Networking protocol design method for cooperative air computing in high confrontation environment

Abstract

The invention discloses a networking protocol design method cooperating with air computing in a high confrontation environment. The networking protocol design method mainly solves the problems that in the prior art, self-organizing network data transmission is poor in reliability, high in protocol complexity and large in control overhead. According to the implementation scheme, a networking protocol frame structure comprising a control time slot and a service time slot is designed, and the whole network topology is obtained in the control time slot; calculating the shortest path according to the whole network topology; calculating Y secondary short paths to be searched according to the shortest path, and selecting m paths with the same hop count from the Y paths as use paths; the source node searches for time slots outside a two-hop range for all nodes; according to the found time slots outside the two-hop range, the source node finds all available time slots for the relay node on the path according to the adjacent matrix of the node; and filling all available time slots of all relay nodes on the path into a data packet, and distributing the available time slots to the relay nodes. The method is high in reliability, low in protocol complexity and low in control overhead, and can be used for the MAC layer and the network layer of the self-organizing network of the unmanned system.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communications, and in particular relates to a networking protocol design method, which can be used for a MAC layer and a network layer of an unmanned system self-organizing network. Background Art

[0002] MAC layer access protocols can be divided into three categories according to the different channel access mechanisms: contention protocols, allocation protocols and hybrid protocols. Among them, contention protocols include ALOHA, CSMA / CA, etc., allocation protocols include TDMA, FDMA, etc., and hybrid protocols include HTDMA, etc.

[0003] In competitive protocols, nodes access channels in a competitive manner. Nodes actively seize channels based on their own business needs, while other nodes will delay seizing channels when they detect that the channel is busy. When multiple nodes compete for channels at the same time, conflicts will occur. In order to reduce the probability of conflicts, after failing to compete for channels, nodes are allowed to randomly back off for a period of time according to the backoff strategy and adjust the time for the next access to the channel.

[0004] In allocation-type protocols, different partitioning methods are used to divide resources into several orthogonal sub-resources, and then these sub-resources are allocated to designated nodes for use.

[0005] In a hybrid protocol, a contention protocol is usually combined with an allocation protocol. This type of protocol mainly uses the TDMA protocol and introduces a contention mechanism when allocating time slots.

[0006] The routing protocols used in self-organizing networks are mainly divided into three categories, namely on-demand routing protocols, table-driven routing protocols and hybrid routing protocols. In the on-demand routing protocol, a node will only start route discovery when it wants to send a data packet to its destination. In the table-driven routing protocol, all nodes maintain one or several routing tables. To ensure the real-time nature of routing information, all nodes update the routing table every once in a while. In the hybrid routing protocol, each layer uses a table-driven routing protocol, and the layers use an on-demand routing protocol. Both the on-demand routing protocol and the table-driven routing protocol are planar routing protocols. The hybrid routing protocol includes both an on-demand routing protocol and a table-driven routing protocol.

[0007] At present, in wireless ad hoc networks, the TDMA protocol is the most widely used allocation protocol in scenarios such as drone ad hoc networks. In the TDMA protocol, the time slot is the smallest unit, and several time slots constitute a time slot frame. The number of time slots contained in each time slot frame is called the time slot length. The time it takes for a node to send a data packet to a neighboring node must be less than the length of a time slot. The traditional TDMA protocol avoids the collision of multi-user frames on the physical channel through certain allocation principles. However, air computing technology uses this collision to achieve the superposition of multi-user signals and improve reliability. Therefore, it is necessary to design a reasonable time slot allocation scheme to enable collaborative nodes to send information in the same time slot.

[0008] The patent document with application number CN202310096038.2 discloses a "multi-hop TDMA time slot allocation scheme suitable for unmanned aerial vehicle self-organizing network". It divides each TDMA time frame into a reservation subframe and a data subframe in chronological order. The reservation subframe contains K reservation periods, and the data subframe contains N data time slots; defines control short frames, service long frames, and time slot occupancy tables; occupies a fixed reservation period according to the node ID and priority; each node that has completed network access sends a reservation application frame and a reservation notification frame to the data time slot in the reservation period it occupies according to the current service demand and time slot occupancy; each node that has completed network access maintains a local time slot occupancy table according to the reservation notification frame; after the reservation subframe of a TDMA time frame ends, the reserved data time slots between the nodes send service long frames for communication. This method realizes the on-demand allocation of TDMA channels while giving priority to the channel requirements of high-priority nodes and the timeliness of data. However, the disadvantage of this scheme is that this access method makes the network control overhead large, and its scalability is limited as the number of nodes increases, which is not suitable for large-scale network scenarios. At the same time, the time slot allocation scheme fails to be combined with the relevant routing mechanism, which has a significant impact on the network performance of the UAV self-organizing network and the reliability cannot be guaranteed.

