Ad-hoc network unmanned aerial vehicle measurement and control system based on dynamic cluster structure and networking communication method thereof

Abstract

The application discloses a kind of networking communication methods of unmanned aerial vehicle measurement and control system of self-organizing network based on dynamic cluster structure, comprising: each node broadcasts own state information by Hello message, and dynamically elects cluster head according to node state information and pre-stores alternative cluster head;Each node calculates link retention time based on the motion vector received by the adjacent node, triggers route early warning in advance and establishes standby path when the link retention time is lower than the early warning threshold, switches to standby path when main path fails;Control plane queue and data plane queue are set up in link layer, and the time slot resource allocation ratio between modulation and coding mode and control plane queue and data plane queue is dynamically adjusted according to channel quality, when channel quality is lower than the preset threshold, time slot resource is preferentially allocated to control plane queue. Method changes route maintenance from post-reconstruction to pre-switching, realizes millisecond-level switching, solves the problem of frequent link breakage under high dynamic topology and long route reconstruction delay.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) communication technology, specifically to a self-organizing UAV telemetry and control system based on a dynamic cluster structure and its networking communication method. Background Technology

[0002] Currently, the mainstream drone telemetry and control solutions include:

[0003] (a) Traditional data transmission radio: It uses a one-to-one data transmission module with a fixed frequency and only supports single-unit operation. The disadvantages of this solution are: it cannot achieve inter-unit networking, the relay needs to rely on additional physical nodes, and it cannot effectively support collaborative operations when the number of drones in the cluster exceeds 3.

[0004] (ii) Wi-Fi Mesh-based modification: Modification using commercial Wi-Fi Mesh chips is attempted to achieve mesh networking among multiple UAVs. However, the Wi-Fi protocol is essentially designed for static or low-speed movement. When the relative speed exceeds 50km / h, the handshake protocol overhead is large and prone to loss of synchronization; moreover, the MAC layer adopts the CSMA / CA mechanism, which results in severe collisions and large latency jitter when there are multiple nodes, making it unsuitable for the deterministic communication requirements of flight control scenarios.

[0005] (III) Solution based on 4G / 5G public network module: This solution utilizes the public cellular network to achieve communication between the UAV and the ground. Its limitations are: it completely fails in signal-free scenarios such as mountainous areas, oceans, and disaster areas, and there are uncertainties in data latency and risks to public network data security.

[0006] In addition to the individual shortcomings of the three types of solutions mentioned above, existing UAV swarm communication technologies as a whole also have the following common deficiencies:

[0007] First, cluster head election and route maintenance are disconnected, making it unable to handle highly dynamic topologies: Existing clustering schemes elect cluster heads based on historical statistics such as relative mobility, and there is no linkage between the election and subsequent route maintenance. The cluster head only assumes topology management responsibilities. When a link is about to break due to the high-speed relative movement of a drone, the network cannot detect it in advance and can only restore communication through a route reconstruction process of >500ms after the link breaks. During the link break-routing-recovery period, the drone is in a dangerous state of being out of contact.

[0008] Second, the separation of dynamic positioning and communication into two stages results in insufficient real-time performance: Existing solutions explicitly divide the system into an offline positioning stage and an online communication stage. After positioning is completed, the UAV's position continues to change during the online stage, leading to a significant timeliness discrepancy between the positioning results and communication decisions. In highly dynamic scenarios, this fragmented architecture causes routing decisions to be based on outdated location information.

[0009] Third, the lack of a service-leveling mechanism means all data is downgraded equally when the channel deteriorates: Existing trunking communication solutions lack a service-leveling concept in their data transmission subsystems—all data is treated the same. When the channel deteriorates, flight control commands and video streams are downgraded equally, leading to a risk of losing "human-in-the-loop" flight control. The root cause is that these solutions are designed for monitoring scenarios and do not involve flight control; therefore, they lack differentiated QoS guarantees for the control plane and data plane. Summary of the Invention

[0010] The purpose of this invention is to provide a self-organizing network UAV telemetry and control system and its networking communication method based on a dynamic cluster structure. By predicting the link breakage time and link status level through link hold time, it can proactively warn and establish backup paths before the link breaks, transforming route maintenance from post-reconstruction to pre-switching, achieving millisecond-level switching, and solving the problems of frequent link breakage and large route reconstruction delay under highly dynamic topologies.

