Networking method with arbitrary time slot recovery and split-fusion capabilities of contention center

Abstract

The present invention discloses a networking method with arbitrary time slot recovery and split-fusion capabilities of a competition center, and belongs to the technical field of swarm drone communication networking. The design of the competition center in the present invention enables each node participating in the networking to have the opportunity to become the network center, avoiding network paralysis caused by the enemy's targeted elimination of the ground center. Even if the network center node is destroyed for various reasons, the remaining nodes will quickly select a new center through competition to form a new network. Arbitrary time slot timing recovery improves the flexibility of the system. No matter which time slot the central node occupies, the slave node can perform timing recovery, which solves the pain point of the central node occupying a fixed time slot in the past. The split-and-re-fusion design solves the problem that the split networks cannot communicate with each other and interfere with each other, and improves the collaborative ability of swarm networking.

Description

Technical Field

[0001] The present invention relates to the technical field of swarm drone communication networking, and in particular to a networking method with the capability of recovering any time slot of a contention center and splitting and merging. Background Art

[0002] In recent years, in the face of sudden natural disasters such as earthquakes, floods, tsunamis, landslides and other rescue work, the demand for emergency networking communications has grown rapidly. The construction of emergency networking communication systems is crucial to reducing the loss of life and property of the people. However, disaster-stricken areas often face the current situation of destroyed or severely lacking communication infrastructure. The networking communication method based on 4G / 5G cannot be restored in a short time and cannot meet the urgent requirements of post-disaster rescue and reconstruction. Emergency networking communications based on satellites and ground stations have narrow bandwidth and large delays, which cannot meet the real-time and high-definition image transmission requirements. In addition, the roads in disaster-stricken areas are often severely damaged, and equipment cannot be quickly transported to the disaster area. For disaster-stricken areas, multi-node and full-area coverage is often required. Therefore, the emergency communication network needs to have the following characteristics: support interconnection between multiple nodes, no equipment (base station) other than communication nodes is required, can be quickly deployed, power on and ready to use, have voice and high-definition video transmission capabilities, strong anti-destruction capabilities, good stability, convenient addition of nodes, split fusion capabilities, and meet the needs of various complex usage scenarios. However, there is no networking method that meets this demand in the prior art. Summary of the invention

[0003] In view of this, the present invention proposes a networking method with the ability to recover any time slot of the competition center and split and merge. In this method, each node participating in the networking has the opportunity to become the network center, avoiding network paralysis caused by damage to the central node, and has the ability to split and merge. In addition, no matter which time slot the central node occupies, the slave node can perform timing recovery.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is:

[0005] A networking method with the capability of recovering any time slot of a contention center and splitting and merging, wherein a time synchronization module is set in each networking node, wherein the networking node includes a drone node and a ground node; and the method comprises the following steps:

[0006] (1) Setting the time frame structure

[0007] There are 63 time slots in a time frame, among which narrow time slots 1 to 50 are downlink time slots, occupied by drone nodes, and transmit downlink narrowband data; narrow time slots 51 and 52 are uplink time slots, occupied by ground nodes, and transmit uplink narrowband data; wide time slots 53 to 60 are downlink time slots, occupied by drone nodes, and transmit downlink broadband data; narrow time slots 61 and 62 are random time slots, randomly occupied by the master node, and are used to avoid mutual collision between time slots in network fusion; time slot 63 is a spectrum sensing time slot, which is used to detect the electromagnetic environment in space; the ground-to-air protection interval is 100Km, and the air-to-air protection interval is 30Km;