[0009] The patent document with application number CN202211415577.X discloses a "routing and time-frequency allocation method and device based on an air-ground integrated ad hoc network". It first obtains the node information in the air-ground integrated ad hoc network, and the node information includes the available channels and edges in the ad hoc network; then obtains the routing and time-frequency allocation problem model and transmission constraints of the ad hoc network based on the available channels and edges; then obtains the weight of each edge corresponding to the data stream, and obtains the minimum weight routing of the data stream based on the weight and routing constraints; finally, according to the minimum weight routing and time-frequency constraints, the routing and time-frequency allocation problem model of the ad hoc network is converted into a time slot channel allocation optimization problem, and the time slot channel allocation optimization problem is optimized and solved based on the continuous convex approximation strategy. This method can improve the resource availability of the node and reduce the potential contention conflict of the candidate routing path by optimizing the routing and time-frequency allocation in the ad hoc network at the same time, thereby reducing the end-to-end delay of the data stream transmission in the ad hoc network. However, since this method only optimizes the end-to-end delay performance, it will reduce the throughput of the network, and the calculation process is complex, the calculation amount is large, and the hardware requirements are high. Summary of the invention

[0010] The purpose of the present invention is to address the deficiencies of the above-mentioned prior art and propose a networking protocol design method for collaborative air computing in a highly confrontational environment, so as to further improve network reliability, reduce protocol complexity, and reduce control overhead.

[0011] The key technologies and principles for achieving the purpose of the present invention are as follows:

[0012] The present invention realizes the design of the networking protocol by coordinating air computing technology in a self-organizing network.

[0013] The above-mentioned air computing is different from the traditional communication and computing scheme of "transmit first and then calculate". Air computing aims to utilize the waveform superposition property of the signal during the transmission process of the wireless multiple access channel to directly demodulate the superimposed signal and obtain the calculation result. This technology can complete the calculation of an objective function while avoiding the collection of individual data through concurrent transmission of nodes. Because air computing uses the superposition of multiple access channels to realize the calculation of functions, this technology is also called computing based on multiple access channels. Air computing can combine different objective functions through the pre-processing function of the sending end, the post-processing function of the receiving end, and the linear superposition of the waveforms transmitted by multiple nodes during the transmission process to complete different computing tasks. Consider a class of objective functions with a summation structure, that is, it satisfies of the form, where is the pre-processing function of the sending node, and ψ(·) is the post-processing function of the access point. Due to the summation structure of the objective function, the architecture of air computing can be used to realize the calculation of the objective function. It can be seen that air computing can avoid the collection of individual data and directly complete the calculation of the function. No matter how many nodes there are, the objective function can be restored by using the channel once, thereby greatly reducing the calculation delay. At the same time, the concurrent transmission of data by multiple nodes can improve the reliability of the receiving end receiving data. The present invention utilizes the characteristic of concurrent transmission of nodes in air computing to realize the design of a self-organizing network networking protocol for collaborative air computing technology, including an access protocol and a routing protocol.

[0014] Based on the above principles, the technical steps of the present invention for designing the networking protocol include:

[0015] (1) Design a frame structure for a networking protocol suitable for collaborative air computing, which includes a control time slot and a service time slot;

[0016] (2) Construct a wireless self-organizing network scenario, including the distribution of source nodes and relay nodes, and use the designed frame structure to obtain the entire network topology within the control time slot to complete the self-organizing convergence of the network;

[0017] (3) Calculate the shortest path based on the entire network topology;

[0018] (4) Calculate the Y second shortest paths to be found based on the shortest path (Y>0, Y is an integer);

[0019] (5) Select m shortest paths from Y paths as the paths to be used (m <Y);

[0020] (6) Initialize the necessary parameters such as the source node, relay node set, adjacency matrix of other nodes, number of time slots, and node available time slot table;

[0021] (7) The source node searches for time slots outside the two-hop range for all nodes based on the node adjacency matrix;

[0022] (8) Based on the time slots outside the two-hop range of the found nodes, the source node finds a set of available time slots for the relay nodes on the path and marks them, and avoids reusing their marks when searching again;

[0023] (9) Repeat step (8) to complete the search for all available time slots of all relay nodes on the path, fill these available time slots into the data packets, and distribute them to the relay nodes.

