Complex and variable-oriented adaptive network generation method and system

Abstract

The invention discloses a method and a system for generating a complex and variable-oriented adaptive network. The method comprises the following steps of: constructing a cross-layer routing architecture consisting of a routing layer and an MAC (Media Access Control) layer; after the MAC layer obtains a first-hop neighbor and a second-hop neighbor, respective backbone nodes are selected in a distributed mode, flooding is conducted to the whole network, and declarations of the flooding backbone nodes are grouped and loaded into beacons; the routing layer starts to run on the backbone nodes, performs secondary relay selection on the backbone nodes, notifies the positions of the relay nodes of the whole virtual backbone network, and establishes a local topology table; meanwhile, a zero-interval triggering type cross-layer cooperation flooding algorithm and a constraint flooding algorithm based on fuzzy vision are provided; and finally, selecting an optimal routing path by taking the media occupancy time as a scale for measuring the advantages and disadvantages of the path. According to the invention, the network topology structure can be adaptively, rapidly and automatically adjusted according to the task and node state, the network delay and overhead are reduced, and rapid routing selection and maintenance can be realized on a large scale.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication networking, in particular to a complex adaptive network generation method and system. BACKGROUND

[0002] With the development of wireless communication networking technology and the large-scale investment of new equipment, the network scale is increasing, and the network structure is becoming more and more complex, especially the networking of unmanned platforms and mobile ad hoc networks. The position and task demand of nodes will change frequently. The traditional network generation method based on static network model will cause network delay to be too long, routing overhead to be too large and seriously affect network performance when applied to complex and variable scenarios such as large scale and high dynamics.

[0003] The prior art has the following technical problems to be solved: First, the routing method has the problems of low efficiency and poor reliability; the ad hoc network topology establishment process is passive, the establishment process has large overhead, and the generated topology is complex (such as many hops and few key hotspot paths), which leads to the defects of low efficiency and poor reliability of the subsequent routing method; Second, the network has large overhead; the topology grouping transmission time is too long in large-scale ad hoc networks, the flooding information is easy to expire, and when the end nodes are far away, the message will spread to the whole network, causing unnecessary overhead. SUMMARY

[0004] The technical problem to be solved by the present application is to optimize the network topology structure in large-scale, high-dynamic and high-density application scenarios, and to provide a complex adaptive network generation method and system. Based on the cross-layer routing architecture design, MAC layer backbone node selection, zero interval trigger cross-layer cooperation flooding, and fuzzy vision based constraint flooding algorithm are proposed to realize fast routing update with the lowest routing maintenance overhead and ensure the adaptive optimization of network topology structure.

[0005] The technical solution adopted by the present application to solve the technical problem is: The present application provides a complex adaptive network generation method, which comprises: Constructing a cross-layer routing architecture composed of a routing layer and a MAC layer, and designing a routing protocol in the MAC layer and the routing layer; After obtaining 1-hop neighbors and 2-hop neighbors, the MAC layer selects respective backbone nodes in a distributed manner, floods to the whole network, and loads the flooding backbone node declaration group into the beacon; The routing layer starts to operate on the backbone nodes, the routing layer performs secondary relay selection of the backbone nodes, informs the position of the relay nodes in the whole virtual backbone network, and establishes a local topology table; Adopting zero interval trigger cross-layer cooperation flooding algorithm: the flooding of relay node information is designed as the mode of slow periodic flooding combined with trigger flooding, and the update of topology information adopts two types of periodic update and trigger update; when the MAC layer judges that the key parameters of the node change, the influence on the 1-hop and 2-hop neighbors is judged, if the influence exceeds the set threshold value, a new relay node declaration packet is broadcasted by informing the routing layer; Adopting constraint flooding algorithm based on fuzzy vision: the spatial distance of topology update packet flooding is limited, that is, the node only sends the topology control message when detecting that the link state changes, and the TTL is used to limit the flooding range of the packet, and only the near-end node is sent; The media occupation time is taken as the scale of measuring the advantages and disadvantages of the path, and the optimal routing path is selected.

