port, and AWGR-2 provides a second port. The first port and the second port are connected. The first port is port 6 of AWGR-1, and the second port is port 6 of AWGR-2. The two ports have the same wavelength swap relationship. Here, the same wavelength swap relationship means that the second wavelength optical signal input from port B of AWGR-1 will be transmitted to port 6 of AWGR-1, and the second wavelength optical signal input from port 6 of AWGR-2 will be transmitted to port B of AWGR-2.

[0112] From the symmetry of the AWGR wavelength routing table, we can see that two nodes in different secondary groups but with the same rack and rack number can communicate directly. For example, combining Table 2 and Table 4, Figure 5The intermediate node (1, 3, 1) uses wavelength λ4 to send the optical signal from port 5 of AWGR-1 to port 8. Port 8 of AWGR-1 is directly connected to port 8 of AWGR-2. At this time, the wavelength of the optical signal is still λ4. According to the wavelength routing table of AWGR, the optical signal is transmitted from port 8 of AWGR-2 to port 5, and finally output from port 5 to node (2, 3, 1).

[0113] When communicating across different secondary groups, at least one relay node is required for intra-group relaying. Assume that node (x, i, m) (node ​​numbered m in rack i of the x-th secondary group) sends data to node (y, j, n). There are several possible ways to send data, two of which are provided below.

[0114] Method 1: Node (x, i, m) first sends data to relay node (y, i, m) within the secondary grouping where node (y, j, n) resides. Then, within the secondary grouping, the data is sent to node (y, j, n) via relay node (y, i, m). The process is as follows: Node (x, i, m) encapsulates the data in a MAC header according to the Ethernet protocol. In the MAC header, the destination address is the MAC address of node (y, j, n), and the source address is the MAC address of node (x, i, m). Based on the symmetry of the AWGR's wavelength routing table and the regularity of the interconnection topology, the relay node is determined to be the mth node in the i-th rack of the y-th secondary grouping, i.e., node (y, i, m). Node (x, i, m) encapsulates the optical frame header for the MAC frame to obtain an optical frame. In the optical frame header, the destination address is the topological address of node (y, j, n), the source address is the topological address of node (x, i, m), and the wavelength is wavelength 4 used from node (x, i, m) to node (y, i, m). The optical frame is inserted into the sending queue corresponding to wavelength 4. When there is an idle optical module, the idle optical module is bound to the wavelength queue corresponding to wavelength 4. The optical module emits an optical signal of wavelength 4. The optical module modulates the optical frame onto the optical signal of wavelength 4 to obtain a fifth optical signal, which is sent to the connected second-level wavelength switching device so that the fifth optical signal reaches node (y, i, m). After receiving the optical frame, node (y, i, m) inserts the optical frame into the receiving queue corresponding to wavelength 4 according to the wavelength in the optical frame header. Node (y, i, m) uses the destination address in the optical frame header to determine that it is not its own topological address and continues to send. For the sending process, see [1]. Figure 3 The description of the two-level switching architecture shown in FIG is omitted here. Thus, the communication process from node (x, i, m) to node (y, j, n) is: node (x, i, m) - node (y, i, m) - node (y, j, m) - node (y, j, n), or node (x, i, m) - node (y, i, m) - node (y, i, n) - node (y, j, n).

[0115] Method 2: Node (x, i, m) first sends data to the relay node (x, j, n) in its secondary group, and then communicates between racks through the relay node (x, j, n) to send the data to node (y, j, n). The process is as follows: Node (x, i, m) sends data to the relay node (x, j, n) in its secondary group. See the corresponding process in the previous article. Figure 3 In method 1 or method 2, the relay node (x, j, n) uses the destination address in the optical frame header to determine that it is not its own topological address and continues to send. For the sending process, see Figure 3 The description of the two-level switching architecture shown in FIG is omitted here. Thus, the communication process from node (x, i, m) to node (y, j, n) is: node (x, i, m) - node (x, j, m) - node (x, j, n) - node (y, j, n). Alternatively, the communication process from node (x, i, m) to node (y, j, n) can be: node (x, i, m) - node (x, j, m) - node (y, j, m) - node (y, j, n).

[0116] Furthermore, within a secondary group, for intra-rack node communication, since the nodes are downstream of the same AWGR, the communication strategy between them follows the primary switching architecture. Within a secondary group, for inter-rack node communication, since the AWGRs between racks are connected, the communication strategy between them follows the secondary switching architecture.

[0117] In addition, in the three-level optical switching network system, the communication between nodes will also obtain the signal contract. The process of obtaining the signal contract is described in the previous article and will not be repeated here.

[0118] The above describes the architecture of a three-level optical switching network system. By analogy, a higher-dimensional multi-level architecture can be constructed. With each additional level of architecture, a corresponding number of AWGRs are added to each rack to provide interconnection at that level, and a certain number of network cards are added to the nodes to connect to the AWGRs at that level. This facilitates the construction of a higher-dimensional multi-level architecture.

[0119] In one optional embodiment, in a multi-stage switching architecture, each node is connected to an electrical switching device, and each node can communicate with each other through the electrical switching device to transmit control information, including but not limited to communication requests and communication consent messages. For example, before a sending node sends data to a receiving node, the sending node sends a communication request to the receiving node via the electrical switching device to obtain a data transmission agreement. The process of sending a communication request is described above and will not be repeated here.

[0120] Optionally, when there are multiple relay nodes in the next hop, the source node can randomly select a relay node when selecting a relay node, or it can first send communication requests to multiple relay nodes through an electrical switching device, establish an optical link with the relay node that first replies to the communication message, and send a stop communication message to the relay node that replies to the communication message later.

[0121] In the disclosed embodiments, a fine-grained all-optical architecture of a decentralized control plane is implemented, and a multi-level and multi-hop interconnection architecture is implemented by optical relay routing, which can achieve the expansion scale of the electrical network.

