Distributed time slot resource scheduling method for multi-machine cluster flight ad hoc network

Abstract

The application provides a multi-machine cluster flight self-organizing network distributed time slot resource scheduling method, and belongs to the technical field of wireless self-organizing network communication. The application constructs a dynamic adaptive time slot superframe structure composed of a neighbor sensing subframe, a time slot reservation competition subframe, a data transmission subframe and a time slot maintenance update subframe; a multi-dimensional dynamic priority evaluation model is established by integrating the service emergency level, the link communication quality, the node flight state, the residual power and the topological correlation degree; each flight node completes distributed time slot reservation and time slot competition conflict arbitration by relying on one-hop neighborhood information; different service types adopt a differentiated time slot scheduling mode of single time slot exclusive, multi-time slot aggregation and time slot fragmentation multiplexing, and time slot resource dynamic maintenance is realized for the node network access, offline and link temporary disconnection scene, and a time slot conflict secondary check and neighborhood load balancing mechanism is additionally added. The application adopts a completely distributed scheduling architecture, and eliminates the single point failure risk existing in the centralized scheduling.

Description

Technical Field

[0001] This invention belongs to the field of wireless ad hoc network communication technology, and more specifically, relates to a distributed time slot resource scheduling method for multi-aircraft cluster flight ad hoc networks. Background Technology

[0002] Multi-aircraft swarm self-organizing networks rely on wireless links between aircraft nodes to form a network, eliminating the need for ground base stations and central control units. They are widely used in scenarios such as collaborative airspace inspection, collaborative target reconnaissance, and emergency communication support. Swarm aircraft are characterized by high-speed maneuverability and dynamic formation changes, leading to continuous changes in network topology, frequent temporary disconnections of nodes, and dynamic additions and removals of neighbors. Time-division multiple access (TDMA) time-slot scheduling technology, with its advantages of controllable conflicts and simple scheduling logic, is widely used in UAV swarm self-organizing communication systems.

[0003] Existing time slot scheduling schemes are mainly divided into two categories: centralized time slot scheduling and static distributed TDMA scheduling. Centralized scheduling relies on a central node to collect information from the entire network and complete time slot allocation. Once the central node is damaged or loses connection, the entire network scheduling system will be directly paralyzed, resulting in poor resilience and difficulty in adapting to complex airspace combat and unsupported operating environments. Static distributed TDMA schemes pre-fix time slot allocation rules and time slot lengths, making it impossible to flexibly adjust according to changes in network topology and service traffic. Scheduling priorities are mostly based on simple divisions of service types, ignoring key states such as aircraft battery power, link quality, and topology relay value, which can easily lead to resource allocation imbalances.

[0004] Meanwhile, existing technologies typically employ a uniform time-slot-exclusive transmission mode, which cannot simultaneously address the differentiated transmission needs of mixed services such as short-packet control commands, large-capacity image backhaul, and periodic low-rate monitoring. This results in significant time-slot idleness and low resource utilization. Furthermore, traditional solutions lack corresponding time-slot management strategies for node network entry, offline status, and brief link interruptions. High-speed aircraft maneuvers causing topology changes can easily generate implicit time-slot conflicts, leading to data packet collisions and service transmission delays. Therefore, there is an urgent need to propose a decentralized, dynamically adaptable, and differentiated time-slot resource scheduling method that supports mixed service transmission and addresses the aforementioned shortcomings of existing technologies. Summary of the Invention

[0005] To address the technical problems of existing centralized time slot scheduling being prone to single-point failures, fixed time slot architecture being difficult to adapt to high-speed maneuvering topology changes in clusters, single dimension for scheduling priority evaluation, and low utilization of time slot resources, this invention provides a distributed time slot resource scheduling method for multi-aircraft cluster flight ad hoc networks to solve the problem of dynamic and reasonable allocation of distributed time slot resources in multi-aircraft cluster flight ad hoc networks.