[0024] Compared with the prior art, the present invention has the following advantages:

[0025] First, the present invention designs a collaborative air computing technology access protocol, which enables multiple relay nodes to complete function calculations while synchronously sending information in the same time slot, avoiding the process of the opposite end node collecting individual data in turn and then performing function calculations, greatly reducing calculation delays, and at the same time improving the reliability of data received by the receiving end.

[0026] Secondly, the present invention designs a collaborative air computing technology routing protocol, so that all nodes in the network can obtain the entire network topology after two epochs, reducing control overhead. At the same time, it calculates multiple paths based on the Dijkstra algorithm, reducing the complexity of the routing protocol. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is the overall flow chart of the implementation of the present invention;

[0028] Figure 2 It is a schematic diagram of the frame structure designed in the present invention;

[0029] Figure 3 It is a multi-path query sub-flow chart in the present invention;

[0030] Figure 4 It is a sub-flow chart of time slot allocation of a single source node in the present invention. DETAILED DESCRIPTION

[0031] The present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0032] Reference Figure 1 , the implementation steps of this example are as follows:

[0033] Step 1: Design a frame structure of a networking protocol suitable for collaborative air computing.

[0034] like Figure 2 As shown, the frame structure of the networking protocol designed in this step includes a control time slot and a service time slot.

[0035] The control time slot includes two time elements, each of which includes M time frames, and one time frame includes N ordinary time slots, where M is the maximum number of users supported by the network, each user has a unique identification ID in [0, M-1], and N is the maximum number of relay hops supported by the network, which is used to transmit networking control information, that is, when the source information of a user node needs to be sent, the source information is sent in the 0 time slot of the time frame of the user node. After sending, the node enters a dormant state in the subsequent time slots within the current time frame period, and no longer receives, processes, or sends any information data until the next time frame period begins. When a node receives information in the nth time slot that does not belong to its own time frame, it is necessary to determine whether the destination node of the information is the current node: if it is the current node, it is received and processed; otherwise, the node immediately forwards the information in the n+1th time slot within the time frame, and enters a dormant state in the subsequent time slots within the current time frame period, and wakes up only at the beginning of the next time frame period.

[0036] The service time slot includes k service time slots, where the value of k depends on the length of the superframe and is a positive integer multiple of the maximum number of users M, and is used for service transmission.

[0037] Step 2: Build a wireless self-organizing network scenario, use the frame structure to obtain the entire network topology within the control time slot, and complete the self-organizing convergence of the network.

[0038] 2.1) Construct a wireless self-organizing network scenario including the distribution of source nodes and relay nodes;

[0039] 2.2) Use the frame structure designed in step 1 to obtain the topology of the wireless self-organizing network in the control time slot, that is, each node broadcasts the unique identification ID and neighbor status of the source node in the control time slot, which is implemented as follows:

[0040] 2.2.1) Send the unique ID of the node in the first epoch period of the control time slot, that is, initiate control information flooding in the 0th time slot of the time frame of the node, which carries the ID of the node. If other nodes receive the flooding in the 0th time slot of the Xth time frame, they set the Xth node as a neighbor, and the corresponding bit position of the neighbor status is 1;

[0041] 2.2.2) Send the neighbor status in the second time element period within the control time slot, that is, initiate a control information flood containing the node ID and the neighbor status collected in the previous control time frame period in the 0 time slot of the current node time frame. By flooding the neighbor status throughout the network, all nodes in the network can obtain the entire network topology and complete the self-organizing convergence of the network.

[0042] Step 3: Calculate the shortest path based on the entire network topology.

[0043] 3.1) Set node, network, and array parameters, including: initializing the source node o, the destination node d, the set V of all nodes in the network topology, the set S of vertices for which the shortest path has been found, the dis array for recording the current shortest path lengths from the source node o to other vertices, and the path array for recording the predecessor nodes of the shortest paths from the source node o to other vertices;

[0044] 3.2) Use the Dijkstra algorithm for the shortest path to find the shortest path from the source node o to the destination node d:

[0045] 3.2.1) Calculate the end point vi of the shortest path starting from the source node o from the set V - S, and add it to the set S;

[0046] 3.2.2) Modify the shortest path lengths from the source node o to any node vk in the set V - S that is reachable. If the distance from the source node o to vk is greater than the sum of the distance from the source node o to vi and the distance from vi to vk, then record the sum of the distance from the source node o to vi and the distance from vi to vk in the dis array, and modify the predecessor node of node vk in the path array to be vi;

[0047] 3.2.3) Repeat the operations in (3.2.1) and (3.2.2) until all vertices are included in S, and then query the shortest path from the source node o to the destination node d from the path array, and mark it as P y (y = 1), where y is the number of paths found currently.

