A method for routing aviation cluster networks

By dynamically adjusting the transmission interval and link performance evaluation value of LLDP messages, the optimal path is selected, which solves the problem of network topology changes in the aviation cluster network, improves communication efficiency and service quality, and meets the flexible combat needs of the dynamic aviation cluster network.

CN114339929BActive Publication Date: 2024-11-01CHINESE AERONAUTICAL RADIO ELECTRONICS RES INST
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202111422087.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-11-26
Publication Date
2024-11-01
Estimated Expiration
2041-11-26

AI Technical Summary

Technical Problem

The existing aviation cluster networks have frequent changes in network topology due to the rapid movement of drones, which is difficult to meet the needs of flexible combat.

Method used

By dynamically adjusting the time interval for the software-defined network controller node to send link layer discovery protocol LLDP messages, calculate the performance evaluation value based on the link bandwidth and transmission delay, and select the optimal routing path for data forwarding.

Benefits of technology

It improves the bandwidth utilization efficiency of the communication link, ensures service quality, meets the time-sensitive dynamic aviation cluster network requirements, and achieves adaptable and efficient service transmission performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114339929B_ABST
    Figure CN114339929B_ABST
Patent Text Reader

Abstract

The present invention discloses an aviation cluster network routing method, comprising: step 1, dynamically adjusting the time interval for a software-defined network controller node to send a link layer discovery protocol LLDP message according to the location information and movement information of the aviation cluster network node; step 2, calculating the performance evaluation value of each link according to the bandwidth and transmission delay of each link; step 3, selecting the optimal routing path for data forwarding according to the performance evaluation value of each link. The technical solution provided by the embodiment of the present invention solves the problem that the existing aviation cluster network will cause the network topology to change frequently due to the rapid movement of drones, thereby making it difficult to meet the flexible combat requirements of future aviation cluster networks.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to, but is not limited to, the field of aviation wireless communication technology, and in particular to an aviation cluster network routing method. Background Art

[0002] With the development of drone technology, more and more UAVs are used in aerial battlefields to complete various combat tasks, such as situational awareness, target monitoring, search, etc. Since it is difficult for a single drone to complete complex tasks, clustering and networking are the development trends of drone technology. The aviation cluster network composed of multiple drones breaks through the limitations of a single drone and has the advantages of low cost and high efficiency. Its networking capability and network performance will have an important impact on future war modes. Although the aviation cluster network has many advantages, its practical application still faces some technical challenges.

[0003] Most of the current aviation cluster networks adopt the traditional distributed network architecture. However, the rapid movement of drones will cause the network topology to change frequently, and node links are prone to interruption, which cannot meet the flexible combat needs of future aviation cluster networks. Summary of the invention

[0004] Purpose of the present invention: An embodiment of the present invention provides an aviation cluster network routing method to solve the problem that the rapid movement of drones in existing aviation cluster networks will cause frequent changes in network topology, making it difficult to meet the flexible combat needs of future aviation cluster networks.

[0005] Technical solution of the present invention: An embodiment of the present invention provides an aviation cluster network routing method, comprising:

[0006] Step 1, dynamically adjusting the time interval for the software defined network controller node to send the link layer discovery protocol LLDP message according to the location information and movement information of the aviation cluster network node;

[0007] Step 2: Calculate the performance evaluation value of each link according to the bandwidth and transmission delay of each link;

[0008] Step 3: According to the performance evaluation value of each link, select the optimal routing path for data forwarding.

[0009] Optionally, in the aviation cluster network routing method as described above, step 1 comprises:

[0010] Step 11, the controller node i obtains the three-dimensional position information of the controller node i and each switch node j in real time, and calculates the node spacing and relative movement speed between the controller node i and each switch node j in real time;

[0011] Step 12: The controller node i calculates the link holding time between the controller node i and each switch node j in real time according to the three-dimensional position information, node spacing and relative movement speed, and uses the minimum link holding time as the time interval for the controller node to periodically send LLDP messages.