[0008] (2) Competing for network central nodes

[0009] A backoff time in units of time frame period is randomly selected, and the channel is continuously monitored during the backoff time; if a cluster head competition frame broadcast by other nodes is received before the backoff time arrives, a narrow time slot is randomly selected in the time frame to transmit a cluster head competition confirmation frame; if a cluster head announcement frame broadcast by other nodes is received before the backoff time arrives, the node enters the network access stage; if a cluster head competition frame broadcast by other nodes is not received after the backoff time ends, the cluster head competition frame is broadcast immediately; after the node broadcasts the cluster head competition frame, it continuously monitors the channel within a time frame length. If a cluster head competition confirmation frame sent by other nodes is received, the cluster head competition frame broadcast is determined to be successful, and the node is elected as the network center node; if a cluster head competition confirmation frame is not received from other nodes within a time frame length, the cluster head competition frame broadcast is determined to have failed, and the node continues to compete for the network center node;

[0010] (3) Access the network from the node

[0011] After receiving the cluster head announcement frame or cluster head competition frame, the slave node performs time slot synchronization; if the network access permission flag in the cluster head announcement frame sent by the central node is 1, the slave node to be connected to the network sends a network access application frame to the central node. After receiving the network access application frame, the central node sends a network access announcement frame to the slave node. After receiving the network access announcement frame, the slave node sends a network access confirmation frame to complete the network access process; if multiple slave nodes send network access application frames to the central node in the same time slot, each node adopts a random backoff method to avoid continuous collisions;

[0012] (4) Time Slot Allocation

[0013] After the slave node successfully joins the network, the central node allocates time slots to the slave node and sends a time slot allocation frame; when the slave node leaves the network, the central node immediately updates the time slot occupancy status of the nodes on the network;

[0014] (5) Arbitrary time slot timing recovery

[0015] The central node timing source is set to 0, that is, local timing recovery, and the slave node timing source is set to the time slot number currently occupied by the central node, thereby achieving recovery time synchronization of any time slot;

[0016] (6) Data transmission

[0017] Each node in the network works according to the received data time slot allocation frame, and sends the message to the corresponding node through the wireless channel in its own transmission time slot;

[0018] (7) De-networking from a node

[0019] When the central node fails to receive any data from a slave node for 3 consecutive seconds, the central node determines that the slave node has been disconnected from the network and immediately sends a data time slot allocation frame to reclaim the occupied time slot resources; when the slave node fails to receive any data from the central node for 5 consecutive seconds, the slave node considers that it has been disconnected from the network and immediately enters the stage of competing for the network central node;

[0020] (8) Network Split and Fusion

[0021] When multiple slave nodes exceed the communication distance and cannot communicate with the central node, causing multiple slave nodes to leave the network, the slave nodes that leave the network form a new communication network by competing for the central node of the network, thereby splitting the original network into multiple networks;

[0022] When the two networks are restored to within the communication distance range, the networks are merged.

[0023] Furthermore, the time system module generates a second pulse with a period of 1s and a time frame period of 200ms.

[0024] Furthermore, in step (2), the backoff time is calculated as follows:

[0025] T c =mod(A Mac ,N)+2+nNT f

[0026] Where, T c is the backoff time, A Mac is the node physical address read from the configuration file, n is the number of failed cluster head contention frames, N is the number of network nodes, T f is the time frame length, and mod() is the remainder function.

[0027] Furthermore, in step (3), the random backoff method is as follows:

[0028] a) The slave node to be connected to the network sends a network access application frame and increases the network access application frame sending counter C by 1;

[0029] b) Randomly generate the timeout period T1 for waiting for the network access announcement frame, where T1 ranges from 6 to 16 and is expressed in time frames;

[0030] c) Wait for the T1 time frame. If a network access notification frame is received before T1 times out, the network access process continues and the timeout timer T1 is canceled;

[0031] d) If the counter C is greater than 3, exit the network access process, clear the counter C, and set a random timeout timer T2, the range of T2 is 1200 to 2400, the unit is time frame;

[0032] e) If T1 times out, return to step a);

[0033] f) If T2 times out, return to step a).