[0048] Step 4, calculate the Y second - shortest paths to be searched according to the shortest path.

[0049] Refer to Figure 3 , and the specific implementation of this step is as follows:

[0050] 4.1) Initialize parameters such as the source node o, the destination node d, the number Y of shortest paths to be found, the list A of shortest paths already found, the candidate list B, and the network topology and node time slot table;

[0051] 4.2) Compare the currently found number of paths y with the number of paths Y to be found:

[0052] If y < Y and there are candidate paths in the candidate list B, then execute step 4.3);

[0053] If y >= Y or there are no candidate paths in the candidate list B, then complete the path query;

[0054] 4.3) Consider each intermediate node on the P y path except the destination node d as a deviation point. Suppose there are x deviation points in total, and each deviation point is denoted as Vi , i = 1.2...x, and traverse each deviation point starting from i = 1, and find V i The shortest path to the destination node d. When finding the shortest path, pay attention to the V i The shortest path to d cannot include o to V i Any node on the path, and from V i The starting edge cannot be the same as the shortest path from V in List A. i The starting edge is the same, and then P y From the starting point to V i The path + the obtained V i The shortest path to the destination node d is used as the candidate path P y+1 , put it into candidate path list B;

[0055] 4.4) After traversing the deviation points, find the path with the smallest weight in list B, which is the required P y+1 , remove the path from list B and put it into list A, and return to step 4.2).

[0056] Step 5: Select m shortest paths from the y paths as the paths to be used.

[0057] Sort the y shortest paths found according to the number of hops, and select the m shortest paths with the least hops from the y paths as the paths to be used. If the m paths with the least hops cannot be found, select all the paths that meet the minimum hop count.

[0058] Step 6, initialize parameters.

[0059] The initialization parameters include: source node, relay node set, adjacency matrix of other nodes, number of time slots, node available time slot table and other necessary parameters. The adjacency matrix of the initialized node is established by each node according to the connection relationship between nodes in the network topology.

[0060] Step 7: The source node searches for time slots outside the two-hop range for all nodes according to the node adjacency matrix.

[0061] Reference Figure 4 , the specific implementation of this step is as follows:

[0062] 7.1) The source node traverses its own adjacency matrix, and each time it finds a neighbor node, it records the time slot as unavailable, and then traverses the adjacency matrix of this neighbor node to find all neighbors of this neighbor node, and records the time slots of all neighbor nodes of this neighbor node as unavailable, thereby obtaining all time slots outside the two-hop range of the source node;

[0063] 7.2) The source node searches for all time slots outside the two-hop range for the relay node in the same manner as in step 7.1), and traverses all relay nodes in turn to find all time slots outside the two-hop range for the relay node.

[0064] Step 8: Based on the time slots outside the two-hop range of the found nodes, the source node finds a group of available time slots for the relay nodes on the path and marks them to avoid repeated use of their marks when searching again.

[0065] 8.1) The source node traverses its own two-hop out-of-range slot table:

[0066] If an available time slot is found, the node first records the time slot in its own time slot table to be used, and then marks the time slot as unavailable in the time slot tables of other nodes within two hops of the node, and executes step 8.2);

[0067] If an available time slot cannot be found, the search for available time slots of all nodes on the path is completed;

[0068] 8.2) According to the number of relays in the current layer relay node set, find an available time slot for the next layer relay:

[0069] If there is a relay node in this layer, execute step 8.2.1);

[0070] If there are multiple relay nodes in this layer, execute step 8.2.2);

[0071] 8.2.1) Traverse the two-hop out-of-range slot table of this node:

[0072] If an available time slot is found, the time slot is first recorded in the time slot table to be used, and then the time slot is marked as unavailable in the time slot tables of all other nodes within the two-hop range of the node, and step 8.3 is executed;

[0073] If an available time slot cannot be found, the search for available time slots of all nodes on the path is completed;

[0074] 8.2.2) Traverse the two-hop out-of-range time slot table of all nodes in this layer of relays in turn:

[0075] If a sharable time slot can be found, the time slots of all relay nodes in this layer are first recorded in their own time slot table to be used, and then the time slot is marked as unavailable in the time slot table of other nodes within two hops of the node, and step 8.3 is executed;

[0076] If a shared time slot cannot be found, the search for available time slots of all nodes on the path is completed;

[0077] 8.3) Determine whether the current layer is the last layer of relay:

[0078] If it is the last layer of relay, return to step 8.1);

[0079] If it is not the last layer of relay, return to step 8.2).

