Satellite laser communication optical network dynamic route optimization distribution method

Abstract

In a satellite optical network system, network routing is particularly important for improving network performance, because the network routing can effectively utilize laser link resources among all satellites of the satellite optical network. Generally, the link cost of a certain inter-satellite link is represented in a time delay manner by using information. Because each node in the satellite optical network moves constantly, the link cost of each inter-satellite link changes constantly, and a network topology structure in which multiple links are backups for each other forms a plurality of ring networks, factors of link distance, transmission time delay, Doppler effect and geographic positions of link satellites need to be comprehensively considered, and the link cost of each inter-satellite link is changed constantly. And searching an optimal route (an optimal optical path) for a source-target satellite node to a connection request. According to the invention, the multi-loop network virtual topological structure of the satellite laser communication optical network is constructed, so that the network topological structure becomes a loop-free tree network topological structure, and broadcast storm, multi-frame copying and instability of an address table caused by a multi-loop redundant link network topological structure can be avoided. Based on the method and the device provided by the invention, an optimal special link can be quickly assigned between source-target user nodes in a satellite optical network system based on a link cost judgment mode, and a data transmission path can be automatically switched according to a physical layer laser link state, so that the problem of dynamic routing distribution of a satellite laser communication optical network can be effectively solved.

Description

Technical field:

[0001] The present invention provides a method for dynamic routing optimization and allocation of satellite high-speed laser communication optical networks, and relates to the field of satellite optical network networking system construction. Background technology:

[0002] In satellite optical networks, network routing is crucial for improving network performance because it effectively utilizes the resources of each intersatellite laser link. Typically, the link cost of an intersatellite link is represented by the time delay of information. Because nodes in satellite optical networks are constantly in motion, the link cost of each intersatellite link varies. The link cost calculation for the routing scheme comprehensively considers factors such as intersatellite link distance, transmission delay, Doppler effect, and the geographic location of the linking satellites to determine the optimal route (optimal optical path) for the connection request between the source and destination satellite node pair. Furthermore, to account for the time-varying nature of the optimal optical path, an optimal routing set is established for all source-destination node pairs in the network topology at each moment within the constellation cycle. Summary of the invention:

[0003] When a connection request arrives from a source-destination satellite node pair, the network will assign the optimal optical path among all available optical routes that meet the connection request based on routing selection. The optimal optical path is determined by whether it minimizes the total link cost. Each optical path for a source-destination node pair consists of several intersatellite laser links connected in series, starting from the source satellite, passing through several intermediate routing nodes, and reaching the destination satellite. Assuming the network is a real-time system providing low-latency services, the total link cost of the optical route should be measured in time. The total link cost of the optical route is composed of two components: the link cost of a single intersatellite link and the link cost introduced by signal processing at the intermediate nodes.

[0004] Each intersatellite link has a transmission delay determined by the distance between the two satellites. Therefore, the link cost of a single intersatellite link is related to the link distance between the satellites. In addition, the Doppler wavelength shift of the intersatellite link increases the processing complexity of the receiver. Since the Doppler effect exists in inter-orbit links but not in intra-orbit links, intra-orbit links should be given priority in routing. Define P j,Dop represents the link cost introduced by the Doppler effect of the jth intersatellite laser link on the optical route. In order to quantify the numerical range of the parameter scale in the link cost function, P j,Dop The value range of is set to [0,1]. The smaller the Doppler wavelength drift of the intersatellite link, the closer the introduced link cost is to 1, and the higher the selection weight of the link. Combining the above two factors, the link cost expression of a single intersatellite laser link can be obtained as:

[0005]

[0006] Where, d j (t i ) represents the link distance of the jth intersatellite link in the optical route, and c represents the speed of light. Both the intersatellite link distance and the Doppler effect are functions of time. As the formula shows, the link cost of a single intersatellite link is also a function of time.