[0034] Furthermore, in step (8), the network fusion method is:

[0035] At the physical level, each node continuously generates a second pulse signal through its own time synchronization module, and calibrates the timing of the central node every 1 second, thereby ensuring the time uniformity of different central nodes and achieving integration between networks;

[0036] If the timing module fails to synchronize timing and cannot be calibrated in real time, fusion can be achieved through the following methods:

[0037] The central node randomly occupies time slots 61 and 62 for broadcasting cluster head announcement frames; the occupancy rule is that the central node uses the locally generated random number t as a basis to select one of the time slots 61 and 62 to send data at the start of each time frame;

[0038] When two central nodes can communicate with each other, they determine each other's MAC address. The central node with a smaller MAC address changes from the sending state to the silent state and enters the slave node network access phase. The slave nodes of the original network to which it belongs also execute the slave node re-entry process.

[0039] When all nodes are connected to the network, a new network is formed and the integration is completed.

[0040] The beneficial effects of the present invention are:

[0041] (1) The design of the competitive center allows every node participating in the network to have the opportunity to become the network center, thus avoiding network paralysis caused by the enemy's targeted elimination of the ground center. Even if the network center node is destroyed for various reasons, the remaining nodes will quickly select a new center through competition and form a new network.

[0042] (2) Arbitrary time slot timing recovery improves the flexibility of the system. No matter which time slot the central node occupies, the slave node can perform timing recovery, solving the pain point of the central node occupying a fixed time slot in the past.

[0043] (3) The split-and-reintegrate design solves the problem that split networks cannot communicate with each other and interfere with each other, thereby improving the collaborative ability of swarm networking. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a time frame structure design diagram in an embodiment of the present invention.

[0045] Figure 2 It is a flow chart of the competition network central node protocol in an embodiment of the present invention.

[0046] Figure 3 It is a behavior flow chart of a network slave node in a slave node network access process in an embodiment of the present invention.

[0047] Figure 4 It is a behavior flow chart of a network center node in a process of slave node accessing the network in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0049] A networking method with the capability of recovering any time slot of a contention center and splitting and merging, comprising the following steps:

[0050] (1) Design time frame structure

[0051] Considering the need for re-integration after splitting, a time synchronization module is added to the system, which generates a second pulse with a period of 1s. To ensure that the period of the second pulse is an integer multiple of the time frame period, the time frame period is designed to be 200ms. 63 time slots are designed in a time frame, among which narrow time slots 1 to 50 are downlink time slots, occupied by airborne nodes, and transmit downlink narrowband data. Narrow time slots 51 and 52 are uplink time slots, occupied by ground nodes, and transmit uplink narrowband data. Wide time slots 53 to 60 are downlink time slots, occupied by airborne nodes, and transmit downlink broadband data such as high-definition video image information. Narrow time slots 61 and 62 are random time slots, randomly occupied by the master node, and are used to avoid collisions between time slots in network integration. Time slot 63 is a spectrum sensing time slot, which is used to detect the electromagnetic environment in space. The ground-to-air protection interval is 100Km, and the air-to-air protection interval is 30Km.

[0052] (2) Competing for network central nodes

[0053] In the competition for the central node of the network, each UAV node and ground node in the network adopts the method of competing to broadcast the cluster head competition frame U1.1 to achieve time synchronization between the nodes in the network and generate the central node of the current network. Before broadcasting the cluster head competition frame U1.1, the UAV node and the ground node first randomly select a period of time before sending the cluster head competition frame U1.1 in units of fixed-length time frames, and continue to monitor the channel during the backoff time. If the cluster head competition frame U1.1 broadcasted by other nodes is received before the backoff time arrives, a narrow time slot is randomly selected in the time frame to transmit the cluster head competition confirmation frame U1.2. If the cluster head announcement frame U2.1 broadcasted by other nodes is received before the backoff time arrives, the node enters the network access stage from the node. If the cluster head competition frame U1.1 broadcasted by other nodes is not received after the backoff time ends, the cluster head competition frame U1.1 is broadcasted immediately.

[0054] For the backoff time, after each node is powered on and reads the configuration parameters, it will set the backoff time for the competing network center node. The backoff time is in fixed-length time frames. In this patent, the backoff time T c Determined by the following formula.