[0006] Further, the cross-layer routing architecture of the application specifically comprises: The MAC layer is used for realizing the discovery of network neighbors, the monitoring of network topology changes and the selection of backbone nodes; the routing layer is used for realizing topology change update, routing maintenance and fast routing selection; the flooding of topology information in the topology change update adopts the mode of slow periodic flooding combined with trigger flooding; the MAC layer and the routing layer adopt linkage design; the space signal layer is further arranged outside the cross-layer routing architecture, and the space signal layer reports the link adaptive parameters to the MAC layer, and the MAC layer determines the MCS value of the link.

[0007] Further, the method for distributing to select respective backbone nodes of the application specifically comprises: The backbone node selection algorithm adopts the distributed domination connected graph algorithm, that is, the UCDS algorithm, the UCDS algorithm generates a UCDS set to constitute a virtual backbone network of the whole network; The UCDS algorithm specifically is: generating the UCDS set based on the UCDS rule, that is, the union of the domination set DS and the connected set CS; the UCDS rule marks each possible DS and CS, the DS is the domination set of the network topology graph, and the CS is the connected set of the network topology graph, then the redundant nodes are removed from the CDS, and the UCDS rule is used in a certain order, that is, DS rule-->CS candidate rule-->CS exception rule-->CS rule, and the UCDS set is correctly selected.

[0008] Further, the UCDS algorithm of the application specifically comprises: 1) DS rule: A certain node i in the topology graph G has the highest domination factor d among all neighbor nodes j of the node i i , then the node i specifies itself as a DS member; it is also possible that the node i has the highest domination factor d among all neighbors k of the neighbor nodes j of the node i i, then node j designates node i as a DS member;

[0009] 2) CS Candidate Rules: For a non-DS member node i, there are neighbor nodes j and k, at least one of which is a DS member, and their DS neighbor node sets are disjoint, then node i is a candidate member of CS ; If there are multiple common neighbor nodes of nodes j and k , then it has the highest dominance factor d i of Members will be selected as CS candidate members;

[0010]

[0011] 3) CS exception rules: For non-DS member node i, if there is a CS member among the common neighbors of neighbor nodes j and k, the CS rule does not apply. If there is only one non-DS member among the neighbor nodes j and k, and node i and the non-DS member have a common DS neighbor node, the CS rule does not apply either. Usage conditions: If non-DS member node i becomes a CS member because another pair of neighbor nodes l and m satisfy the CS rule, the CS exception rule does not apply;

[0012] 4) Non-CS rules: For a non-DS member node i, if any of its neighbor nodes j and k are directly connected, then node i is definitely not a CS member;

[0013] Among them, DS is the dominating set of the network topology graph, and CS is the connected set of the network topology graph; DSN k is the set of DS neighbor node identification numbers of node k, N i is the set of neighbor node identification numbers of node i, d i is the neighbor degree of node i, that is, the number of neighbors of the node, DSN j is the set of DS neighbor node identification numbers of node j, and I, j, k, l, and m refer to the node identification numbers in the network topology respectively.

[0014] Furthermore, the method for performing secondary backbone node secondary relay selection by the routing layer of the present invention includes: When the MAC layer selects the backbone node and informs the whole network of the backbone node and the normal nodes hung thereunder, the routing layer starts to operate on the backbone node; the backbone node routing is realized by using a link state type routing protocol algorithm, including: 1) backbone node neighbor discovery, since the MAC layer performs 1-hop neighbor and 2-hop neighbor discovery of all nodes, 1-hop backbone neighbors and 2-hop backbone neighbors of all backbone nodes are obtained by default; 2) secondary relay selection of the backbone node, referring to the backbone node selection algorithm, the relay node of the backbone node is declared by using flooding, and the position of the relay node in the whole virtual backbone network is informed; 3) establishing a local topology table; 4) finding a route.

[0015] Further, in the secondary relay selection of the backbone node, the backbone node selection algorithm refers to the MAC backbone node selection algorithm, each node in the virtual backbone network independently selects its own relay node set, so that all data control packets are forwarded through the relay node set, and through the forwarding, the control packets sent by the node can be transmitted to all two-hop neighbors; The requirement for calculating the relay node set is that the node and the relay node must be a bidirectional symmetric link, and the group sent by the node can reach all strict two-hop neighbor nodes through the forwarding of the relay node.