[0007] The second part of the total link cost of optical routing is the forwarding signal processing time at the intermediate node. j,p is the processing time of the router at the receiving end of the j-th intersatellite link on the optical route, and the link cost introduced by the intermediate node forwarding signal processing is expressed as:

[0008] W j，router (t i )=t j，p (2)

[0009] Let N hops represents the number of intersatellite link hops included in the path. Then, the total link cost of the optical route can be expressed as the sum of two parts:

[0010]

[0011] Right now:

[0012]

[0013] By using formula (4) to calculate the link cost of the optical route, the optimal route (optimal optical path) set can be established for all node pairs in the satellite optical network topology. According to the system communication design plan, the inter-satellite multi-constellation network data exchange function is realized, and a redundant backup network topology is constructed. Each routing exchange node establishes a link with the adjacent routing exchange node to realize a multi-loop link topology. The routing exchange node is used to exchange data between routing exchange nodes in a software-defined manner to build an optimal laser network path between users, user nodes, intermediate nodes, and user nodes and users. A single-node multi-link connection mode is realized, and a single-path network data exchange function is realized with other nodes, with real-time redundant network data exchange function. The schematic diagram of the network topology of the satellite laser communication optical network system is shown in the attached figure [1].

[0014] The exchange of network data can only be achieved on the basis of the construction of network topology, so the construction of network topology is the foundation. The construction of network topology requires the use of certain algorithms to virtualize the multi-loop network topology into a loop-free tree network topology. The multi-loop redundant link network topology will cause broadcast storms, multi-frame replication and address table instability, so it is necessary to dynamically manage links in the network to avoid loops by blocking certain links. When a normally working link is disconnected due to a fault or data is blocked, other links are immediately activated to quickly replace the faulty link to ensure the normal operation of the network. Network topology link switching design of satellite laser communication optical network system:

[0015] 1) When a network topology is being constructed, when a source node needs to send data to a destination node but does not have a specific path to the destination, the source node initiates the route discovery process. The source node broadcasts a Route Request (RREQ) packet, which contains the source node's address, the destination node's address, a unique request ID, and other information (such as the source address and hop counter).

[0016] 2) The intermediate node receives the pathfinding packet and performs inspection, path recording, forwarding, and destruction operations:

[0017] a) Check: The intermediate node checks the pathfinding packet to confirm whether it has been processed (identified by the source address). If so, the pathfinding packet is discarded.

[0018] b) Record the path: If the pathfinding packet is received for the first time, the node records the source of the pathfinding packet and updates its own routing table.

[0019] c) Forwarding: The node selects the next node according to the routing strategy and forwards the pathfinding packet.

[0020] d) Destruction: The intermediate node will increase the value of the hop counter. When the hop counter reaches the preset value, the intermediate node will no longer forward the pathfinding packet and destroy the pathfinding packet.

[0021] 3) After receiving the pathfinding packet, the destination node will check the information in the pathfinding packet and then update its own routing table. At this time, the destination node can send data to the source node according to the routing table.

[0022] 4) By analogy, after the source node receives the RREQ from the destination node, the source node can send data to the destination node.

[0023] 5) Network conditions may change (such as node failure, link congestion, etc.), so it is necessary to periodically send pathfinding packets to maintain and update path information.

[0024] 6) When the information in its own routing table reaches the preset survival time and no same pathfinding packet is received again, the information in this routing table is destroyed.

[0025] 7) Link Probe Packet: This packet contains the user data network switching unit node number, the destination network switching unit node number, the number of data forwarding hops (the number of hops is reduced by 1 for each switching center), and the data link selection (selecting the path with the shortest number of hops or the path with the shortest latency). The link probe packet is sent by the user data network switching unit and destroyed by an intermediate node (if the hop count is 0, it is destroyed to avoid data storms) or the destination user data node (to complete the link probe function).

[0026] Constructing a multi-loop virtual topology for a satellite laser communication optical network can transform the network into a loop-free tree-like structure, avoiding the broadcast storms, multiple frame duplication, and address table instability that can occur with multi-loop redundant link topologies. The method and apparatus of the present invention, based on link cost, can rapidly assign an optimal dedicated link between source and target user nodes in a satellite optical network system. The method and apparatus can also autonomously switch data transmission paths based on the status of the physical layer laser link, effectively solving the problem of dynamic routing allocation in satellite laser communication optical networks. Description of the drawings:

[0027] Figure 1 This is a schematic diagram of the network topology of the satellite laser communication optical network system of the present invention.