[0055] T c =mod(A Mac ,N)+2+nNT f (1)

[0056] In the above formula, A Mac is the node physical address read from the configuration file; n is the number of failed attempts to send cluster head competition frames U1.1, N is the number of network nodes, T f is the time frame length, and mod() is the remainder function.

[0057] (3) Access the network from the node

[0058] The slave node can complete time slot synchronization after receiving the cluster head announcement frame U2.1 or the cluster head competition frame U1.1. If the network access permission flag in the cluster head announcement frame U2.1 sent by the network center node is 1, the node to be networked sends a network access application frame U2.2 to the control node. After the network center node receives the network access application frame U2.2, it sends a network access announcement frame U2.3 to the slave node. After receiving the network access announcement frame U2.3, the slave node sends a network access confirmation frame U2.4 to complete the network access process. If multiple slave nodes send network access application frames U2.2 to the network center node in the same time slot, data collision may occur. Each node adopts a random backoff method to avoid continuous collisions. The random backoff process is as follows:

[0059] a) The node to be networked sends a network access application frame U2.2 and increases the network access application frame U2.2 sending counter C by 1;

[0060] b) Randomly generate the timeout time T1 for waiting for the network access announcement frame U2.2, with the random range of T1 being 6 to 16, in time frames;

[0061] c) Wait for the T1 time frame. If the network access notification frame U2.2 is received before T1 times out, the network access process continues and the timeout timer T1 is canceled;

[0062] d) If the counter C is greater than 3, exit the network access process, clear the counter C, and set a random timeout timer T2, the random range of T2 is 1200 to 2400, and the unit is time frame;

[0063] e) If T1 times out, return to step a);

[0064] f) If T2 times out, return to step a).

[0065] (4) Time Slot Allocation

[0066] After the network is successfully connected, the central node allocates time slots to each node according to the established plan and sends the time slot allocation frame U3.1. When the node enters or leaves the network, the central node immediately updates the time slot occupancy of the nodes in the network.

[0067] (5) Arbitrary time slot timing recovery

[0068] Fill in 0 for the central node timing source selection for local timing recovery, and fill in the time slot number occupied by the current central node for the slave node timing source selection, so that time synchronization can be restored in any time slot.

[0069] (6) Data transmission

[0070] Each node in the network works according to the received data time slot allocation frame U3.1, and sends the message to the corresponding node through the wireless channel within its own sending time slot.

[0071] (7) De-networking from a node

[0072] According to the application scenario, the main consideration is the exit of slave nodes from the network caused by factors such as drones being shot down and being unable to continue normal communication due to terrain. In this case, when the network center node cannot receive any data from a node for 3 consecutive seconds, the network center node determines that the slave node has exited the network, and will immediately send a data time slot allocation frame U3.1 to recover the occupied time slot resources; when the slave node cannot receive any data from the network center node for 5 consecutive seconds, the network slave node believes that it has been disconnected from the network, and immediately enters the competition network node stage, and the backoff time is determined by formula (1).

[0073] (8) Split and merge

[0074] According to the application scenario, it is considered that multiple drones are affected by their own or external factors, exceeding the communication distance and unable to communicate with the current central node, causing multiple slave nodes to exit the network, and multiple slave nodes form a new communication network through the competition center mode. At this time, the original network is split into two or more networks. In the subsequent flight process, due to other factors such as operation, control, and setting, the two networks are restored to within the communication distance and meet the conditions for fusion. In order to smoothly achieve fusion, dual designs are carried out at the physical level and the algorithm level. Fusion can be achieved if either side plays a role. At the physical level, a time synchronization module is added to each node. The second pulse signal generated by it calibrates the timing of the central node every 1s, ensuring the time uniformity of different central nodes, avoiding conflicts between different time slots, and ensuring mutual communication between nodes. All nodes can be regarded as a communication network. At the algorithm level, even if the timing of the time synchronization module fails and time calibration cannot be performed, the design of the fusion algorithm can also achieve fusion from another dimension. The specific algorithm is implemented as the central node randomly occupying time slots 61 and 62 for broadcasting cluster head announcement frame U2.1. The occupation rule is that the central node uses the locally generated random number t as the basis (the random number seed is the local power-on time) at the beginning of each time frame to select one of the time slots 61 and 62 to send data (if t=1, select time slot 61 to send; if t=2, select time slot 62 to send; if t=0, do not send). When the two central nodes are able to communicate with each other and determine each other's MAC address, the central node with a smaller MAC address changes from the sending state to the silent state and enters the slave node network access link. The slave nodes of the original network to which it belongs also execute the slave node re-entry process. When all nodes are connected to the network, a new network is formed and the fusion is completed. The design at the algorithm level ensures the success rate of network fusion when the time system module cannot provide time all the time, which is an effective supplement to the network fusion at the physical level.