[0080] Step 9: Fill the available time slots into the data packets and distribute them to the relay nodes.

[0081] The source node fills all available time slots found into the data packet and distributes it to the relay node. That is, the source node first fills all available time slot numbers of each layer into the data packet according to the layer number, and then sends it to the relay node in its first service time slot;

[0082] After receiving this data packet, the relay node forwards it until all relays receive this packet.

[0083] The above description is only a specific example of the present invention and does not constitute any limitation to the present invention. It is obvious that for professionals in this field, after understanding the content and principles of the present invention, they may make various modifications and changes in form and details without departing from the principles and structures of the present invention. However, these modifications and changes based on the ideas of the present invention are still within the scope of protection of the claims of the present invention.

[0084] It should be noted that the step numbers in the specification and claims of the present invention are only for a clear description of the implementation scheme of the present invention to facilitate understanding, and the sequence of the step numbers is not limited.

Claims

1. A networking protocol design method for collaborative air computing in a highly confrontational environment, characterized in that: include: (1) Design a frame structure for a networking protocol suitable for collaborative air computing, which includes a control time slot and a service time slot; (2) Construct a wireless self-organizing network scenario, including the distribution of source nodes and relay nodes, and use the designed frame structure to obtain the entire network topology within the control time slot to complete the self-organizing convergence of the network; (3) Calculate the shortest path based on the entire network topology; (4) Calculate the Y second shortest paths to be found based on the shortest path (Y>0, Y is an integer); (5) Select m shortest paths from Y paths as the paths to be used (m <Y); (6) Initialize the necessary parameters such as the source node, relay node set, adjacency matrix of other nodes, number of time slots, and node available time slot table; (7) The source node searches for time slots outside the two-hop range for all nodes based on the node adjacency matrix; (8) Based on the time slots outside the two-hop range of the found nodes, the source node finds a set of available time slots for the relay nodes on the path and marks them, and avoids reusing their marks when searching again; (9) Repeat step (8) to complete the search for all available time slots of all relay nodes on the path, fill these available time slots into the data packets, and distribute them to the relay nodes.

2. The method according to claim 1, characterized in that The control time slot and service time slot in step (1) have the following structure: The control time slot is used to transmit networking control information, which includes two time units, each time unit includes M time frames, and one time frame includes N ordinary time slots, where M is the maximum number of users supported by the network, each user has a unique identification ID in [0, M-1], and N is the maximum number of relay hops supported by the network; The service time slot is used for service transmission, and includes k service time slots, where the value of k depends on the length of the superframe and is a positive integer multiple of the maximum number of users M.

3. The method according to claim 1, characterized in that In step (2), the designed frame structure is used to obtain the entire network topology in the control time slot by broadcasting source information by each node in the control time slot. The source information includes the unique identification ID of the source node and the neighbor status. The implementation is as follows: The unique ID of the node is sent in the first epoch period of the control time slot, that is, the source control information flood is sent in the 0th time slot of the time frame of the node, which carries the ID of the node and the rest of the bits are all 0. If other nodes receive the flood in the 0th time slot of the Xth time frame, the Xth node is set as a neighbor and the corresponding bit position of the neighbor status is 1; The neighbor status is sent in the second time element period within the control time slot, that is, the control information flooding containing the node ID and the neighbor status collected in the previous control time frame period is initiated in the 0 time slot of the node time frame. By flooding the neighbor status throughout the network, all nodes in the network can obtain the entire network topology.

4. The method according to claim 1, characterized in that: In step (3), the shortest path is calculated based on the entire network topology, which is implemented as follows: (3a) Setting node, network and array parameters, including: initializing the source node o, the destination node d, the set V of all nodes in the network topology, the set S of vertices for recording the shortest paths obtained, the dis array for recording the current shortest path lengths from the source point o to other vertices, and the path array for recording the shortest path predecessor nodes from the source point o to other vertices; (3b) Use the Dijkstra algorithm for the shortest path to find the shortest path from the source node o to the destination node d: (3b1) Select the end point vi of the shortest path starting from the source node o from the set V - S and add it to the set S; (3b2) Modify the length of the shortest path reachable from the source node o to any node vk in the set V - S, record it in the dis array, and modify the predecessor node of the node vk in the path array to be vi; (3b3) Repeat (3b1) and (3b2) until all vertices are included in S. Query the shortest path from the source node o to the destination node d from the path array and mark it as P. y (y=1), y is the number of paths currently found.