[0011] This invention is achieved through the following technical solution:

[0012] In a first aspect, the first embodiment of the present invention provides a networking communication method for a self-organizing UAV telemetry and control system based on a dynamic cluster structure, comprising:

[0013] After multiple drones are powered on, they enter a listening-competition state. Each node broadcasts its own status information through Hello messages, dynamically elects a cluster head based on the node status information, and pre-stores candidate cluster heads.

[0014] Each node collects motion vectors by periodically broadcasting Hello messages. Based on the received motion vectors of neighboring nodes, each node calculates the link hold time and determines the link status level according to the link hold time. When the link hold time is lower than the warning threshold, a routing warning is triggered in advance and a backup path is established. When the primary path fails, the node switches to the backup path.

[0015] A control plane queue and a data plane queue are established at the link layer to monitor channel quality in real time. The modulation and coding scheme and the time slot resource allocation ratio between the control plane queue and the data plane queue are dynamically adjusted according to the channel quality. When the channel quality is lower than a preset threshold, the time slot resources are forcibly allocated to the control plane queue first.

[0016] Furthermore, the node status information includes normalized remaining power, number of neighboring nodes, and location stability indicators. The node status information dynamically elects a cluster head and pre-stores candidate cluster heads, specifically including:

[0017] Each node calculates its election weight using a comprehensive weighting formula. The node with the highest weight is elected as the primary cluster head, and the node with the second highest weight is stored as a candidate cluster head. The formula for calculating the comprehensive weight W is as follows:

[0018] ;

[0019] Where α, β, and γ are preset weighting coefficients for remaining power, relative position stability, and number of neighboring nodes, respectively; X is the remaining power; S is the relative position stability; and Z is the number of neighboring nodes.

[0020] The relative position stability S is calculated as follows: A three-dimensional position history sequence of length N is maintained at each node. The relative position stability is calculated using the displacement increment and its standard deviation σ between adjacent sampling points. The calculation formula is:

[0021] S = 1 / (1+σ).

[0022] Furthermore, the specific formula for calculating the link hold time is as follows:

[0023] ;

[0024] in, R is the link hold time, and R is the effective communication radius. This is the current distance. , These are the velocity vectors of nodes A and B, respectively.

[0025] Furthermore, the specific method for establishing the backup path includes:

[0026] When the link hold-up time is lower than the warning threshold but higher than the handover threshold, the system enters a warning state. The node that calculates the link hold-up time sends a warning message to the source node. Upon receiving the message, the source node sends a probe message to the alternative next hop. Based on the response, the source node selects the node with the smallest number of hops and the highest RSSI as the backup next hop and sends a backup path confirmation message to confirm the establishment of the backup path. The source node and the backup next hop exchange keep-alive messages during the cycle to ensure the effectiveness of the backup path.

[0027] When the link hold-up time is lower than the handover threshold and the backup path has not been established or has failed, an emergency state is entered. The source node broadcasts a Fast RREQ message. Neighbor nodes that receive the message and can reach the target node reply with a Fast RREP message. The source node sets the first replying node it receives as the backup next hop.

[0028] Furthermore, the specific method for switching to the backup path when the primary path fails includes: when the link hold time drops to 0 or the primary path is actually broken, the FPGA of the source node modifies the routing table entry at the link layer, switches the next hop of the target node to the backup next hop, and after the switch is completed, sends a path release message to the original next hop to release resources, and the backup path becomes the new primary path.

[0029] Furthermore, the control plane queue includes heartbeat packets, remote control commands, and telemetry keyframes, while the data plane queue includes video streams and sensor batch data.

[0030] Furthermore, the dynamic adjustment of the modulation and coding scheme and the time slot resource allocation ratio between the control plane queue and the data plane queue based on channel quality specifically includes:

[0031] When the signal quality is higher than the first threshold, a high modulation method is adopted, and time slot resources are mainly allocated to the data plane queue.

[0032] When the channel quality is below the second threshold, it falls back to the robust modulation scheme and forces the time slot resources to be preferentially allocated to the control plane queue, suspending or reducing the code rate of the data plane queue.

[0033] If the channel quality is between the first and second thresholds, the current modulation scheme remains unchanged.

[0034] Secondly, another embodiment of the present invention provides a self-organizing network UAV telemetry and control system based on a dynamic cluster structure, comprising: a radio frequency front-end module, a baseband processing unit, a flight control interface unit, a positioning and timing module, and a network control and scheduling unit.