[0016] Further, the zero interval trigger type cross-layer cooperation flooding algorithm specifically includes: The flooding of the relay node information is designed as a mode combining slow periodic flooding and trigger type flooding, and the update of the topology information is divided into two types of periodic update and trigger type update; the periodic update is used for ensuring that the routing information in the routing table is not deleted due to timeout, and the trigger type update is used for ensuring that the network topology change can be quickly flooded to the whole network, and the reaction time of the routing is improved; When the MAC layer judges that the key parameters of the node change, including neighbor nodes, link state, queue exception, the influence on the 1-hop and 2-hop neighbors is judged, if the influence exceeds the set threshold value, the routing layer is informed to broadcast a new relay node declaration packet; The time frequency of the flooding is only related to the change frequency of the network, for the self-organizing network, the topology change depends on the moving speed and distance of the platform, and only when the platform moves more than N MCS intervals is the update of the topology triggered.

[0017] Further, the constraint flooding algorithm based on fuzzy vision specifically includes: The constraint flooding algorithm based on fuzzy vision is used to limit the spatial distance of the topology update packet flooding, that is, the node only sends the topology control message when detecting the link state change, and the TTL is used to limit the flooding range of the TC packet, and only the near-end node is sent; the distribution rule is as follows: assuming that the counter is counter=N*2i-1; 1) At the time of te, counter=1, at this time N=1, i=1, the node detects that the link change occurs in the previous 21-1te time, at this time the topology change message with TTL=2l is sent; 2) At the time of 2te, counter=2, at this time N=1, i=2, the node detects that the link change occurs in the previous 22-1te time, at this time the topology change message with TTL=22 is sent; 3) At the time of 3te, counter=3, at this time N=3, i=1, the node detects that no link change occurs in the previous 2i-1te time, at this time the topology change message is not sent; thereafter, the same is true, so as to limit the propagation range of the topology change message.

[0018] Further, the method for selecting the optimal routing path specifically comprises: The media occupation time, that is, the transmission time, is taken as the scale for measuring the advantages and disadvantages of the path, and the optimal routing path is selected from the routing table based on the minimum media occupation time.

[0019] The application provides a complex and variable adaptive network generation system, comprising: A memory for storing executable computer programs; A processor for executing the executable computer programs stored in the memory to realize the complex and variable adaptive network generation method.

[0020] The application has the following beneficial effects: (1) The cross-layer routing architecture places the functions of network neighbor discovery, network topology change monitoring and backbone node selection in the MAC layer, and the routing layer is only responsible for topology change update, routing maintenance and fast routing selection, and the MAC layer and the routing layer are designed in linkage, thereby bringing synchronous gain of overhead and delay, and being more suitable for large-scale, high-dynamic self-organizing networks compared with on-demand routing and proactive routing design. (2) Zero-interval trigger cross-layer cooperation flooding algorithm, supports dynamic adjustment of transmission strategy in the process of data packet transmission according to the current state of the network (such as channel quality, node energy state, etc.), such as selecting the optimal transmission path, adjusting the data transmission rate, etc., so as to reduce the data transmission delay and improve the data transmission success rate, at the same time, through the zero-interval trigger mechanism, it can effectively reduce the communication delay between nodes and improve the response speed of the network, which is suitable for wireless communication scenarios that require high real-time and reliability; (3) Constraint flooding algorithm based on fuzzy vision The connection state, signal strength, channel quality and other parameters between nodes are often dynamically changing and uncertain, the constraint flooding algorithm based on fuzzy vision supports processing multiple factors using fuzzy logic, avoiding the local optimal problem that may be caused by the traditional routing selection based on a single parameter (such as the shortest path), supporting intelligent processing of network uncertainty and improving the global optimality of routing selection, and providing more optimized strategy for data transmission. BRIEF DESCRIPTION OF DRAWINGS

[0021] The application will be further described below in combination with the drawings and examples, wherein: Figure 1 is a cross-layer routing architecture; Figure 2 is a backbone network constructed based on the UCDS algorithm; Figure 3 is a virtual backbone network routing schematic diagram; Figure 4 is a relay node selection algorithm flowchart in the backbone network; Figure 5 is a link throughput statistics based on media occupancy time routing metric. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical scheme and advantages of the application more clear, the application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the application and do not limit the application.