[0012] Optionally, in the aviation cluster network routing method as described above, step 11 includes:

[0013] Step 11a, the controller node sends two consecutive time slots (t1, t2) of LLDP messages to the SDN switch node in a broadcast manner, so that each switch node that receives the LLDP message sends a node response message to the controller node, and the node response message includes the location, movement speed, trajectory and direction information of the switch node;

[0014] Step 11b, controller node i reads the three-dimensional position information (a) of controller node i and each switch node j in real time from the node response message. i ,b i ,c i ),(a j ,b j ,c j ), and calculates in real time:

[0015] (1) The node distance between controller node i and each switch node j is:

[0016]

[0017] (2) The relative movement speed between controller node i and each switch node is:

[0018]

[0019] Among them, a i ,b i ,c i is the three-dimensional coordinate information of controller node i at time t, a j ,b j ,c j is the three-dimensional coordinate information of switch node j at time, is the speed difference between switch node j and controller node i in the x direction, is the speed difference between switch node j and controller node i in the y direction, is the speed difference between switch node j and controller node i in the z direction.

[0020] Optionally, in the aviation cluster network routing method as described above, step 12 includes:

[0021] Step 12a: The controller node projects the node motion graph in the three-dimensional space onto the relative motion plane where the controller node i and the switch node j are located according to the node spacing and the relative motion speed, and calculates the vertical distance between the controller node i and the motion path of the adjacent switch node j as: in, is the relative position vector from switch node j to controller node i;

[0022] Step 12b: The controller node calculates the change in node spacing from t1 to t2 based on the node spacing and relative motion speed: ΔS = S i,j (t2)-S i,j (t1); where S i,j (t1) is the distance between controller node i and switch node j at time t1, S i,j (t2) is the distance between controller node i and switch node j at time t2;

[0023] Step 12c, calculate the link holding time between the controller node i and each switch node j in real time. The link holding time when the two nodes are moving in opposite directions is: If two nodes move towards each other, the link holding time is Where R is the signal coverage range transmitted by controller node i;

[0024] Step 12d: Controller node i calculates the link holding time of all switch nodes directly connected to it, and selects the minimum link holding time T min (i,j)=Min(T(i,j)),j∈neighbor(i), the minimum link holding time is used as the time interval T for the controller node to periodically send LLDP messages lldp =T min (i,j).

[0025] Optionally, in the aviation cluster network routing method as described above, step 2 includes:

[0026] Step 21, the switch node directly connected to the controller node i receives and forwards the LLDP message, and establishes links between all switch nodes;

[0027] Step 22, each switch node transmits its neighbor switch node list, adjacent link bandwidth and transmission delay to the controller node, so that the controller node creates an overall topology view relationship of the aviation cluster network according to the received information;

[0028] In step 23, the controller node calculates the routing metric of each link as a performance evaluation value by normalizing the bandwidth and transmission delay of each link.

[0029] Optionally, in the aviation cluster network routing method as described above, step 23 includes:

[0030] Step 23a: The controller node normalizes the link bandwidth to obtain the bandwidth of each link: Among them, B k is the actual bandwidth of link k, is the minimum bandwidth of link k, is the maximum value of the bandwidth of link k;

[0031] Step 23b: The controller node normalizes the link transmission delay to obtain the transmission delay of each link: Among them, D k is the actual bandwidth of link k, is the minimum bandwidth of link k, is the maximum value of the bandwidth of link k;

[0032] Step 23c, calculate the routing metric R of each link k =w B ×B+w D ×D, where R k represents the routing metric of link k, w B 、w D is the weight vector w B 、w D , and satisfies w B +w D =1.

[0033] Optionally, in the aviation cluster network routing method as described above, step 3 includes:

[0034] Step 31, the source switch node of the data to be forwarded sends a routing request message containing the destination switch node to the controller node;

[0035] Step 32: The controller node calculates the routing metric R of the relevant link. k , selecting an optimal routing path from the source switch node to the destination switch node, so that the source switch node adopts the optimal routing path to forward data packets.