[0006] A distributed time-slot resource scheduling method for multi-aircraft cluster flight self-organizing networks includes the following steps: A dynamic time-slot superframe structure is constructed, consisting of neighbor-aware subframes, time-slot reservation contention subframes, data transmission subframes, and time-slot maintenance and update subframes. A multi-dimensional dynamic priority evaluation model is established based on service urgency level, link communication quality, node flight status, remaining battery power, and topology correlation. Each node performs distributed time-slot reservation, conflict arbitration, and time-slot resource allocation based on one-hop neighborhood information. Different time-slot scheduling modes such as single-time-slot exclusivity, multi-time-slot aggregation, and time-slot fragmentation reuse are enabled for different service types. Dynamic maintenance and smooth switching of time-slot resources are performed for scenarios such as node joining the network, node leaving the network, and temporary link disconnection.

[0007] Preferably, the superframe period and duration of each subframe in the dynamic time-slot superframe structure can be adaptively adjusted based on cluster size, flight speed, and service traffic.

[0008] Preferably, the weight parameters of each evaluation dimension in the multi-dimensional dynamic priority evaluation model can be dynamically adjusted according to the real-time network load.

[0009] Preferably, when a time slot preemption conflict occurs, arbitration is completed based on the dynamic priority assessment results, with high-priority nodes obtaining the right to use the time slot and low-priority nodes re-searching for available time slots.

[0010] Preferably, during the time slot maintenance and update phase, time slot conflict verification, idle time slot reclamation, and network-wide load balancing are completed in real time.

[0011] Compared with the prior art, the present invention has the following beneficial effects: In this invention, a four-partition decoupled superframe architecture is adopted to decouple the topology awareness and time slot scheduling process, adapt to the high dynamic topology change scenario of flight clusters, and reduce the probability of time slot collision.

[0012] In this invention, a multi-dimensional dynamic priority evaluation mechanism is adopted, abandoning the fixed priority strategy, so as to tilt time slot resources toward core business and key nodes, and the scheduling strategy is more intelligent and reasonable.

[0013] In this invention, a purely distributed neighborhood autonomous scheduling method is adopted, which does not rely on a central control node, eliminates the risk of single point of failure, and improves the network's resilience.

[0014] In this invention, multiple time slot reuse modes are set up to adapt to different business needs, and with the idle time slot reclamation mechanism, the utilization rate of time slot resources across the entire network is effectively improved.

[0015] This invention includes a complete dynamic topology maintenance mechanism to enable nodes to quickly join the network and smoothly recover from communication failures, thereby improving network communication stability and fault tolerance.

[0016] In this invention, the parameters of the entire solution can be adaptively adjusted, which can adapt to different cluster sizes and different operating scenarios, and has good versatility and scalability. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall superframe timing structure of the present invention; Figure 2 This is a flowchart of the overall time slot resource scheduling process of the present invention. Detailed Implementation

[0018] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0019] This embodiment provides a distributed time-slot resource scheduling method for multi-aircraft swarm flight self-organizing networks, applied to a multi-aircraft swarm flight self-organizing peer-to-peer communication network composed of 5 to 30 UAVs. The swarm can perform tasks such as airspace collaborative inspection, target collaborative reconnaissance, and emergency air-to-ground communication support. The network architecture is a completely decentralized distributed architecture, with all flight communication nodes within the swarm having completely equal status. There is no ground control base station, swarm head node, or central scheduling unit. The node flight speed is 0-30 m / s, and the effective wireless communication distance of the node is 2500 m. High-speed maneuvers, turns, and formation changes of the aircraft will continuously change the network topology, making it highly susceptible to temporary node disconnection and dynamic addition or removal of one-hop neighbors. The network simultaneously carries three typical services: low-latency emergency command and control services, high-definition image / video high-capacity backhaul services, and periodic node status monitoring low-rate services. Each node is equipped with a local communication scheduling and processing unit, independently completing topology awareness, priority calculation, time slot reservation, conflict arbitration, and time slot resource management.

[0020] The complete implementation process of the scheduling method of this invention is as follows: Step 1: Construct a dynamically adaptive time-slot superframe structure The design employs a periodically cyclical dynamic time-slot superframe, with each complete superframe serving as a scheduling cycle. In this basic embodiment, the initial configuration sets the total duration of a single superframe to 200ms. The complete superframe is divided into four independent and functionally decoupled subframes in chronological order: a neighbor-aware subframe, a time-slot reservation contention subframe, a data transmission subframe, and a time-slot maintenance and update subframe.