[0035] The radio frequency front-end module is used for the transmission and reception of multi-band signals;

[0036] The baseband processing unit includes a heterogeneous computing architecture of FPGA and ARM, where the FPGA is used to process physical layer data and the ARM is used to process protocol stack data.

[0037] The flight control interface unit is used to connect to the UAV's flight control computer and receive remote control commands and telemetry data feedback.

[0038] The positioning and timing module is used to acquire the location information of the UAV and provide clock synchronization for the ad hoc network.

[0039] The network control and scheduling unit performs the method described in the first embodiment above.

[0040] Furthermore, the RF front-end module includes a power amplifier, a low-noise amplifier, and a programmable filter connected in sequence.

[0041] Furthermore, the flight control interface unit includes a UART interface and a CAN bus interface, and the positioning and timing module includes a real-time dynamic differential global positioning module for positioning.

[0042] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0043] This invention provides a self-organizing network UAV telemetry and control system and its networking communication method based on a dynamic cluster structure. Through a three-level linkage closed-loop architecture of perception, prediction, and protection, it achieves a fundamental leap from passive response to proactive prevention. By predicting link breakage time and link status level through link hold time, it proactively issues warnings and establishes backup paths before link breakage, transforming route maintenance from post-reconstruction to pre-switching, achieving millisecond-level switching. This solves the problems of frequent link breakage and long route reconstruction delays under highly dynamic topologies, avoiding the safety hazards of UAVs being out of contact during the "link breakage-pathfinding-recovery" period in traditional solutions. This method is applicable to multi-UAV collaborative tasks in areas with no or weak infrastructure, such as mountainous areas, sea areas, and disaster sites, including but not limited to swarm collaborative search, formation flight, and emergency communication relay scenarios. Attached Figure Description

[0044] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:

[0045] Figure 1 This is a schematic diagram of a self-organizing network UAV telemetry and control system based on a dynamic cluster structure, provided in the first embodiment of the present invention.

[0046] Figure 2 A flowchart illustrating a networking communication method for a self-organizing UAV telemetry and control system based on a dynamic cluster structure, provided in the first embodiment of the present invention;

[0047] Figure 3 This is a flowchart of the cluster head election and handover process in an ad hoc network.

[0048] Figure 4 This diagram illustrates the dynamic adjustment of modulation and coding schemes and the allocation of time slot resources between control plane queues and data plane queues based on channel quality.

[0049] Figure 5 This diagram illustrates proactive route maintenance based on mobility prediction in a bee colony cooperative search scenario. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0051] like Figure 1As shown, the first embodiment of this invention provides a self-organizing network UAV telemetry and control system based on a dynamic cluster structure, integrated inside the UAV body or as an external module. The system includes: a radio frequency (RF) front-end module, a baseband processing unit, a flight control interface unit, a positioning and timing module, and a network control and scheduling unit. The RF front-end module is used for transmitting and receiving multi-band signals and includes a power amplifier (PA), a low-noise amplifier (LNA), and a programmable filter. It supports a wide frequency band coverage of 300MHz - 6GHz and adopts a MIMO (Multiple-Input Multiple-Output) 2x2 architecture to improve link reliability. The baseband processing unit includes a heterogeneous computing architecture of FPGA and ARM. The FPGA is used to process physical layer data, including OFDM (Orthogonal Frequency Division Multiplexing) modulation and demodulation, forward error correction (FEC), and fast Fourier transform (FFT). The ARM is used to process higher-level protocol stack data; the higher-level protocol stack includes self-organizing network routing protocols, TCP / IP acceleration, and flight control data pass-through. The flight control interface unit connects to the UAV's flight control computer, receiving remote control commands and telemetry data feedback. The flight control interface unit includes a UART interface and a CAN bus interface. The positioning and timing module acquires the UAV's position information and provides clock synchronization for the ad hoc network. This module incorporates an RTK-GPS (Real-time Dynamic Differential Global Positioning System) module, which not only provides positioning but also uses the IEEE 1588 protocol to provide high-precision clock synchronization for the ad hoc network, ensuring the accuracy of TDMA (Time Division Multiple Access) time slot allocation. The network control and scheduling unit is used for cluster head election, route maintenance, time slot allocation, service priority control, and abnormal condition handling. This unit employs a network communication method for an ad hoc UAV telemetry and control system based on a dynamic cluster structure, as provided in another embodiment of this invention, to implement these functions.