[0036] Optionally, in the aviation cluster network routing method described above, the controller node in step 32 selects the optimal routing path in the following manner:

[0037] The routing metric R of the link associated with the source switch node and the destination switch node k As a reward function, the Q-learning algorithm is used to select the optimal routing path from the source switch node to the destination switch node.

[0038] Beneficial effects of the invention: The embodiment of the invention proposes an aviation cluster network routing method, and its main implementation steps include: first, in the network topology discovery stage, dynamically adjust the sending interval of LLDP messages in the SDN controller according to the node movement speed and direction to improve the communication link efficiency; then, in the routing stage, calculate the link performance evaluation value according to each link bandwidth and transmission delay, and use it as the reward function in the Q learning algorithm to select the optimal routing path that can guarantee the service quality (QoS). The routing method of this method makes full use of the movement and link information of the node, designs a routing method that adapts to the node behavior, can improve the link bandwidth utilization efficiency, and can better meet the time-sensitive dynamic aviation cluster network requirements, so as to achieve the purpose of adapting to the mobility of the aviation cluster network and improving the service transmission performance.

[0039] Using the aviation cluster network routing method provided by the embodiment of the present invention, the SDN controller sends LLDP messages at fixed time intervals to discover and maintain the network topology. However, if the fixed LLDP message time interval is too short, bandwidth waste will increase, or if the interval is too long, the delay in the topology discovery process will increase. This makes the topology information stored in the SDN control node possibly inaccurate, and the routing path calculated according to this network topology does not conform to the real-time state of the network node, reducing the efficiency of message transmission. Therefore, in the aviation cluster network routing method proposed by the present invention, the sending time interval of the SDN controller LLDP message is dynamically adjusted according to the node movement speed and direction, etc., which can improve the bandwidth utilization efficiency of the communication link and more accurately construct the network topology; in the routing stage, the routing metric that guarantees the service QoS is designed according to the bandwidth and delay of each link, and the machine learning algorithm is used to select the path that can obtain the best service transmission quality, which can better meet the time-sensitive dynamic airborne network requirements. Conventional SDN routing only makes routing decisions based on the number of node hops, and it is difficult to ensure the quality of service transmission.

[0040] Furthermore, the technical solution of the embodiment of the present invention will play an important role in ensuring the link bandwidth utilization efficiency of wireless communication networking through adaptive improvement, provide a certain basis for improving the development level of wireless networks, and have significant economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings are used to provide a further understanding of the technical solution of the present invention and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the technical solution of the present invention and do not constitute a limitation on the technical solution of the present invention.

[0042] Figure 1 A flow chart of an aviation cluster network routing method provided by an embodiment of the present invention;

[0043] Figure 2 A flowchart of an aviation cluster network routing method provided for a specific embodiment of the present invention;

[0044] Figure 3 It is a schematic diagram of an aviation cluster network with sparsely distributed nodes in a specific embodiment of the present invention;

[0045] Figure 4 Schematic diagram of the relationship between the node distance and the relative motion speed between node i and node j in an embodiment of the present invention;

[0046] Figure 5 It is a schematic diagram of projecting a node motion graph in a three-dimensional space onto a relative motion plane where two nodes are located in an embodiment of the present invention;

[0047] Figure 6 This is the overall topology view of the aviation cluster network created by the controller according to the node information in the embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present invention more clear, the embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other arbitrarily without conflict.

[0049] The steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions. Also, although a logical sequence is shown in the flowchart, in some cases, the steps shown or described can be performed in a sequence different from that shown here.

[0050] The above background technology has explained that most of the current aviation cluster networks adopt the traditional distributed network architecture. However, due to the rapid movement of drones, the network topology will change frequently, and node links are prone to interruption, which cannot meet the flexible combat needs of future aviation cluster networks.

[0051] To solve this problem, based on the centralized control characteristics of software-defined networks (SDN), software-defined aviation cluster networks are designed to obtain global network information, such as link information, network topology, etc., to optimize network routing performance, thereby reducing the impact of failures, and have good application prospects in the field of airborne communication networks. However, due to the high dynamic and dense nodes of aviation cluster networks, direct application of SDN technology to aviation cluster networks still requires solving some key problems.