[0075] The design of the competition center in this method allows each node participating in the networking to have the opportunity to become the network center, avoiding network paralysis caused by the enemy's targeted elimination of the ground center. Even if the network center node is destroyed for various reasons, the remaining nodes will quickly select a new center through competition to form a new network. Arbitrary time slot timing recovery improves the flexibility of the system. No matter which time slot the central node occupies, the slave node can perform timing recovery, solving the previous pain point of the central node occupying a fixed time slot. The split-and-reintegrate design solves the problem that the split networks cannot communicate with each other and interfere with each other, and improves the collaborative ability of the swarm networking.

[0076] Here is a more specific example:

[0077] A networking method with the capability of recovering any time slot of a contention center and splitting and merging, the method mainly includes the following contents:

[0078] 1. Time frame structure design

[0079] Considering the need for re-fusion after splitting, a time module is added to the system, which generates a second pulse with a period of 1s. To ensure that the period of the second pulse is an integer multiple of the time frame period, the time frame period is designed to be 200ms. Figure 1 As shown, 63 time slots are designed in a time frame, including 54 narrow time slots with a transmission rate of 36Kbps, 8 wide time slots with a transmission rate of 532.8Kbps, and 1 spectrum sensing time slot. Narrow time slots 1 to 50 are downlink time slots, occupied by airborne nodes, and transmit downlink narrowband data. Narrow time slots 51 and 52 are uplink time slots, occupied by ground nodes, and transmit uplink narrowband data. Wide time slots 53 to 60 are downlink time slots, occupied by airborne nodes, and transmit downlink broadband data such as high-definition video image information. Narrow time slots 61 and 62 are random time slots, randomly occupied by the master node, and are used to avoid mutual collision between time slots in network fusion. Time slot 63 is a spectrum sensing time slot, which is used to detect the electromagnetic environment in space. The ground-to-air protection interval is 100Km, and the air-to-air protection interval is 30Km.

[0080] 2. Competitive Network Central Nodes

[0081] like Figure 2 As shown in the figure, in the stage of competing for the central node of the network, each UAV node and ground node in the network adopts the method of competing to broadcast the cluster head competition frame U1.1 to achieve time synchronization between the nodes in the network and generate the central node of the current network. Before broadcasting the cluster head competition frame U1.1, the UAV node and the ground node first randomly select a period of time before sending the cluster head competition frame U1.1 in units of fixed-length time frames, and continue to monitor the channel during the backoff time. If the cluster head competition frame U1.1 broadcasted by other nodes is received before the backoff time arrives, a narrow time slot is randomly selected in the time frame to transmit the cluster head competition confirmation frame U1.2. If the cluster head announcement frame U2.1 broadcasted by other nodes is received before the backoff time arrives, the node enters the stage of joining the network from the node. If the cluster head competition frame U1.1 broadcasted by other nodes is not received after the backoff time ends, the cluster head competition frame U1.1 is broadcasted immediately.

[0082] For the backoff time, after each node is powered on and reads the configuration parameters, it will set the backoff time for the competing network center node. The backoff time is in fixed-length time frames. In this patent, the backoff time T c Determined by the following formula.