[0021] Within the neighbor-aware subframe (base duration 30ms), all online flight nodes in the cluster periodically broadcast standardized heartbeat data packets. The heartbeat data packets carry information including: the aircraft's current 3D position coordinates, real-time flight speed, flight attitude information, remaining onboard battery power, a list of currently running services, the time slot number already occupied, the signal-to-noise ratio of the communication link with neighboring nodes, and the node topology relay identifier. Each node continuously listens for heartbeat messages broadcast by neighboring nodes within its one-hop communication range. After parsing the messages, it updates its locally maintained neighbor topology information table and network-wide time slot resource occupancy table in real time. The topology table records the online status, motion parameters, and communication quality of neighboring nodes; the time slot occupancy table marks which node occupies each time slot, the time slot usage mode, and the service type. If a node's heartbeat message is not received for several consecutive superframe cycles, the node is considered offline.

[0022] In the time slot reservation contention subframe (base duration 40ms), nodes with data transmission needs, based on the data volume and latency metrics of their pending services, generate a standardized time slot reservation request message and broadcast it to their one-hop neighbor. The reservation message carries the number of time slots requested, the service type, the maximum tolerable latency for the service, and the estimated data length. All neighboring nodes receive the reservation request, temporarily store the reservation information, and await unified arbitration.

[0023] In the data transmission subframe (base duration 100ms), nodes complete the sending and receiving of service data packets within their corresponding time slot intervals based on the time slot allocation result obtained in the previous round of competition arbitration; nodes that do not obtain time slot resources enter a silent listening state and wait for the next scheduling cycle.

[0024] The time slot maintenance and update subframe (base duration 30ms) phase does not involve large-scale business data transmission. It mainly performs background scheduling and maintenance work: completes time slot latent conflict detection, offline node identification, idle time slot reclamation, and neighboring network load statistics. At the same time, based on the current network operation indicators, it updates the priority weight and superframe duration parameters used in the next superframe cycle.

[0025] Furthermore, the superframe supports a dynamic adaptive parameter tuning mechanism: when the number of cluster nodes increases, the average flight speed of the aircraft increases, and the network topology changes more rapidly, the system automatically shortens the total duration of the superframe and increases the frequency of topology awareness; when the cluster size decreases, the aircraft cruises at low speed, and the topology is relatively stable, the superframe period is appropriately extended to reduce the wireless communication overhead caused by the continuous broadcasting of nodes and reduce power consumption.

[0026] Step 2: Establish a multi-dimensional dynamic priority evaluation model Abandoning the traditional static strategy of prioritizing solely based on service type, a dynamic priority evaluation model with multi-indicator weighting and fusion is constructed. Five core evaluation dimensions are selected: service urgency level, link communication signal-to-noise ratio, node flight stability, node remaining battery power, and node topology correlation. All indicators are normalized to the range of 0 to 1, and the overall node scheduling priority is calculated through a weighted summation method. In the formula: A is the business urgency level score, with emergency control commands receiving the highest score, followed by image and video backhaul services, and periodic status monitoring services receiving the lowest score; B is the link communication signal-to-noise ratio score, with higher link signal-to-noise ratios indicating stronger transmission reliability and thus higher scores; C is the node flight stability score, with scores appropriately lowered for nodes exhibiting violent attitude maneuvers or high-speed changes of direction; D is the node's remaining battery power score, with lower remaining battery power resulting in higher scores, prioritizing communication permissions for critical nodes with low battery power; E is the node topology correlation score, with higher scores assigned to relay nodes that undertake cross-regional data forwarding and act as bridges for network connectivity. These are the weight coefficients corresponding to each dimension.

[0027] Basic default weight configuration: .

[0028] The weights support real-time adaptive correction: when there is widespread network congestion, the service urgency level weight w1 is automatically increased; when a large number of nodes within the cluster are low on power, the remaining power weight w4 is increased; when the formation undergoes large-scale maneuvers and the link fluctuates frequently, the link signal-to-noise ratio weight w2 is increased. The weight update results are locally refreshed in the time slot maintenance update subframe, and officially take effect in the next round of reservation competition cycle.