[0023] Example 1 As Figure 1As shown, the embodiment of the application comprehensively considers the dynamic nature of node position and task, and proposes a complex and variable adaptive network generation method, which can automatically adjust the network topology according to the task and node state, reduce network delay and overhead. The method is as follows: a cross-layer integrated design idea is adopted to cooperatively design the routing and the MAC layer (media access control layer), fully utilize the link information of the MAC layer to trigger the change and maintenance of the routing, and propose algorithms such as MAC layer backbone node selection, routing layer secondary backbone selection, zero interval trigger cross-layer cooperative flooding, fuzzy vision based constraint flooding, fast routing acquisition, and routing selection based on media occupation time in multi-hop scenario, which can meet the fast routing selection and maintenance in a large scale, reduce the overhead and delay. Specifically, it includes: (1) Cross-layer routing architecture design A cross-layer routing design is adopted to design the routing protocol in the media access control layer and the inline network layer, realize the transparent transmission of IP layer (Internet Protocol layer) data, and simultaneously design the MAC layer and the routing layer to bring synchronous gain in overhead and delay. The cross-layer routing architecture is as shown in Figure 1 .

[0024] (2) Backbone node selection algorithm After obtaining the 1-hop neighbors and 2-hop neighbors, the MAC layer selects the respective backbone nodes in a distributed manner, and floods to the whole network. The backbone node declaration information packet is also loaded into the beacon. The backbone node selection algorithm adopts a distributed domination graph algorithm, which can quickly and distributedly construct a virtual backbone network.

[0025] (3) Zero interval trigger cross-layer cooperative flooding algorithm The topology information (relay node information) flooding is designed as a combination of slow periodic flooding and trigger flooding mode. The update of the topology information is divided into periodic update and trigger update. The periodic update is used to ensure that the routing information in the routing table is not deleted due to timeout. The trigger update is used to ensure that the network topology changes can be quickly flooded to the whole network, thereby improving the reaction time of the routing. When the MAC layer judges that the key parameters of the node change (neighbor nodes, link state, queue exception, etc.), the influence on the 1-hop and 2-hop neighbors is judged. If the influence exceeds the set threshold value, a new relay node declaration packet is broadcasted by the routing layer. Thus, the time frequency of the topology information flooding is limited, and the time frequency of the flooding is only related to the change frequency of the network.

[0026] (4) Fuzzy vision based constraint flooding algorithm The fuzzy vision-based constraint flooding algorithm is used to limit the spatial distance of the topology update packet flooding, i.e. the node only sends the topology control message when detecting the link state change, and the TTL is used to limit the flooding range of the TC packet, only sent to the near nodes, further reducing the overhead.

[0027] The distribution rule is as follows: assuming that the counter is counter=N*2i-1. 1) At the time of te, counter=1, at this time N=1, i=1, the node detects that the link change occurs within the previous 21-1te time, at this time the topology change message with TTL=2l is sent.

[0028] 2) At the time of 2te, counter=2, at this time N=1, i=2, the node detects that the link change occurs within the previous 22-1te time, at this time the topology change message with TTL=22 is sent. 3) At the time of 3te, counter=3, at this time N=3, i=1, the node detects that no link change occurs within the previous 2i-1te time, at this time no topology change message is sent. Thereafter, this is followed to limit the propagation range of the topology change message.