[0083] T c =mod(A Mac ,N)+2+nNT f (2)

[0084] In the above formula, AMac is the node physical address read from the configuration file; n is the number of failed attempts to send cluster head competition frames U1.1, N is the number of network nodes, T f is the time frame length, and mod() is the remainder function.

[0085] 3. Access the network from the node

[0086] The slave node can complete time slot synchronization after receiving the cluster head announcement frame U2.1 or the cluster head competition frame U1.1. If the network access permission flag in the cluster head announcement frame U2.1 sent by the network center node is 1, the node to be networked sends a network access application frame U2.2 to the control node. After receiving the network access application frame U2.2, the network center node sends a network access announcement frame U2.3 to the slave node. After receiving the network access announcement frame, the slave node sends a network access confirmation frame U2.4 to complete the network access process. The behaviors of the network slave node and the network center node in the slave node network access process are as follows: Figure 3 and Figure 4 shown.

[0087] If multiple slave nodes send access application frames U2.2 to the network center node in the same time slot, data collision may occur. Each node adopts random backoff to avoid continuous collision. The random backoff process is as follows:

[0088] a) The node to be networked sends a network access application frame U2.2 and increases the network access application frame U2.2 sending counter C by 1;

[0089] b) Randomly generate the timeout time T1 for waiting for the network access announcement frame U2.2, with the random range of T1 being 6 to 16, in time frames;

[0090] c) Wait for the T1 time frame. If the network access notification frame U2.2 is received before T1 times out, the network access process continues and the timeout timer T1 is canceled;

[0091] d) If the counter C is greater than 3, exit the network access process, clear the counter C, and set a random timeout timer T2, the random range of T2 is 1200 to 2400, and the unit is time frame;

[0092] e) If T1 times out, return to step a);

[0093] f) If T2 times out, return to step a).

[0094] 4. Time Slot Allocation

[0095] After the network is successfully connected, the central node allocates time slots to each node according to the established plan and sends the time slot allocation frame U3.1. When the node enters or leaves the network, the central node immediately updates the time slot occupancy of the nodes in the network.

[0096] 5. Timing recovery of any time slot

[0097] Fill in 0 for the central node timing source selection for local timing recovery, and fill in the time slot number occupied by the current central node for the slave node timing source selection, so that time synchronization can be restored in any time slot.

[0098] 6. Data Transmission

[0099] Each node in the network works according to the received data time slot allocation frame U3.1, and sends the message to the corresponding node through the wireless channel within its own sending time slot.

[0100] 7. De-networking from a node

[0101] According to the application scenario, the main consideration is the exit of slave nodes from the network caused by factors such as drones being shot down and being unable to continue normal communication due to terrain. In this case, when the network center node cannot receive any data from a node for 3 consecutive seconds, the network center node determines that the slave node has exited the network, and will immediately send a data time slot allocation frame U3.1 to recover the occupied time slot resources; when the slave node cannot receive any data from the network center node for 5 consecutive seconds, the network slave node believes that it has been disconnected from the network, and immediately enters the competition network node stage, and the backoff time is determined by formula (2).