[0029] Step 3: Distributed time slot reservation and neighborhood collaborative conflict arbitration During the time slot reservation competition subframe period, multiple nodes are highly likely to simultaneously initiate requests for the same idle time slot, resulting in time slot resource contention conflicts. When time slot resource contention occurs within the neighborhood, all participating nodes uniformly invoke the aforementioned multi-dimensional dynamic priority evaluation model to independently calculate the comprehensive priority value of each competing node; the priority results are compared among them, and the node with the highest comprehensive priority value obtains the right to use the target time slot. Nodes that fail to win the competition voluntarily abandon their time slot application, search their local time slot resource occupancy table, select other unoccupied idle time slots, and re-initiate a time slot application in the next reservation cycle. The entire arbitration process is completed solely through local information exchange between one-hop neighboring nodes, without requiring any central node for centralized computation and instruction issuance, achieving fully distributed autonomous collaborative scheduling.

[0030] Step 4: Configure differentiated time slot resource reuse scheduling modes for different services. To adapt to the scenario of mixed services coexisting in a cluster network, three time slot scheduling modes are designed, and nodes automatically match the corresponding transmission mechanism according to their own service type: (1) Single-timeslot exclusive mode: for emergency command and control services. This type of service has short data packets and strict latency requirements. Nodes apply for a single independent time slot, which is exclusively used by the node during the entire transmission cycle to prevent other nodes from preempting it, thus maximizing the low latency and stable transmission of commands.

[0031] (2) Multi-slot aggregation mode: for high-bandwidth big data backhaul services such as high-definition aerial images and real-time video. Nodes continuously apply for an adjacent idle time slot, and aggregate multiple time slots to form a continuous and uninterrupted transmission channel to meet the continuous transmission needs of large-capacity data streams and avoid the additional overhead caused by frequent segmented data transmission.

[0032] (3) Time slot fragmentation multiplexing mode: for low-rate monitoring services where a large number of nodes periodically report their location, attitude, and power. Multiple non-real-time low-priority nodes share the same time slot and take turns sending short messages within the time slot according to the pre-equal time fragmentation. This significantly reduces the idle time slot resources and increases the number of nodes the network can accommodate.

[0033] Step 5: Dynamic maintenance of time slot resources and smooth communication handover for highly dynamic topologies The continuous maneuvering of swarm aircraft leads to three typical scenarios: node network access, node offline, and temporary link connectivity issues. This invention sets corresponding time slot maintenance strategies: 1) New node joining scenario: Newly started or newly joined flight nodes in the cluster first continuously listen to the heartbeat messages of multiple superframe period neighbor-aware subframes, and synchronously obtain the current network superframe timing rules, time slot allocation mechanism, neighbor node list and time slot occupancy status; after completing network timing synchronization, they initiate time slot application in the next time slot reservation competition subframe, smoothly access the distributed scheduling system, and carry out normal communication.

[0034] 2) Node voluntary network exit / failure offline scenario: When a node detects that a neighboring node has not sent heartbeat messages for several consecutive cycles, it determines that the node has returned to the network and exited the cluster or has gone offline due to hardware failure. The system immediately releases all time slot resources occupied by the offline node, marks them as idle time slots, and makes them available for other nodes in the cluster to re-request and reuse.

[0035] 3) Temporary Link Disconnection Scenario: When an aircraft maneuvers at high speed and temporarily flies out of the communication coverage area, causing a brief interruption of the one-hop link, the system will not immediately reclaim the time slot resources that the node has successfully reserved; the time slot occupancy right is retained. After the aircraft returns to the communication coverage area and the link is restored, the node does not need to compete for time slots again, and can directly use the original time slots to resume data transmission, avoiding data lag and service interruption caused by repeated time slot reservations.

[0036] Step 6: Secondary check for time slot conflicts and load balancing control across the entire network During the time slot maintenance and subframe update phase, background scheduling optimization is performed. Each node compares its local time slot occupancy record with the time slot occupancy information reported by neighboring nodes to identify implicit time slot conflicts arising from distributed contention. If two nodes are detected simultaneously occupying the same time slot, causing a signal collision, a secondary priority arbitration is initiated in the next superframe cycle to reallocate time slot resources and eliminate implicit interference. Simultaneously, each node calculates the overall time slot occupancy rate within its one-hop neighborhood and sets a load threshold of 85%. When the time slot occupancy rate in a local area continuously exceeds the threshold, the area network is deemed overloaded. The scheduling mechanism appropriately restricts the right of nodes in that area to apply for new time slots and guides non-real-time monitoring services to stagger transmission times, balancing the time slot load across the entire cluster and preventing the continued spread of local congestion from affecting critical business communications.