[0052] One of the key issues in applying SDN technology to aviation cluster networks is the need to study reliable and efficient routing protocols, because SDN network routing mainly uses the Link Layer Discovery Protocol (LLDP) data packets sent by the controller to establish the local routing table of the node. In order to ensure that the controller can obtain complete network topology information, LLDP messages are generally sent periodically to detect network changes. Although the implementation of fixed LLDP interval detection of neighbor nodes is simple, this method cannot guarantee accurate detection of cluster nodes in highly dynamic aviation cluster networks. Therefore, the aviation cluster network routing method provided by the embodiment of the present invention takes into account that in aviation cluster network applications, the flight path of the node is usually pre-planned, and its position and motion information can be obtained through the positioning system and inertial sensors. These auxiliary information can be used to calculate the movement trend of the aviation cluster network nodes.

[0053] Therefore, based on the movement and location information of network nodes, the present invention proposes a software-defined aviation cluster network routing method, and its main implementation steps include: first, in the network topology discovery phase, dynamically adjust the sending interval of LLDP messages in the SDN controller according to the node movement speed and direction to improve the communication link efficiency; then, in the routing phase, calculate the link performance evaluation value according to each link bandwidth and transmission delay, and use it as the reward function in the Q learning algorithm to select the optimal routing path that can guarantee the service quality (QoS). The routing method of this method makes full use of the movement and link information of the node, designs a routing method that adapts to the node behavior, can improve the link bandwidth utilization efficiency, and can better meet the time-sensitive dynamic aviation cluster network requirements, so as to adapt to the mobility of the aviation cluster network and improve the service transmission performance.

[0054] The present invention provides the following specific embodiments which can be combined with each other. The same or similar concepts or processes may not be described in detail in some embodiments.

[0055] Figure 1 A flowchart of an aviation cluster network routing method provided by an embodiment of the present invention. The technical problem to be solved by the embodiment of the present invention is: how to adaptively adjust the time interval for the SDN controller to send LLDP messages according to the location and movement information of the aviation cluster network nodes, and then use the link performance evaluation value to select the optimal path to make routing decisions to ensure service quality (QoS). Based on the above technical problems, the aviation cluster network routing method provided by the embodiment of the present invention specifically includes the following execution steps:

[0056] Step 1, dynamically adjusting the time interval for the software defined network controller node to send the link layer discovery protocol LLDP message according to the location information and movement information of the aviation cluster network node;

[0057] Step 2: Calculate the performance evaluation value of each link according to the bandwidth and transmission delay of each link;

[0058] Step 3: According to the performance evaluation value of each link, select the optimal routing path for data forwarding.

[0059] In the embodiment of the present invention, the specific implementation of step 1 may include:

[0060] Step 11, the controller node i obtains the three-dimensional position information of the controller node i and each switch node j in real time, and calculates the node spacing and relative movement speed between the controller node i and each switch node j in real time;

[0061] Step 12: The controller node i calculates the link holding time between the controller node i and each switch node j in real time according to the three-dimensional position information, node spacing and relative movement speed, and uses the minimum link holding time as the time interval for the controller node to periodically send LLDP messages.

[0062] In an implementation of the embodiment of the present invention, the specific implementation process of the above step 11 may include:

[0063] Step 11a, the controller node sends two consecutive time slots (t1, t2) of LLDP messages to the SDN switch node in a broadcast manner, so that each switch node that receives the LLDP message sends a node response message to the controller node, and the node response message includes the location, movement speed, trajectory and direction information of the switch node;

[0064] Step 11b, controller node i reads the three-dimensional position information (a) of controller node i and each switch node j in real time from the node response message. i ,b i ,c i ),(a j ,b j ,c j ), and calculates in real time:

[0065] (1) The node distance between controller node i and each switch node j is:

[0066]

[0067] (2) The relative movement speed between controller node i and each switch node is:

[0068]

[0069] Among them, a i ,b i ,c iis the three-dimensional coordinate information of controller node i at time t, a j ,b j ,c j is the three-dimensional coordinate information of switch node j at time, is the speed difference between switch node j and controller node i in the x direction, is the speed difference between switch node j and controller node i in the y direction, is the speed difference between switch node j and controller node i in the z direction.