[0102] 8. Split and Fusion

[0103] According to the application scenario, it is considered that multiple drones are affected by their own or external factors, exceeding the communication distance and unable to communicate with the current central node, causing multiple slave nodes to exit the network, and multiple slave nodes form a new communication network through the competition center mode. At this time, the original network is split into two or more networks. In the subsequent flight process, due to other factors such as operation, control, and setting, the two networks are restored to within the communication distance and meet the conditions for fusion. In order to smoothly achieve fusion, dual designs are carried out at the physical level and the algorithm level. Fusion can be achieved if either side plays a role. At the physical level, a time synchronization module is added to each node. The second pulse signal generated by it calibrates the timing of the central node every 1s, ensuring the time uniformity of different central nodes, avoiding conflicts between different time slots, and ensuring mutual communication between nodes. All nodes can be regarded as a communication network. At the algorithm level, even if the timing of the time synchronization module fails and time calibration cannot be performed, the design of the fusion algorithm can also achieve fusion from another dimension. The specific algorithm is implemented as the central node randomly occupying time slots 61 and 62 for broadcasting cluster head announcement frame U2.1. The occupation rule is that the central node uses the locally generated random number t as the basis (the random number seed is the local power-on time) at the beginning of each time frame to select one of the time slots 61 and 62 to send data (if t=1, select time slot 61 to send; if t=2, select time slot 62 to send; if t=0, do not send). When the two central nodes are able to communicate with each other and determine each other's MAC address, the central node with a smaller MAC address changes from the sending state to the silent state and enters the slave node network access link. The slave nodes of the original network to which it belongs also execute the slave node re-entry process. When all nodes are connected to the network, a new network is formed and the fusion is completed. The design at the algorithm level ensures the success rate of network fusion when the time system module cannot provide time all the time, which is an effective supplement to the network fusion at the physical level.

[0104] In summary, the present invention first proposes a networking method with the ability to recover any time slot of a competition center and split and merge. The design of the competition center allows each node participating in the networking to have the opportunity to become the network center, avoiding network paralysis caused by damage to the central node. Even if the network center node is destroyed for various reasons, the remaining nodes will quickly select a new center through competition to form a new network. In addition, the timing recovery of any time slot improves the flexibility of the system. No matter which time slot the central node occupies, the slave node can perform timing recovery, solving the pain point of the central node occupying a fixed time slot in the past.

[0105] The present invention also has the ability of splitting and merging. Splitting and merging means that in actual use, multiple drones performing the same target mission are affected by their own or external factors and cannot communicate with the current central node, causing multiple slave nodes to exit the network, and multiple slave nodes form a new communication network through the competitive center mode. At this time, the original network is split into two or more networks. In the subsequent flight process, due to other factors such as operation, control, and setting, the two networks are restored to within the communication distance, and through the design of the physical layer and the algorithm layer, the two split networks are re-fused into one network. This split-and-re-fusion design solves the problem that the split networks cannot communicate with each other and interfere with each other, and improves the collaborative ability of the swarm networking.