[0037] Preferably, all heartbeat messages and time slot reservation messages incorporate a Cyclic Redundancy Check (CRC) code. The receiving end performs integrity and correctness checks on the messages, filters out erroneous messages caused by spatial electromagnetic interference, and prevents the topology information table and time slot occupancy table from being tampered with by erroneous data, ensuring reliable scheduling decisions. Preferably, the neighbor topology table and time slot occupancy table stored locally by nodes are equipped with an information aging timeout mechanism; if a valid heartbeat message from the corresponding node is not received within a set time limit, expired records are automatically cleared, saving onboard embedded hardware storage resources. Preferably, when a node fails to compete for time slot reservations multiple times consecutively, a random backoff mechanism is introduced, randomly delaying the initiation time of the next reservation message, avoiding a continuous conflict storm caused by a large number of nodes simultaneously initiating applications. Preferably, each node periodically compiles scheduling operation indicators, including the number of time slot conflicts, the overall utilization rate of time slot resources, the average transmission latency of various services, and the service packet loss rate; the statistical results serve as the data basis for adjusting superframe parameters and dynamically correcting priority weights, enabling continuous adaptive optimization of the scheduling strategy. Preferably, after a low battery warning threshold is triggered, the node automatically increases its own scheduling priority while appropriately reducing the frequency of heartbeat message broadcasts to balance communication scheduling needs with aircraft endurance needs.

[0038] Implementation and operational results: In simulation and field testing environments with 5-30 UAV swarms, this invention offers significant advantages over traditional fixed time-division multiple access (TDMO) scheduling and centralized time-slot scheduling schemes: time-slot resource utilization is increased by 25%-32%; the average transmission latency of emergency command and control services is stably controlled within 25ms; the incidence of latent time-slot conflicts is significantly reduced in scenarios involving high-speed formation maneuvers and frequent topology changes; and since there is no central scheduling node in the network, the loss or damage of a small number of aircraft will not paralyze the entire network scheduling mechanism. It can fully meet the engineering requirements for highly dynamic topologies, mixed service carrying, and highly reliable distributed dynamic allocation of time-slot resources in unreliable environments within multi-aircraft swarm self-organizing networks.

[0039] The embodiments of the present invention are given for illustrative and descriptive purposes only, and are not intended to be exhaustive or to limit the invention to the forms disclosed. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of the invention, and to enable those skilled in the art to understand the invention and to design various embodiments with various modifications suitable for a particular purpose.

Claims

1. A distributed time-slot resource scheduling method for multi-aircraft cluster flight self-organizing networks, characterized in that: A self-organizing multi-aircraft cluster flight network with a decentralized architecture is applied. The network consists of several flight nodes that dynamically join and leave the network. Each flight node forms a peer-to-peer network and independently makes time slot resource decisions. The method includes the following steps: S1. Construct a distributed time-slot frame structure adapted to highly dynamic flight scenarios, dividing a single superframe period into four functional subframes: neighbor awareness subframe, time slot reservation contention subframe, data transmission subframe, and time slot maintenance and update subframe. The neighbor awareness subframe is used by each flight node to detect the topology, link quality, and node status information of neighboring nodes in real time. The time slot reservation contention subframe is used to realize distributed time slot preemption and reservation. The data transmission subframe carries service data transmission. The time slot maintenance and update subframe completes time slot conflict detection, resource release, and scheduling strategy iteration. S2. Establish a multi-dimensional dynamic priority evaluation model to calculate the time slot scheduling priority for each flight node in real time. The multi-dimensional dynamic priority evaluation model integrates five core parameters: node flight status, business urgency level, link communication quality, remaining power, and topology correlation. It abandons the fixed priority allocation mechanism and realizes dynamic adaptive updating of priority according to cluster flight status and business needs. S3. Distributed time slot reservation scheduling is completed based on local neighborhood information. Each flight node only needs to interact with the status of one-hop neighboring nodes and time slot occupancy information to autonomously complete idle time slot screening, time slot competition arbitration, dedicated time slot reservation and reuse authorization, without the need for unified scheduling by a central node. In the event of a conflict scenario where multiple nodes simultaneously compete for the same time slot, priority is determined by the multi-dimensional dynamic priority evaluation model. High-priority nodes obtain the right to use the time slot, while low-priority nodes automatically avoid the slot and re-screen for idle time slots. S4. Implement a differentiated time slot resource reuse mechanism. Based on the business data volume, transmission latency requirements and link load status of each flight node, distinguish between three scheduling modes: single time slot exclusive transmission, multi-time slot aggregation transmission, and time slot fragmentation reuse. Dynamically reclaim and re-allocate idle time slots of low-load nodes to improve time slot resource utilization. S5 enables dynamic maintenance and smooth switching of time slot resources under highly dynamic topologies, real-time monitoring of the movement, network entry, network exit, and link breakage reconstruction status of cluster flying nodes, immediate release and reclamation of time slot resources occupied by offline nodes, rapid allocation of idle time slot resources to newly joined nodes, and real-time fine-tuning of time slot scheduling strategies to address link jitter caused by high-speed movement, ensuring the continuity of cluster network communication.