[0052] This invention provides a self-organizing network UAV telemetry and control system based on a dynamic cluster structure. It enables collaborative operation of the self-organizing network without relying on infrastructure. Positioning information is injected into the routing table in real time via Hello messages, forming a closed loop rather than two separate stages of positioning and communication. RTK-GPS provides centimeter-level positioning, and IEEE 1588 provides microsecond-level clock synchronization; both work together to ensure real-time accuracy of current distance and motion vector. By predicting link breakage time and link status level through link hold time, proactive warnings are issued and backup paths are established before link breakage, transforming route maintenance from post-reconstruction to pre-switching, achieving millisecond-level switching. This solves the problems of frequent link breakage and large route reconstruction delays under highly dynamic topologies.

[0053] like Figure 2 As shown, another embodiment of the present invention provides a networking communication method for a self-organizing UAV telemetry and control system based on a dynamic cluster structure, comprising:

[0054] S1: Dynamic cluster head election and network initialization: After multiple drones are powered on, they enter a listening-competition state. Each node broadcasts its own status information through Hello messages, dynamically elects a cluster head based on the node status information, and pre-stores candidate cluster heads.

[0055] S2: Active route maintenance based on motion prediction: Each node collects motion vectors by periodically broadcasting Hello messages. Based on the received motion vectors of neighboring nodes, each node calculates the link hold time and determines the link status level according to the link hold time. When the link hold time is lower than the warning threshold, a route warning is triggered in advance and a backup path is established. When the primary path fails, the node switches to the backup path.

[0056] S3: Cross-layer resource mapping and dynamic scheduling: A control plane queue and a data plane queue are established at the link layer. The channel quality is monitored in real time. The modulation and coding scheme and the time slot resource allocation ratio between the control plane queue and the data plane queue are dynamically adjusted according to the channel quality. When the channel quality is lower than a preset threshold, the time slot resources are forcibly allocated to the control plane queue first.

[0057] like Figure 3 As shown, in this embodiment, a fixed ground station is not used as the central node. Instead, a temporary cluster head is elected through a comprehensive weighting algorithm. The node with the highest weight is elected as the primary cluster head, and the node with the second highest weight is stored simultaneously as a candidate cluster head.

[0058] The formula for calculating the overall weight W is:

[0059] ;

[0060] Where α, β, and γ are preset weighting coefficients for remaining battery power, relative position stability, and number of neighboring nodes, respectively; X is the remaining battery power; S is the relative position stability; and Z is the number of neighboring nodes. The remaining battery power is directly read from the power management module and normalized to the [0,1] interval. The number of neighboring nodes is obtained by counting the number of different node IDs received in the most recent Hello message period.

[0061] The relative position stability S is calculated as follows: Each node maintains a three-dimensional position history sequence of length N (default N=10), corresponding to a 1-second history window. The relative position stability is calculated using the displacement increment and its standard deviation σ between adjacent sampling points. The calculation formula is:

[0062] S = 1 / (1+σ).

[0063] When a node hovers or flies at a constant speed, S approaches 1; when a node makes a sharp turn or performs a maneuver, S drops significantly. This metric ensures that nodes located at the geometric center of the formation and with stable flight paths receive higher election weights. The drone with the longest endurance, located at the geometric center of the formation, and with the highest connectivity acts as the cluster head, responsible for communicating with the ground station and maintaining intra-cluster routing. If the cluster head's battery level falls below a threshold, a "soft handover" is triggered, smoothly transferring the cluster head role to the next best node, achieving a zero-packet-loss role transition.

[0064] The standard AODV protocol only initiates route error correction (RERR) and route reconstruction when a link breaks, resulting in communication interruption (the black hole period). AODV stands for On-Demand Plane Distance Vector Routing: a classic mobile ad hoc network routing protocol that establishes routes only when communication is needed, making it suitable for resource-constrained nodes.

[0065] In this embodiment, inertial navigation data (IMU) and GPS information are directly injected into the routing table through the physical layer interface. When node A sends a Hello message, it carries its own motion vector, including: velocity vector v, acceleration a, and latitude and longitude coordinates (x, y, z). The Hello message includes node ID, timestamp, location, velocity, acceleration, battery level, queue status, sequence number, and checksum field.