[0070] In one implementation of the embodiment of the present invention, the specific implementation process of the above step 12 may be:

[0071] Step 12a: The controller node projects the node motion graph in the three-dimensional space onto the relative motion plane where the controller node i and the switch node j are located according to the node spacing and the relative motion speed, and calculates the vertical distance between the controller node i and the motion path of the adjacent switch node j as: in, is the relative position vector from switch node j to controller node i;

[0072] Step 12b: The controller node calculates the change in node spacing from t1 to t2 based on the node spacing and relative motion speed: ΔS = S i,j (t2)-S i,j (t1); where S i,j (t1) is the distance between controller node i and switch node j at time t1, S i,j (t2) is the distance between controller node i and switch node j at time t2;

[0073] Step 12c, calculate the link holding time between the controller node i and each switch node j in real time. The link holding time when the two nodes are moving in opposite directions is: If two nodes move towards each other, the link holding time is Where R is the signal coverage range transmitted by controller node i;

[0074] Step 12d: Controller node i calculates the link holding time of all switch nodes directly connected to it, and selects the minimum link holding time T min (i,j)=Min(T(i,j)),j∈neighbor(i), the minimum link holding time is used as the time interval T for the controller node to periodically send LLDP messages lldp =T min (i,j).

[0075] In an embodiment of the present invention, the specific implementation of step 2 may include:

[0076] Step 21, the switch node directly connected to the controller node i receives and forwards the LLDP message, and establishes links between all switch nodes;

[0077] Step 22, each switch node transmits its neighbor switch node list, adjacent link bandwidth and transmission delay to the controller node, so that the controller node creates an overall topology view relationship of the aviation cluster network according to the received information;

[0078] In step 23, the controller node calculates the routing metric of each link as a performance evaluation value by normalizing the bandwidth and transmission delay of each link.

[0079] In an implementation of the embodiment of the present invention, the specific implementation process of the above step 23 may include:

[0080] Step 23a: The controller node normalizes the link bandwidth to obtain the bandwidth of each link: Among them, B k is the actual bandwidth of link k, is the minimum bandwidth of link k, is the maximum value of the bandwidth of link k;

[0081] Step 23b: The controller node normalizes the link transmission delay to obtain the transmission delay of each link: Among them, D k is the actual bandwidth of link k, is the minimum bandwidth of link k, is the maximum value of the bandwidth of link k;

[0082] Step 23c, calculate the routing metric R of each link k =w B ×B+w D ×D, where R k represents the routing metric of link k, w B 、w D is the weight vector w B 、w D , and satisfies w B +w D =1.

[0083] In an embodiment of the present invention, the specific implementation of the above step 3 may include:

[0084] Step 31, the source switch node of the data to be forwarded sends a routing request message containing the destination switch node to the controller node;

[0085] Step 32: The controller node calculates the routing metric R of the relevant link.k , selecting an optimal routing path from the source switch node to the destination switch node, so that the source switch node adopts the optimal routing path to forward data packets.

[0086] In an implementation of the embodiment of the present invention, the controller node in step 32 may select the optimal routing path in the following manner:

[0087] The routing metric R of the link associated with the source switch node and the destination switch node k As a reward function, the Q-learning algorithm is used to select the optimal routing path from the source switch node to the destination switch node.

[0088] The basic principle of the aviation cluster network routing method provided by the embodiment of the present invention is: in the aviation cluster network, each node can obtain its position, trajectory, direction, speed and other information through the positioning system; in the network topology discovery stage, the time interval for the controller to send LLDP messages is dynamically adjusted according to the node movement speed and direction to improve the communication link efficiency; in the routing stage, the link performance evaluation value is calculated according to the bandwidth and transmission delay of each link and used as a reward function, and the Q reinforcement learning algorithm is used to select the optimal routing path, which can better meet the time-sensitive dynamic airborne network requirements.