2. The distributed time-slot resource scheduling method for multi-aircraft cluster flight self-organizing networks according to claim 1, characterized in that, In step S1, the neighbor sensing subframe contains several broadcast probe time slots of equal duration. Each flying node periodically broadcasts a heartbeat frame in the corresponding probe time slot. The heartbeat frame carries the node ID, real-time latitude and longitude, flight speed, heading angle, remaining battery power, current service type, link signal-to-noise ratio, and occupied time slot number information. Each flying node constructs a real-time updated one-hop neighbor topology table and time slot resource occupancy table by listening to the heartbeat frames of neighboring nodes. The topology update cycle adapts to the maximum flight speed of the cluster, achieving millisecond-level dynamic topology synchronization.

3. The distributed time-slot resource scheduling method for multi-aircraft cluster flight self-organizing networks according to claim 1, characterized in that, In step S2, the priority calculation formula of the multi-dimensional dynamic priority evaluation model is as follows: P=α·P1+β·P2+γ·P3+δ·P4+ε·P5; Where P represents the overall scheduling priority of the node, the larger the value, the higher the priority; P1 represents the urgency level of the service, with reconnaissance and early warning, and emergency command services having a higher priority than regular data transmission services. P2 is the link quality weight, which is assigned in real time based on the link signal-to-noise ratio and bit error rate. The better the link quality, the higher the weight. P3 is the flight state weight, which is assigned based on the node's flight attitude and position stability to avoid the problem of high-speed and violently maneuvering nodes seizing critical time slots. P4 represents the weight of remaining battery power. The lower the remaining battery power, the higher the weight, ensuring that nodes with low battery power complete core data transmission first. P5 represents the topological correlation weight, with higher weights for core network nodes and multi-neighbor nodes. α, β, γ, δ, and ε are adaptive adjustment coefficients that are dynamically adjusted in real time based on the overall service load and network status of the cluster, and α+β+γ+δ+ε=1.

4. The distributed time-slot resource scheduling method for multi-aircraft cluster flight self-organizing networks according to claim 1, characterized in that, In step S3, the distributed time slot contention arbitration process specifically includes: S31. Each flight node sends a time slot reservation request frame in the time slot reservation competition subframe according to its own service transmission needs. The request frame includes node ID, reservation time slot number, service requirement parameters and real-time priority value. S32. After receiving the reservation request frame, the neighboring node compares the local time slot resource occupancy table to determine whether the target time slot is free. If the time slot is free, the reservation request is directly accepted and the time slot is pre-occupied. If the time slot has been reserved, priority arbitration is triggered. S33. During priority arbitration, the real-time comprehensive priority of the conflicting nodes is compared. The node with the highest priority locks the right to use the target time slot, and the other conflicting nodes terminate the current reservation and re-select idle time slots from the remaining competitive time slots. S34. After a successful reservation, the node marks the time slot occupancy status in the local time slot resource table and sends a reservation confirmation frame to neighboring nodes to complete the synchronization of local time slot status across the entire network.