[0066] After receiving the data, neighbor node B calculates the link hold time. The formula is:

[0067] ;

[0068] in, R is the link hold time, and R is the effective communication radius. This is the current distance. , These are the velocity vectors of nodes A and B, respectively.

[0069] The link status level can be determined based on the link holding time. The link status levels include normal status, warning status, emergency status and disconnection status, and different levels correspond to different threshold ranges.

[0070] When the link hold-up time is lower than the warning threshold, a routing warning is triggered in advance and a backup path is established. The method for establishing the backup path is as follows:

[0071] When the link hold-up time is lower than the warning threshold (default 10 seconds) but higher than the handover threshold (default 2 seconds), node A enters a warning state and begins probing and establishing a backup path. At this time, neighbor node B, which calculates the link hold-up time, sends a Route Warning message to source node A. This message includes the failed link identifier, the expected failure time, and a recommended backup next hop field. Upon receiving the Route Warning, node A first queries its local routing table to find other next hops reachable from the target node besides node B. If a backup next hop exists, node A sends a Route Probe message to each. This 2-byte message contains only the target node ID and sequence number. If the backup next hop is reachable from the target node, it replies with a Route Probe ACK message, which includes the hop count and RSSI value. Node A selects the node with the lowest hop count and highest RSSI as the backup next hop and sends a Backup Path Confirm message to confirm the establishment of the backup path. Subsequently, Node A exchanges keep-alive messages with the backup next hop every 200ms to ensure the backup path remains valid during a handover. If three consecutive keep-alive timeouts occur, the backup path is marked as invalid, and Node A must re-execute the discovery process.

[0072] When the link hold-up time falls below the handover threshold, an emergency state is entered. If the backup path has not yet been established or has failed at this time, Node A immediately broadcasts a Fast RREQ message with TTL=2, limiting the number of hops to 2. If a node that receives this message can reach the target node and the number of hops does not exceed 2, it immediately replies with a Fast RREP message. Node A sets the first node to reply as the backup next hop. The entire process is completed within the link hold-up time.

[0073] When the link hold-up time drops to 0 or the primary path is actually broken, Node A's FPGA directly modifies the routing table entry at the link layer, without going through the ARM higher-level protocol stack. The switching time is less than 1ms, switching the target node's next hop to the backup next hop, and simultaneously sending a path release message to the original next hop to release resources. After the switch is complete, the backup path becomes the new primary path, and the node continues to maintain a keep-alive mechanism with the new backup next hop to cope with possible subsequent link changes.

[0074] The routing switch has been transformed from traditional post-event reconstruction to pre-event switching, reducing the switching time from more than 500ms to milliseconds, achieving seamless switching. Even at relative node speeds of 200km / h, end-to-end communication latency remains below 30ms. Through the pre-switching mechanism, link interruption time is reduced from a second-level black hole to near 0ms, avoiding the safety risks of drones being out of contact during the link interruption-routing-recovery period of traditional solutions.

[0075] like Figure 4 As shown, two virtual queues are set up at the link layer:

[0076] Control plane queue: Contains heartbeat packets, remote control commands, and telemetry keyframes. Requirements: Latency < 20ms, packet loss rate < 0.1%.

[0077] Data plane queue: Contains video streams and batch sensor data. Requirements: Best-effort, bandwidth priority.

[0078] The baseband processing unit monitors SNR (Signal-to-Noise Ratio) and RSSI (Received Signal Strength Indicator) in real time. It dynamically adjusts the modulation and coding scheme and the time slot resource allocation ratio between the control plane queue and the data plane queue based on channel quality. Specifically, this includes:

[0079] When the channel quality is good (e.g., SNR > 15dB), 64QAM (quadrature amplitude modulation) high modulation mode is adopted, and time slot resources are mainly allocated to the data plane queue (video stream).

[0080] When the channel quality deteriorates (e.g., SNR < 5dB) or strong interference is detected, it automatically falls back to BPSK (Binary Phase Shift Keying) robust modulation mode and forces the priority allocation of time slot resources to the control plane queue. For example, it forces the time slot ratio of the control plane queue to be increased from about 15% to about 55%, and suspends or reduces the code rate of the data plane queue to ensure that flight control is not lost and to guarantee the reliability of the control plane.

[0081] If the channel quality is between 5dB < SNR < 15dB, the current modulation scheme remains unchanged.

[0082] An upper / lower threshold hysteresis strategy is adopted to avoid frequent jitter switching of modulation mode.