[0089] Using the aviation cluster network routing method provided by the embodiment of the present invention, the SDN controller sends LLDP messages at fixed time intervals to discover and maintain the network topology. However, if the fixed LLDP message time interval is too short, bandwidth waste will increase, or if the interval is too long, the delay in the topology discovery process will increase. This makes the topology information stored in the SDN control node possibly inaccurate, and the routing path calculated according to this network topology does not conform to the real-time state of the network node, reducing the efficiency of message transmission. Therefore, in the aviation cluster network routing method proposed by the present invention, the sending time interval of the SDN controller LLDP message is dynamically adjusted according to the node movement speed and direction, etc., which can improve the bandwidth utilization efficiency of the communication link and more accurately construct the network topology; in the routing stage, the routing metric that guarantees the service QoS is designed according to the bandwidth and delay of each link, and the machine learning algorithm is used to select the path that can obtain the best service transmission quality, which can better meet the time-sensitive dynamic airborne network requirements. Conventional SDN routing only makes routing decisions based on the number of node hops, and it is difficult to ensure the quality of service transmission.

[0090] Furthermore, the technical solution of the embodiment of the present invention will play an important role in ensuring the link bandwidth utilization efficiency of wireless communication networking through adaptive improvement, provide a certain basis for improving the development level of wireless networks, and have significant economic benefits.

[0091] The implementation of the aviation cluster network routing method provided by the embodiment of the present invention is described in detail below through some specific embodiments.

[0092] Figure 2 A flowchart of an aviation cluster network routing method provided by a specific embodiment of the present invention, Figure 3 FIG. 1 is a schematic diagram of an aviation cluster network with sparsely distributed nodes in a specific embodiment of the present invention. Figure 2 The aviation cluster network with sparsely distributed nodes shown is a connected network composed of multiple nodes that can communicate with each other within the coverage of the radio station, and the nodes can transmit data by establishing a multi-hop wireless network. Assume that node 4 is configured with an SDN controller, and the other nodes are SDN switches installed with the Openflow protocol. The aviation cluster network routing method provided in this specific embodiment may include the following implementation steps:

[0093] Step (1): The SDN controller of the aviation cluster network broadcasts two consecutive time slots (t1, t2) of LLDP messages to the SDN switch nodes.

[0094] Specifically, the LLDP message includes: 1) the primary address of the node; 2) the current location, movement speed, trajectory and direction of the node; 3) the address of the neighboring node; 4) the link maintenance timer value with the neighboring node, etc.

[0095] Step (2): Each SDN switch node that receives the LLDP message sends a node response message to the SDN controller, and transmits the location, movement speed, trajectory and direction information obtained by the node positioning to the SDN controller.

[0096] Step (3): SDN controller node i reads the three-dimensional position information (a) of SDN controller node i and corresponding switch node j from the node response message. i ,b i ,c i ),(a j ,b j ,c j ), and calculate the node distance between controller i and switch node j at time t as:

[0097] The relative speed of motion is: like Figure 4 , which is a schematic diagram of the relationship between the node distance and the relative movement speed between the node i and the node j in an embodiment of the present invention.

[0098] Step (4): The controller node projects the node motion graph in the three-dimensional space onto the relative motion plane where the two nodes are located based on the node spacing and relative motion speed calculated in step (3), and calculates the distance between node i and the motion path of the adjacent switch node j perpendicularly; like Figure 5 As shown, it is a schematic diagram of projecting a node motion graph in a three-dimensional space onto a relative motion plane where two nodes are located in an embodiment of the present invention.

[0099] Step (5): Based on the node spacing and relative movement speed, calculate the node spacing change from t1 to t2 as ΔS = S i,j (t2)-S i,j (t1), where ΔS is greater than zero, indicating that the two nodes are moving in opposite directions, and ΔS is less than zero, indicating that the two nodes are moving towards each other.

[0100] For example, the change in node spacing from t1 to t2 is ΔS=1000-600=400m, and the two nodes move in opposite directions.