[0029] Embodiment 2 In another specific embodiment of the present application, the specific implementation steps of the method are given: 1. Cross-layer routing architecture As shown in the figure, the cross-layer routing design is distributed in the media access control layer and the inline network layer (two-layer routing). The spatial signal layer reports the link adaptive parameters to the MAC layer, and the MAC layer determines the MCS value of the link according to the adaptive rate selection algorithm. Figure 1 2. Backbone node selection

[0030] After obtaining the 1-hop neighbors and 2-hop neighbors, the MAC layer distributes the selection of the respective backbone nodes, and floods to the entire network, and the flooding backbone node declaration information packet is also loaded into the beacon. The backbone node selection algorithm adopts a kind of distributed domination graph algorithm, which can generate a UCDS set (network node set) to constitute a virtual backbone network of the entire network, as shown in the figure. Figure 2

[0031] The UCDS rule embodies the UCDS algorithm idea, and can generate the UCDS set, i.e. the union set of the domination set DS and the connected set CS. The rule uses techniques such as shotgun method to mark each possible DS and CS, and then removes the redundant nodes from the CDS. The UCDS rule needs to be used in a certain order, i.e. DS rule-->CS candidate rule-->CS exception rule-->CS rule, to correctly select the UCDS set. As shown in the table:​​

[0032] UCDS algorithm flow includes: 1) DS rule: DS rule: a node i in the topology graph G has the highest domination factor di among all its neighbor nodes j, then node i designates itself as a DS member; it is also possible that node i has the highest domination factor di among all neighbors k of its neighbor node j, then node j designates node i as a DS member.

[0033]

[0034] 2) CS candidate rule: For a non-DS member node i, there exist neighbor nodes j and k, at least one of which is a DS member, and the DS neighbor node sets of the two are disjoint, then node i is a candidate member of CS ; if there are multiple in the common neighbor nodes of nodes j and k, the CS candidate member with the highest domination factor di will be selected.

[0035]

[0036]

[0037] 3) CS exception rule: For a non-DS member node i, when there is a CS member in the common neighbor nodes of neighbor nodes j and k, the CS rule does not need to be executed; when there is only one non-DS member in neighbor nodes j and k, and node i has a common DS neighbor node with the non-DS member, the CS rule also does not need to be executed.

[0038] Use condition: if non-DS member node i becomes a CS member due to another pair of neighbor nodes l and m satisfying the CS rule, the CS exception rule does not need to be executed.

[0039] 4) Non-CS rule: For a non-DS member node i, any of its neighbor nodes j and k are directly connected, then node i is definitely not a CS member.

[0040] 3. Virtual backbone routing When the MAC layer selects the backbone nodes and informs the entire network of the backbone nodes and the information of the ordinary nodes hanging below them, the routing layer starts to operate on the backbone nodes, as shown in Figure 3 .

[0041] The backbone node routing is implemented by using a link state type routing protocol algorithm. It is roughly divided into the following steps: 1) Backbone node neighbor discovery. Similarly, since the MAC layer performs 1-hop neighbor and 2-hop neighbor discovery of all nodes, the 1-hop backbone neighbor and 2-hop backbone neighbor of all backbone nodes are already obtained by default. This step can be omitted.

[0042] 2) Secondary relay selection of the backbone node. Referring to the backbone node selection algorithm, a simple flooding is used to declare the relay node of the backbone node, to inform the position of the relay node in the entire virtual backbone network, and to further reduce the routing and forwarding overhead.

[0043] 3) Establishing a local topology table.

[0044] 4) Finding a route.

[0045] (1) Secondary backbone network relay selection algorithm Similarly, referring to the MAC backbone node selection algorithm, each node in the virtual backbone network independently selects its own relay node set, so that all data control packets are forwarded through the relay node set. Through the forwarding, the control packets sent by the node can be transmitted to all two-hop neighbors.

[0046] The requirement for calculating the relay node set is that the node and the relay node must be a bidirectional symmetric link, and the packets sent by the node can reach all strict two-hop neighbor nodes through the forwarding of the relay node. The flowchart of the backbone network relay node selection algorithm is shown in Figure 4 .

[0047] (2) Relay flooding The relay node information of the virtual backbone network must be continuously flooded to the entire virtual backbone network to ensure that the topology is updated in time and effectively, and the speed of route convergence is ensured. However, even after the secondary relay node selection, the number of relay nodes is still very large relative to an infinite scale network. The relay node flooding is described as follows: 1) The range of relay node declaration packet flooding: the entire network.

[0048] 2) Nodes participating in flooding: nodes declared as relay nodes.

[0049] 3) The number of times of flooding: each relay node only forwards once.