Claims

1. A networking method with the capability of recovering any time slot of a contention center and splitting and merging, characterized in that: A time synchronization module is set in each networking node, wherein the networking node includes a drone node and a ground node; and the following steps are included: (1) Setting the time frame structure There are 63 time slots in a time frame, among which narrow time slots 1 to 50 are downlink time slots, occupied by drone nodes, and transmit downlink narrowband data; narrow time slots 51 and 52 are uplink time slots, occupied by ground nodes, and transmit uplink narrowband data; wide time slots 53 to 60 are downlink time slots, occupied by drone nodes, and transmit downlink broadband data; narrow time slots 61 and 62 are random time slots, randomly occupied by the master node, and are used to avoid mutual collision between time slots in network fusion; time slot 63 is a spectrum sensing time slot, which is used to detect the electromagnetic environment in space; the ground-to-air protection interval is 100Km, and the air-to-air protection interval is 30Km; (2) Competition for network center nodes A backoff time in units of time frame period is randomly selected, and the channel is continuously monitored during the backoff time; if a cluster head competition frame broadcast by other nodes is received before the backoff time arrives, a narrow time slot is randomly selected in the time frame to transmit a cluster head competition confirmation frame; if a cluster head announcement frame broadcast by other nodes is received before the backoff time arrives, the node enters the network access stage; if a cluster head competition frame broadcast by other nodes is not received after the backoff time ends, the cluster head competition frame is broadcast immediately; after the node broadcasts the cluster head competition frame, it continuously monitors the channel within a time frame length. If a cluster head competition confirmation frame sent by other nodes is received, the cluster head competition frame broadcast is determined to be successful, and the node is elected as the network center node; if a cluster head competition confirmation frame is not received from other nodes within a time frame length, the cluster head competition frame broadcast is determined to have failed, and the node continues to compete for the network center node; (3) Access the network from the node After receiving the cluster head announcement frame or cluster head competition frame, the slave node performs time slot synchronization; if the network access permission flag in the cluster head announcement frame sent by the central node is 1, the slave node to be connected to the network sends a network access application frame to the central node. After receiving the network access application frame, the central node sends a network access announcement frame to the slave node. After receiving the network access announcement frame, the slave node sends a network access confirmation frame to complete the network access process; if multiple slave nodes send network access application frames to the central node in the same time slot, each node adopts a random backoff method to avoid continuous collisions; (4) Time slot allocation After the slave node successfully joins the network, the central node allocates time slots to the slave node and sends a time slot allocation frame; when the slave node leaves the network, the central node immediately updates the time slot occupancy status of the nodes on the network; (5) Timing recovery in any time slot The central node timing source is set to 0, that is, local timing recovery, and the slave node timing source is set to the time slot number currently occupied by the central node, thereby achieving recovery time synchronization of any time slot; (6) Data transmission Each node in the network works according to the received data time slot allocation frame, and sends the message to the corresponding node through the wireless channel in its own transmission time slot; (7) De-networking from a node When the central node fails to receive any data from a slave node for 3 consecutive seconds, the central node determines that the slave node has been disconnected from the network and immediately sends a data time slot allocation frame to reclaim the occupied time slot resources; when the slave node fails to receive any data from the central node for 5 consecutive seconds, the slave node considers that it has been disconnected from the network and immediately enters the stage of competing for the network central node; (8) Network division and integration When multiple slave nodes exceed the communication distance and cannot communicate with the central node, causing multiple slave nodes to leave the network, the slave nodes that leave the network form a new communication network by competing for the central node of the network, thereby splitting the original network into multiple networks; When the two networks are restored to within the communication distance range, the networks are merged; The network integration methods are as follows: At the physical level, each node continuously generates a second pulse signal through its own time synchronization module, and calibrates the timing of the central node every 1 second, thereby ensuring the time uniformity of different central nodes and achieving integration between networks; If the timing module fails to synchronize timing and cannot be calibrated in real time, fusion can be achieved through the following methods: The central node randomly occupies time slots 61 and 62 for broadcasting cluster head announcement frames; the occupancy rule is that the central node uses the locally generated random number t as a basis to select one of the time slots 61 and 62 to send data at the start of each time frame; When two central nodes can communicate with each other, they determine each other's MAC address. The central node with a smaller MAC address changes from the sending state to the silent state and enters the slave node network access phase. The slave nodes of the original network to which it belongs also execute the slave node re-entry process. When all nodes are connected to the network, a new network is formed and the integration is completed.

2. The networking method with the capability of recovering any time slot of the contention center and splitting and merging according to claim 1 is characterized in that: The time system module generates a second pulse with a period of 1s and a time frame with a period of 200ms.

3. The networking method with the capability of recovering any time slot of the contention center and splitting and merging according to claim 1 is characterized in that: In step (2), the backoff time is calculated as follows: ； In the formula, For the retreat time, is the node physical address read from the configuration file, n is the number of failed attempts to send cluster head contention frames, N is the number of network nodes, is the time frame length, mod() is the remainder function.

4. The networking method with the capability of recovering any time slot of the contention center and splitting and merging according to claim 1 is characterized in that: In step (3), the random backoff method is as follows: a) The slave node to be connected to the network sends a network access application frame and increases the network access application frame sending counter C by 1; b) Randomly generate the timeout period T1 for waiting for the network access announcement frame. The range of T1 is 6 to 16, and the unit is time frame; c) Wait for the T1 time frame. If the network access notification frame is received before T1 times out, the network access process continues and the timeout timer T1 is canceled; d) If the counter C is greater than 3, exit the network access process, clear the counter C, and set a random timeout timer T2, the range of T2 is 1200 to 2400, the unit is time frame; e) If T1 times out, return to step a); f) If T2 times out, return to step a).

Images