[0101] Step (6): Calculate the link holding time between nodes i and j. If the two nodes are moving in opposite directions, the link holding time is In this specific embodiment, it is assumed that the calculated link holding time T = 16 seconds; in addition, if the two nodes are moving towards each other, the link holding time Wherein, R is the signal coverage range transmitted by controller node i.

[0102] Step (7): Controller node i calculates the link holding time of all switch nodes directly connected to it, and selects the minimum link holding time as T min (i, j) = Min (T (i, j)), where j ∈ neighbor (i), and the time interval for the SDN controller node to periodically send LLDP messages is T lldp =T min (i, j), in this specific embodiment, it is assumed that the selected LLDP message time interval is 15 seconds.

[0103] After the above steps, the LLDP message sending interval can be dynamically adjusted according to the node movement speed and direction, which is different from the traditional fixed sending interval mode. For example, the original fixed = 10 seconds, after dynamic adjustment = 15 seconds, etc., the LLDP message sending cycle is extended, so that more bandwidth resources are used to send data, improving the network bandwidth utilization efficiency.

[0104] Step (8): The switch nodes directly connected to the controller node i forward the LLDP message and establish links between all switch nodes.

[0105] Step (9): Each switch node transmits its neighbor switch node list, adjacent link bandwidth and transmission delay to the SDN controller. The controller receives the switch node information and creates an overall topology view of the aviation cluster network. Figure 6 As shown, it is the overall topology view of the aviation cluster network created by the controller according to the node information in an embodiment of the present invention.

[0106] Step (10): On the controller side, normalize the bandwidth of each link to Among them B k is the actual bandwidth of link k, is the minimum link bandwidth, The maximum link bandwidth.

[0107] Step (11): On the controller side, normalize the transmission delay of each link to Where D k is the actual bandwidth of link k, is the minimum link bandwidth, D k max The maximum value of the link bandwidth.

[0108] Step (12): On the controller side, calculate the routing metric of link k as R k =w B ×B+w D ×D, where the weight vector w B , w D It can be adjusted according to network performance requirements and meet w B +w D =1.

[0109] For example, for the link between nodes 10 and 11, we calculate B = 0.6, D = 0.8, and select w B =1 / 2, w D =1 / 2, then we can calculate R k =0.7.

[0110] Step (13): When a switch node in the network (for example, network node 8) needs to forward data, it sends a routing request message containing the destination switch node to the SDN controller.

[0111] Step (14): On the SDN controller side, the routing metric R of link k (including the above-mentioned source node and target node) is k As a reward function, the Q learning algorithm is used to intelligently select the optimal routing path from the source node to the destination node and send it to the routing request node (ie, the source node).

[0112] For example, after Q learning, the selected path is 8→6→9→11→10.

[0113] Step (15): The cluster network source node forwards data packets using the obtained optimal routing path; corresponding to this specific embodiment, the cluster network source node 8 forwards data packets according to the obtained routing message (8→6→9→11→10).

[0114] Although the embodiments disclosed in the present invention are as above, the contents described are only embodiments adopted to facilitate understanding of the present invention and are not intended to limit the present invention. Any technician in the field to which the present invention belongs can make any modifications and changes in the form and details of implementation without departing from the spirit and scope disclosed in the present invention, but the patent protection scope of the present invention shall still be subject to the scope defined in the attached claims.