[0050] It can be seen that the overhead of the flooding is directly related to the frequency of the flooding and the number of forwarding times of the flooding. Therefore, the patent proposes to limit the relay node declaration packet flooding in time and space.

[0051] Zero-interval trigger cross-layer cooperative flooding algorithm: The flooding of information in large diameter self-organizing network is that the topology information is easy to expire because of the long time of packet transmission. But in fact, the routing is not much related to the accuracy of the information of remote nodes, and the routing only concerns the next hop node, and when the data packet reaches the next hop node, the next hop node will reselect the route, and finally it will iterate to the vicinity of the destination node, and the routing flooding information in the vicinity of the destination node is timely and accurate. Therefore, the patent designs the flooding of topology information (relay node information) as a mode combining slow periodic flooding and triggered flooding. The update of topology information is divided into periodic update and triggered update. Periodic update is used to ensure that the routing information in the routing table is not deleted due to timeout. Triggered update is used to ensure that the network topology change can be quickly flooded to the whole network, thereby improving the reaction time of routing.

[0052] When the MAC layer judges that the key parameters of the node change (neighbor nodes, link state, queue exception, etc.), the influence on its 1-hop and 2-hop neighbors is judged, and if the influence exceeds the set threshold value, the routing layer is informed to generate a new relay node declaration packet and broadcast it.

[0053] Therefore, the time frequency of topology information flooding is limited, and the time frequency of flooding is only related to the frequency of network change. For self-organizing network, the change of topology depends on the moving speed and distance of the platform. Only when the platform moves more than N MCS intervals is the update of topology triggered.

[0054] Constraint flooding algorithm based on fuzzy vision: In a large-scale network, in order to avoid the diffusion of messages to the whole network and cause unnecessary overhead, the constraint flooding algorithm based on fuzzy vision is used to limit the spatial distance of topology update packet flooding, that is, the node only sends topology control message when detecting link state change, and uses TTL to limit the flooding range of such TC packet, only sends to nearby nodes, further reduces the overhead.

[0055] The distribution rule is as follows: assuming that the counter is counter=N*2i-1. 1) At time te, counter=1, at this time N=1, i=1, the node detects that the link changes in the previous 21-1te time, and sends the topology change message with TTL=2l.

[0056] 2) At time 2te, counter=2, at this time N=1, i=2, the node detects that the link changes in the previous 22-1te time, and sends the topology change message with TTL=22. 3) At 3te, counter = 3, N = 3, i = 1, the node detects that no link change has occurred in the time period of 2i-1te, and no topology change message is sent. This is followed by analogy to limit the propagation range of topology change message.

[0057] (3) Routing calculation The routing calculation of the adaptive self-organizing network is a typical multi-path multi-rate routing. The main calculation idea is to select the optimal path from the routing table according to certain rules by using the table lookup method.

[0058] The traditional shortest path measurement algorithm (DSR, OLSR) is not suitable for large-scale self-organizing networks, and it is likely to cause data to be transmitted on a one-hop slow multi-hop path, occupy more transmission time, and cause the network throughput to decrease. In view of the topology flooding, each node has the route and link state to other virtual backbone nodes, a transmission time measurement algorithm is proposed to measure the best route, which can make the network obtain greater throughput and lower transmission delay.

[0059] Transmission time measurement algorithm: The best one-hop path is the time spent in transmitting a data packet. If the time spent in transmitting a data packet to a destination node is shorter, the time spent in occupying network resources (medium, queue, calculation, storage, etc.) is shorter. The shorter the resource consumption time, the more the limited network resources can play a role, so the medium occupation time (i.e. transmission time) is used as a measure of the pros and cons of the path. As shown in the scenario, the optimal route selected based on the medium occupation time is not necessarily the one with the least number of hops, but it can reduce the network transmission time and balance the network load. Figure 5

[0060] As shown in Figure 5 The simulation results show that the routing protocol can realize the rapid construction and maintenance of the network multi-path backbone routing, and can realize the balance of the network throughput by using the medium occupation time routing measurement, and the routing overhead occupies no more than 10% of the network bandwidth (50-node network size).