[0028] Figure 2 This is a schematic diagram of the overall link of the present invention

[0029] Figure 3 This is a schematic diagram of the link detection data packet transmission of the present invention Implementation method:

[0030] For the convenience of introduction, the implementation method is described by taking a satellite laser communication optical network system consisting of 5 satellites as an example, see Figure [2]. Each satellite is a network routing switch node with 4 data ports (ports without links are not considered for the time being). Each network node in the system performs multi-loop path detection and establishes a virtual data path. The user data network routing switch sends the local user data port information at the link data port. The intermediate node network routing switch receives the user data port information and stores it, and then selects it according to the port information during data exchange.

[0031] To prevent repeated network transmission from occupying the data channel, the pathfinding packet propagation level is set to 3. Based on the link distance and transmission time, combined with the characteristics of the constellation, the link packet is sent once per second, and the data packet is not sent repeatedly for 800ms.

[0032] Analyze the construction process of the link propagation path based on the overall link diagram:

[0033] Assume that node A1 has a user interface. Its user data is transmitted via the A1 network routing switch unit in the form of a link detection packet. This packet contains information such as node connectivity and the number of hops. Based on the physical layer laser link connections, node A1 can connect to three network routing switch units: B1, B2, and B3. Node A1 needs to send link detection packets on its ports A1-2, A1-3, and A1-4 to communicate the node number of the A1 network routing switch unit to the adjacent network routing switch units.

[0034] 1) The intermediate node network switching unit processes the link detection packet. Port B1-1 of node B1 receives and stores the link detection packet. Port B1-1 also receives information about node A1. When storing, it filters packets based on the number of hops (the one with the smallest hop number is stored) or time information (the source port information is not updated within 800ms of receiving the port information). At this point, B1 is able to send data to A1 via port B1-1 at any time.

[0035] 2) B1 receives the link detection packet and forwards it. The decision criteria are: if the hop count is not 0 and B1 has not forwarded the pathfinding packet within the previous 800ms, then it forwards the packet (forwarding it on each link port, such as B1-3 and B1-4). Otherwise, it does not forward the packet. Other intermediate node network switching units process the link detection packet in a similar manner.

[0036] 3) Taking the left transmission path as an example, port B1-4 transmits the A1 pathfinding packet to port B2-2, but at this time port B2-1 has already received the same pathfinding packet sent by port A1-3 and forwarded it to ports B2-2, B2-3, and B2-4. The time B2 receives the A1 pathfinding packet forwarded by port B1-4 is less than 800ms of the last forwarding time. At this time, B2 does not need to forward the pathfinding packet forwarded by port B1-4 to port B2-2.

[0037] 4) The target user receives the link detection data packet through the network switching unit link, stores the node information, and destroys the link detection data packet.

[0038] 5) Analyze the link propagation path in combination with the overall link diagram. See the attached figure [3].

Claims

1. This paper proposes a link-cost-based network optimal routing (optimal optical path) method that comprehensively considers factors such as link distance, transmission delay, Doppler effect, and the geographic location of the link satellite to find the optimal route (optimal optical path) for the connection request between the source and target satellite nodes. The method can effectively reflect the time-varying characteristics of the optimal optical path and effectively utilize the inter-satellite laser link resources of the satellite optical network.

2. The present invention proposes a method for virtualizing a multi-loop network topology structure into a loop-free tree network topology structure, performing dynamic link management in the network, and avoiding the generation of loops by blocking certain links. When a normally working link is disconnected due to a fault or data is blocked, other links are immediately activated and quickly replace the position of the faulty link, which can effectively ensure the normal operation of the network and avoid broadcast storms, multi-frame duplication and address table instability problems caused by the multi-loop redundant link network topology structure.

3. This invention proposes a network path search method based on link detection packets. In an optical network system, each network node performs multi-loop path detection and establishes a virtual data path. The user data network routing switch sends local user data port information at the link data port. The intermediate node network routing switch receives and stores the user data port information, and then selects the port information during data exchange.

Images