Claims

1. An aviation cluster network routing method, characterized in that: The following steps are involved: Step 1, dynamically adjusting the time interval for the software defined network controller node to send the link layer discovery protocol LLDP message according to the location information and movement information of the aviation cluster network node; Step 2: Calculate the performance evaluation value of each link according to the bandwidth and transmission delay of each link; Step 3: Select the optimal routing path for data forwarding based on the performance evaluation value of each link; Wherein, the step 1 comprises: Step 11, the controller node i obtains the three-dimensional position information of the controller node i and each switch node j in real time, and calculates the node spacing and relative movement speed between the controller node i and each switch node j in real time; Step 12, the controller node i calculates the link holding time between the controller node i and each switch node j in real time according to the three-dimensional position information, the node spacing and the relative movement speed, and uses the minimum link holding time as the time interval for the controller node to periodically send LLDP messages; The step 11 comprises: Step 11a, the controller node sends two consecutive time slots (t1, t2) of LLDP messages to the SDN switch node in a broadcast manner, so that each switch node that receives the LLDP message sends a node response message to the controller node, and the node response message includes the location, movement speed, trajectory and direction information of the switch node; Step 11b, controller node i reads the three-dimensional position information (a) of controller node i and each switch node j in real time from the node response message. i ,b i ,c i ),(a j ,b j ,c j ), and calculates in real time: (1) The node distance between controller node i and each switch node j is: (2) The relative movement speed between controller node i and each switch node is: Among them, a i ,b i ,c i is the three-dimensional coordinate information of controller node i at time t, a j ,b j ,c j is the three-dimensional coordinate information of switch node j at time, is the speed difference between switch node j and controller node i in the x direction, is the speed difference between switch node j and controller node i in the y direction, is the speed difference between switch node j and controller node i in the z direction; The step 12 comprises: Step 12a: The controller node projects the node motion graph in the three-dimensional space onto the relative motion plane where the controller node i and the switch node j are located according to the node spacing and the relative motion speed, and calculates the vertical distance between the controller node i and the motion path of the adjacent switch node j as: in, is the relative position vector from switch node j to controller node i; Step 12b: The controller node calculates the change in node spacing from t1 to t2 based on the node spacing and relative motion speed: ΔS = S i,j (t2)-S i,j (t1); where S i,j (t1) is the distance between controller node i and switch node j at time t1, S i,j (t2) is the distance between controller node i and switch node j at time t2; Step 12c, calculate the link holding time between the controller node i and each switch node j in real time. The link holding time when the two nodes are moving in opposite directions is: If two nodes move towards each other, the link holding time is Where R is the signal coverage range transmitted by controller node i; Step 12d: Controller node i calculates the link holding time of all switch nodes directly connected to it, and selects the minimum link holding time T min (i,j)=Min(T(i,j)),j∈neighbor(i), the minimum link holding time is used as the time interval T for the controller node to periodically send LLDP messages lldp =T min (i,j).

2. The aviation cluster network routing method according to claim 1, characterized in that: The step 2 comprises: Step 21, the switch node directly connected to the controller node i receives and forwards the LLDP message, and establishes links between all switch nodes; Step 22, each switch node transmits its neighbor switch node list, adjacent link bandwidth and transmission delay to the controller node, so that the controller node creates an overall topology view relationship of the aviation cluster network according to the received information; In step 23, the controller node calculates the routing metric of each link as a performance evaluation value by normalizing the bandwidth and transmission delay of each link.

3. The aviation cluster network routing method according to claim 2, characterized in that: The step 23 comprises: Step 23a: The controller node normalizes the link bandwidth to obtain the bandwidth of each link: Among them, B k is the actual bandwidth of link k, is the minimum bandwidth of link k, is the maximum value of the bandwidth of link k; Step 23b: The controller node normalizes the link transmission delay to obtain the transmission delay of each link: Among them, D k is the actual bandwidth of link k, is the minimum bandwidth of link k, is the maximum value of the bandwidth of link k; Step 23c, calculate the routing metric R of each link k =w B ×B+w D ×D, where R k represents the routing metric of link k, w B 、w D is the weight vector w B 、w D , and satisfies w B +w D =1.

4. The aviation cluster network routing method according to claim 3, characterized in that: The step 3 comprises: Step 31, the source switch node of the data to be forwarded sends a routing request message containing the destination switch node to the controller node; Step 32: The controller node calculates the routing metric R of the relevant link. k , selecting an optimal routing path from the source switch node to the destination switch node, so that the source switch node adopts the optimal routing path to forward data packets.

5. The aviation cluster network routing method according to claim 4, characterized in that: The way in which the controller node selects the optimal routing path in step 32 is: The routing metric R of the link associated with the source switch node and the destination switch node k As a reward function, the Q-learning algorithm is used to select the optimal routing path from the source switch node to the destination switch node.