[0061] It should be understood that the size of the serial number of each step in the above embodiment does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0062] It should be understood that those skilled in the art can improve or change according to the above description, and all these improvements and changes should belong to the protection scope of the appended claims of the present application.​

Claims

1. A method for complex and changeable adaptive network generation, characterized in that: The method includes: Build a cross-layer routing architecture consisting of the routing layer and the MAC layer, and design routing protocols at the MAC layer and the routing layer; After obtaining the 1-hop and 2-hop neighbors, the MAC layer selects the respective backbone nodes in a distributed manner, floods them to the entire network, and loads the flooded backbone node declaration packet into the beacon; The routing layer starts to operate on the backbone node, performing secondary relay selection on the backbone node, notifying the entire virtual backbone network of the location of the relay node, and establishing a local topology table; A zero-interval triggered cross-layer cooperative flooding algorithm is used: relay node information is flooded using a combination of slow periodic flooding and triggered flooding. Topology information is updated using both periodic and triggered updates. When the MAC layer detects a change in a node's key parameters, it determines the impact on its one-hop and two-hop neighbors. If the impact exceeds a set threshold, the routing layer is notified to generate a new relay node declaration packet for broadcast. A constrained flooding algorithm based on fuzzy vision is used: the spatial distance of topology update packet flooding is restricted, that is, a node only sends a topology control message when it detects a link status change, and the TTL is used to limit the flooding range of this packet, sending it only to the nearest node; The media occupancy time is used as a measure of the quality of the path to select the optimal routing path.

2. The method for complex and changeable adaptive network generation according to claim 1, characterized in that: The cross-layer routing architecture specifically includes: The MAC layer is used to discover network neighbors, monitor network topology changes, and select backbone nodes. The routing layer is used to update topology changes, maintain routes, and perform fast route selection. The flooding of topology information during topology change updates uses a combination of slow periodic flooding and triggered flooding. The MAC and routing layers are designed to be linked. In addition to the cross-layer routing architecture, a spatial signaling layer is also provided. The spatial signaling layer reports link adaptation parameters to the MAC layer, which then determines the MCS value of the link.

3. The method for complex and changeable adaptive network generation according to claim 1, characterized in that: The method of selecting respective backbone nodes in a distributed manner specifically includes: The backbone node selection algorithm adopts the distributed dominant connection graph algorithm, namely the UCDS algorithm, which generates a UCDS set to form a virtual backbone network of the entire network; The UCDS algorithm is specifically as follows: based on the UCDS rules, the UCDS set is generated, that is, the union of the dominating set DS and the connected set CS; the UCDS rules mark each possible DS and CS, DS is the dominating set of the network topology graph, and CS is the connected set of the network topology graph, and then remove redundant nodes from the CDS. The UCDS rules are used in a certain order, that is, DS rules --> CS candidate rules --> CS exception rules --> CS rules, to correctly select the UCDS set.

4. The method for generating a complex and changeable adaptive network according to claim 3, characterized in that: The UCDS algorithm specifically includes: 1) DS rules: A node i in the topological graph G has the highest dominance factor d among all its neighboring nodes j. i , then node i designates itself as a DS member; it is also possible that node i has the highest dominance factor d among all neighbors k of its neighbor node j. i , then node j designates node i as a DS member; 2) CS Candidate Rules: For a non-DS member node i, there are neighbor nodes j and k, at least one of which is a DS member, and their DS neighbor node sets are disjoint, then node i is a candidate member of CS ; If there are multiple common neighbor nodes of nodes j and k , then it has the highest dominance factor d i of Members will be selected as CS candidate members; 3) CS exception rules: For non-DS member node i, if there is a CS member among the common neighbors of neighbor nodes j and k, the CS rule does not apply. If there is only one non-DS member among the neighbor nodes j and k, and node i and the non-DS member have a common DS neighbor node, the CS rule does not apply either. Usage conditions: If non-DS member node i becomes a CS member because another pair of neighbor nodes l and m satisfy the CS rule, the CS exception rule does not apply; 4) Non-CS rules: For a non-DS member node i, if any of its neighbor nodes j and k are directly connected, then node i is definitely not a CS member; Among them, DS is the dominating set of the network topology graph, and CS is the connected set of the network topology graph; DSN k is the set of DS neighbor node identification numbers of node k, N i is the set of neighbor node identification numbers of node i, d i is the neighbor degree of node i, that is, the number of neighbors of the node, DSN j is the set of DS neighbor node identification numbers of node j, and I, j, k, l, and m refer to the node identification numbers in the network topology respectively.