[0122] In the present disclosure, the terms "first" and "second" are used to distinguish between identical or similar items having substantially the same effects and functions. It should be understood that there is no logical or temporal dependency between "first" and "second," nor is there a limitation on quantity or order of execution. It should also be understood that although the following description uses the terms "first," "second," etc. to describe various elements, these elements should not be limited by the terms. These terms are merely used to distinguish one element from another. For example, without departing from the scope of the various examples, a first optical signal may be referred to as a second optical signal, and similarly, a second optical signal may be referred to as a first optical signal. Both the first optical signal and the second optical signal may be optical signals, and in some cases, may be separate and distinct optical signals.

[0123] The term "at least one" in the present disclosure means one or more, and the term "plurality" in the present disclosure means two or more.

[0124] The above description is merely a specific embodiment of the present disclosure, but the scope of protection of the present disclosure is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present disclosure, and such modifications or substitutions should be included in the scope of protection of the present disclosure. Therefore, the scope of protection of the present disclosure should be based on the scope of protection of the claims.

Claims

1. An optical switching network system, characterized in that: including a plurality of nodes and at least one level of wavelength switching device connected to each node; The first-level wavelength switching device connected to each node is connected to each first-level wavelength switching device located in the same-level group; In the case of multiple levels of wavelength switching devices, the i-th level wavelength switching device connected to each node is connected to at least one i-th level wavelength switching device in each other group within the same level group, the different levels of wavelength switching devices connected to each node are connected to the same node, the i-1th level group belongs to the i-th level group, and i is greater than or equal to 2; Each node is configured to, when communication with a destination node is not directly connected, determine a relay node with direct communication among the multiple nodes, and send a first optical signal to the relay node through a wavelength switching device on the direct communication link, so as to send first data carried on the first optical signal to the destination node through the relay node; Each wavelength switching device is used to output the received optical signal through the wavelength switching function.

2. The system according to claim 1, wherein: The wavelength switching device at each level is an arrayed waveguide grating router AWGR; Two wavelength switching devices having a connection relationship are connected through a first port and a second port. The first port and the second port correspond to the same wavelength switching relationship. The first port belongs to one of the two wavelength switching devices, and the second port belongs to the other wavelength switching device.

3. The system according to claim 1 or 2, characterized in that For each node, the topological address of the node includes the number of each level group to which the node belongs and the number of the node, and the number of the node is the number of the node under the wavelength switching device to which the node is connected; Each node is configured to determine the relay node based on the topological address of the node and the topological address of the destination node.

4. The system according to claim 3, characterized in that Each node is also used to determine the topological address corresponding to the physical communication address of the destination node in the correspondence between the physical communication address and the topological address, and obtain the topological address of the destination node. The physical communication address is the address of the node in the upper layer encapsulation of the network used by the system.

5. The system according to claim 3 or 4, characterized in that For each node, the i-th level wavelength switching device connected to the node is connected to the target i-th level wavelength switching device in each other group within the same level group. The target i-th level wavelength switching device is the i-th level wavelength switching device connected to the node with the same level 1 to level i-1 group number as the node.

6. The system according to any one of claims 3 to 5, characterized in that In the case where the sending node that sends the first optical signal and the destination node are located in the first-level group, the relay node and the sending node are connected to the same first-level wavelength switching device, and the node number in the topological address of the relay node is the same as the node number in the topological address of the destination node; or, the relay node and the destination node are connected to the same first-level wavelength switching device, and the node number in the topological address of the relay node is the same as the node number in the topological address of the sending node.

7. The system according to any one of claims 3 to 6, characterized in that Each node is configured to determine, in a correspondence between topological addresses and wavelengths, a first wavelength corresponding to the topological address of the relay node, and modulate the first data onto an optical signal of the first wavelength to obtain the first optical signal.

8. The system according to any one of claims 3 to 7, characterized in that Each node is also used to, when communication with the destination node is direct, determine the second wavelength corresponding to the topological address of the destination node in the correspondence between the topological address and the wavelength, modulate the first data onto the optical signal of the second wavelength to obtain a second optical signal, and send the second optical signal to the wavelength switching device connected to the direct communication link to send the second optical signal to the destination node.

9. The system according to any one of claims 1 to 8, characterized in that Each node includes at least one level of optical module, and the system also includes at least one level of wavelength division multiplexer. For each node, the j-th level optical module is connected to the j-th level wavelength division multiplexer, and the j-th level wavelength division multiplexer is connected to the j-th level wavelength switching device, where j is greater than or equal to 1, and each level of optical module includes multiple optical modules; Each node is configured to, before sending the first optical signal to the relay node, determine a first optical module used to communicate with the relay node; generating the first optical signal by the first optical module; The first optical signal is sent to a connected wavelength switching device through a first wavelength division multiplexer, and the first optical module is connected to the first wavelength division multiplexer.

10. The system according to claim 9, characterized in that Each optical module is a wavelength-tunable optical module; Each node is configured to, before sending the first optical signal to the relay node, tune the first optical module to a target wavelength used for communicating with the relay node, so as to establish a binding relationship between the first optical module and the target wavelength; After the communication with the relay node is completed, the binding relationship between the first optical module and the target wavelength is released.

11. The system according to any one of claims 1 to 10, characterized in that The system also includes an electrical exchange device; Each node is connected to the electrical switching device; Each node is further configured to transmit control information to other nodes via the electrical switching device.

12. The system according to claim 11, characterized in that Each node is further configured to, before sending the first optical signal to the relay node, send a communication request to the relay node through the electrical switching device if the communication authorization of the relay node has not been obtained, and receive a communication consent message sent by the relay node through the electrical switching device.