5. The distributed time-slot resource scheduling method for multi-aircraft cluster flight self-organizing networks according to claim 1, characterized in that, In step S4, the adaptation rules for the three scheduling modes are as follows: Single-time-slot dedicated transmission mode: For command and control services with low latency and small data volume, a fixed dedicated single time slot is allocated to each node for exclusive use throughout the process, avoiding time slot switching and data conflicts, and ensuring the stability of command transmission; Multi-slot aggregation transmission mode: For high-speed services such as high-definition images and big data backhaul, multiple adjacent idle time slots are continuously aggregated to form an aggregated time slot block to meet the needs of high bandwidth transmission. Time slot sharding multiplexing mode: For routine monitoring services with low data rates and multiple nodes, a single time slot is divided into multiple micro-time slot shards and allocated to multiple low-load nodes for time-sharing multiplexing, thereby achieving refined utilization of time slot resources.

6. The distributed time-slot resource scheduling method for multi-aircraft cluster flight self-organizing networks according to claim 1, characterized in that, In step S5, the dynamic maintenance of time slot resources specifically includes node network entry scheduling, node network exit scheduling, and link reconfiguration scheduling: When a new node joins the network, it first listens to the neighbor perception subframe of the entire network to obtain the current network time slot resource distribution status, automatically selects idle time slots to initiate a reservation request, and quickly accesses the network after distributed arbitration to complete the time slot resource allocation. The network access adaptation latency is lower than the preset threshold. When a node actively leaves the network or goes offline abnormally, neighboring nodes determine the node's offline status by detecting heartbeat frame loss, immediately release all time slot resources occupied by the node, and mark it as a reusable idle time slot for other nodes to reschedule. For scenarios involving temporary link breaks and topology reconstruction caused by high-speed cluster maneuvers, the reserved time slot permissions of nodes are retained. Once the link is restored, there is no need to re-reserve, and data transmission can be resumed directly, avoiding communication jitter caused by frequent time slot scheduling.

7. The distributed time-slot resource scheduling method for multi-aircraft cluster flight self-organizing networks according to claim 1, characterized in that, The method also includes a time slot conflict secondary verification and fault tolerance optimization mechanism. During the execution of data transmission subframes, each node monitors time slot transmission conflicts, data packet loss, and bit error anomalies in real time. If a latent time slot conflict is detected, the secondary arbitration mechanism of the time slot maintenance and update subframe is immediately triggered to reallocate time slot resources. At the same time, for transmission anomalies caused by brief link interference, time slot switching is not triggered. Data retransmission is completed through a local buffer retransmission mechanism to improve scheduling fault tolerance.

8. The distributed time-slot resource scheduling method for multi-aircraft cluster flight self-organizing networks according to claim 1, characterized in that, The method also includes a cluster load balancing scheduling mechanism, in which each node statistically analyzes its own time slot resource occupancy rate, service load rate and channel utilization rate in real time, and obtains the local load distribution of the cluster through information exchange with neighboring nodes; for high-load areas, some low-priority services are automatically diverted to surrounding idle time slot resources, and new time slot reservation requests are restricted for overloaded nodes, so as to achieve load balancing of time slot resources across the entire network and avoid local network congestion.

9. The distributed time-slot resource scheduling method for multi-aircraft cluster flight self-organizing networks according to claim 1, characterized in that, The subframe duration and number of time slots in the time slot frame structure support dynamic adaptive adjustment. Each flight node can autonomously iteratively optimize the superframe period and the ratio of each subframe according to the cluster node size, flight maneuver speed and overall service traffic. When the number of nodes increases and the business traffic increases, the number of contention time slots and data transmission time slots will be automatically increased; in high-speed node mobility scenarios, the neighbor perception subframe period will be shortened to improve the topology update frequency and adapt to highly dynamic flight networking scenarios.

10. The distributed time-slot resource scheduling method for multi-aircraft cluster flight ad hoc networks according to any one of claims 1-9, characterized in that, The method adopts a fully distributed peer-to-peer scheduling logic, which does not require the participation of ground base stations or central control nodes. All time slot reservations, conflict arbitrations, resource allocations, and maintenance and update operations are completed autonomously by each flight node based on local neighborhood information. At the same time, the method achieves network-wide scheduling consistency through local information synchronization, effectively avoiding the problem of network-wide scheduling paralysis caused by central node failures, and is suitable for complex airspace and cluster flight networking operation scenarios without reliance.

Images