5. The method for complex and changeable adaptive network generation according to claim 1, characterized in that: The method for performing secondary backbone node secondary relay selection by the routing layer includes: After the MAC layer selects a backbone node and notifies the entire network of the backbone node and its subordinate nodes, the routing layer begins operating on the backbone node. Backbone node routing is implemented using a link state routing protocol algorithm, including: 1) Backbone node neighbor discovery: Since the MAC layer discovers the one-hop and two-hop neighbors of all nodes, the one-hop and two-hop neighbors of all backbone nodes are obtained by default. 2) Backbone node secondary relay selection: Refer to the backbone node selection algorithm and use flooding to declare the backbone node's relay nodes, notifying the entire virtual backbone network of the relay node's location; 3) Establish a local topology table; 4) Find the route.

6. The method for complex and changeable adaptive network generation according to claim 5, characterized in that: The backbone node selection algorithm in the backbone node secondary relay selection refers to the MAC backbone node selection algorithm. Each node in the virtual backbone network independently selects its own relay node set so that all data control packets are forwarded through this relay node set. Through this forwarding, the control packets sent by the node can be delivered to all other two-hop neighbors. The requirements for calculating the relay node set are: there must be a bidirectional symmetric link between the node and the relay node, and the packets sent by the node can reach all strictly two-hop neighbor nodes through forwarding by the relay node.

7. The method for complex and changeable adaptive network generation according to claim 1, characterized in that: The zero-interval triggered cross-layer cooperative flooding algorithm specifically includes: The flooding of relay node information is designed to combine slow periodic flooding with triggered flooding. Topology information updates are divided into two types: periodic updates and triggered updates. Periodic updates are used to ensure that routing information in the routing table does not time out and be deleted. Triggered updates are used to ensure that network topology changes can be quickly flooded to the entire network, improving routing response time. When the MAC layer determines that the key parameters of the node have changed, including neighbor nodes, link status, and queue anomalies, it determines the impact on its one-hop and two-hop neighbors. If the impact exceeds the set threshold, it notifies the routing layer to generate a new relay node declaration packet for broadcast; The temporal frequency of flooding is only related to the frequency of network changes. For self-organizing networks, topology changes depend on the platform movement speed and distance. Only when the platform movement exceeds the constraint of N MCS intervals will the topology update be triggered.

8. The method for complex and changeable adaptive network generation according to claim 1, characterized in that: The fuzzy vision-based constrained flooding algorithm specifically includes: A constrained flooding algorithm based on fuzzy vision is used to limit the spatial distance of topology update packet flooding. That is, a node only sends a topology control message when it detects a link status change. The time-to-live (TTL) is used to limit the flooding range of such TC packets, sending them only to the nearest node. The distribution rules are as follows: Assume that the counter is counter = N*2i-1; 1) At time te, counter = 1, N = 1, i = 1, and the node detects a link change within the previous 21-1te time. At this time, it sends a topology change message with TTL = 2l; 2) At time 2te, counter = 2, N = 1, i = 2, the node detects a link change within the previous 22-1te time, and sends a topology change message with TTL = 22; 3) At time 3te, counter = 3, N = 3, i = 1, and the node detects that no link change occurred within 2i-1te before. At this time, no topology change message is sent. The same process is repeated to limit the propagation range of topology change messages.

9. The method for generating a complex and changeable adaptive network according to claim 1, characterized in that: The method for selecting the optimal routing path specifically includes: The media occupancy time, that is, the transmission time, is used as a measure of the quality of the path. Based on the minimum media occupancy time, the optimal routing path is selected from the routing table.

10. A complex and changeable adaptive network generation system, characterized by: include: a memory for storing executable computer programs; The processor is configured to implement the complex and changeable adaptive network generation method according to any one of claims 1 to 9 when executing the executable computer program stored in the memory.

Images