5. The method according to claim 1, characterized in that Step (4) Calculate the Y second - shortest paths to be searched according to the shortest path, and the implementation is as follows: (4a) Initialize parameters such as the source node o, the destination node d, the number Y of shortest paths to be found, the list A of the shortest paths already found, the candidate list B, and the network topology and node time - slot table; (4b) Compare the currently obtained number of paths y with the number of paths Y to be found: If y < Y and there are candidate paths in the candidate list B, then execute step (4c); If y >= Y or there are no candidate paths in the candidate list B, then execute step (4e); (4c) Put the y Each point on the path except the destination node d is considered as a deviation point. Suppose there are x deviation points in total, and each deviation point is recorded as V i , i = 1.2...x, and traverse each deviation point starting from i = 1, and find V i The shortest path to the destination node d is P y From the starting point to V i The path + the obtained V i The shortest path to the destination node d is used as the candidate path P y+1 , put it into candidate path list B; (4d) After traversing all the deviation points, find the path with the smallest weight in list B, which is the required P y+1 , remove the path from list B, put it into list A, and return to step (4b); (4e) Sort the y shortest paths found according to the number of hops, and select the m shortest paths with the fewest hops from the y paths as the paths to be used. If the m shortest paths with the fewest hops cannot be found, then select all the paths that meet the condition of the fewest hops.

6. The method according to claim 1, characterized in that: In step (6), the adjacency matrix of the nodes is initialized, and each node establishes its own adjacency matrix according to the connection relationship between the nodes in the network topology.

7. The method according to claim 1, characterized in that: In step (7), the source node finds the time - slots outside the two - hop range for all source nodes and all relay nodes according to the adjacency matrix of the nodes, and the implementation is as follows: The source node traverses its own adjacency matrix. For each neighbor node found, record the current time - slot as unavailable, then traverse the adjacency matrix of this neighbor node, find all the neighbors of this neighbor node, and record the time - slots of all the neighbor nodes of this neighbor node as unavailable, so as to obtain all the time - slots outside the two - hop range of the source node; The source node uses this method to find all the time - slots outside the two - hop range for the relay nodes. Traverse all the relay nodes in turn to find all the time - slots outside the two - hop range for the relay nodes.

8. The method according to claim 1, characterized in that In step (8), according to the time - slots outside the two - hop range of the nodes found, the source node finds a set of available time - slots for the relay nodes on the path, and the implementation is as follows: (8a) The source node traverses its time - slot table outside the two - hop range: If an available time - slot can be found, first record this time - slot in its own time - slot table to be used, and then mark this time - slot as unavailable in the time - slot tables outside the two - hop range of all the nodes within the two - hop range of the source node, and execute step (8b); If an available time - slot cannot be found, then complete the search for the available time - slots of all the nodes on the path; (8b) According to the number of relays contained in the current - layer relay node set, find an available time - slot for the next - layer relay: If there is one relay node in this layer, execute step (8b1); If there are multiple relay nodes in this layer, execute step (8b2); (8b1) Traverse the time - slot table outside the two - hop range of this node: If an available time slot can be found, the node first records the time slot in its own time slot table to be used, and then marks the time slot as unavailable in the time slot tables of all other nodes within the two-hop range of the node, and executes step (8c); If an available time slot cannot be found, the search for available time slots of all nodes on the path is completed; (8b2) Traverse the two-hop out-of-range time slot table of all nodes in this layer of relays in turn: If a sharable time slot can be found, the time slots of all relay nodes in this layer are first recorded in their own time slot table to be used, and then the time slot is marked as unavailable in the time slot tables of other nodes within two hops of the node, and step (8c) is executed; If a shared time slot cannot be found, the search for available time slots of all nodes on the path is completed; (8c) Determine whether the current layer is the last layer of relay: If it is the last layer of relay, return to step (8a); If it is not the last layer of relay, return to step (8b).

9. The method according to claim 1, characterized in that: In step (9), all available time slots found are filled into the data packet and distributed to the relay nodes. The source node first fills the serial numbers of all time slots available in each layer into the data packet according to the layer number, and then sends it to the relay node in its first service time slot. After receiving this data packet, the relay node forwards it until all relays receive this packet.

Images