[0083] By implementing dual-queue protection and a priority inversion mechanism in case of channel degradation, reliable transmission of flight control commands is ensured even under adverse channel conditions. In the event of any node failure, the remaining nodes automatically reassemble the network within one second, demonstrating a decentralized self-healing capability.

[0084] This method also provides the following abnormal operating condition handling mechanisms:

[0085] (1) Short-term positioning anomaly: When the positioning information is abnormal, short-term prediction is made based on inertial navigation and historical status, and recalibration is performed after the positioning is restored.

[0086] (2) Strong interference scenario: trigger frequency / bandwidth reconfiguration, increase control plane priority and limit data plane load.

[0087] (3) Critical node failure: triggers rapid re-election and route reconstruction to restore intra-cluster and ground links.

[0088] (4) Queue overload: Enable data plane rate limiting or discarding strategy to prioritize control plane latency constraints.

[0089] The following describes in detail how the methods provided in the embodiments of the present invention are applied to specific application scenarios.

[0090] Example 1: Bee swarm collaborative search scenario

[0091] Imagine a formation of six drones performing a search mission in a mountainous area with no 4G signal. Each drone is equipped with the self-organizing network drone telemetry and control system based on the dynamic cluster structure of this invention.

[0092] Takeoff Phase (Network Initialization and Cluster Head Election):

[0093] After the six drones are powered on, they enter the "listening-competition" state and execute the dynamic cluster head election process in step S1: each node collects and reports local status quantities, including remaining power, position stability indicators, and number of neighbors; the node weight is calculated according to the comprehensive weight calculation formula; the node with the highest weight is selected as the cluster head node; the cluster head publishes the cluster membership relationship and the initial time slot table, and member nodes complete the cluster entry confirmation.

[0094] In this embodiment, aircraft No. 5 was selected as the "air cluster head" due to its highest overall weight, and established a backbone link with the ground command node 10km away, while the other 5 aircraft established an intra-cluster mesh connection.

[0095] Task execution phase (service carrying and cross-layer scheduling):

[0096] like Figure 5As shown, the ground command vehicle issues a "search along the river" command. Downlink: This command is received by Unit 5 (cluster head). Unit 5 searches its routing table and finds the target UAV (Unit 3) in its adjacency list. It forwards the command to Unit 3 via a single hop, with a delay of only 2ms. Uplink: Unit 3 acquires a suspected target video stream (data plane queue), while simultaneously needing to transmit its own GPS / battery data (control plane queue). At this time, the FPGA inside Unit 3's telemetry and control equipment detects slight channel interference (SNR fluctuation). The scheduling in step S3 is executed: the equipment automatically reduces the video bitrate from 10Mbps to 2Mbps and increases the forward error correction bitrate from 1 / 2 to 3 / 4, while ensuring that control plane data is sent at the very beginning of the time slot. The video quality seen at the ground station is slightly degraded, but the remote control and telemetry data remain stable, with no "steering wheel jamming" or "signal loss" phenomena.

[0097] Unexpected Situations (Link Breakage Prediction):

[0098] During the turn, the distance between Unit 5 and Unit 3 increased, with their relative speed reaching 30 m / s. According to step S2, Unit 5 sent a Hello packet carrying its own motion vector. Unit 3 calculated that the link maintenance time was only 0.5 seconds remaining. Unit 3 preemptively broadcast to the network that "Unit 5's link is about to break," and notified Unit 1 (located on the other side of Unit 5) to establish a backup link with Unit 5. When the link between Unit 5 and Unit 3 completely broke, Unit 5 had already seamlessly switched to relay communication through Unit 1, with a communication interruption time of 0 ms, achieving a seamless handover.

[0099] Self-healing:

[0100] If cluster head 5 (unit 5) suddenly crashes and fails due to an accident (such as a bird strike), the system executes a self-healing process linked by steps S1 and S2: Member nodes determine cluster head failure after heartbeat timeout, and the remaining 5 cluster heads detect the heartbeat timeout within 200ms; a re-election is triggered, generating a new cluster head (e.g., unit 2) based on a comprehensive weight: the 5 cluster heads re-execute the election algorithm, and unit 2, being centrally located and having the most neighbors, is selected as the new cluster head. Unit 2 automatically initiates a reconnection request to the ground command vehicle, and the ground station software interface displays "Link reconstruction complete," restoring task data flow. The entire process is unnoticed by the ground operator; the new cluster head synchronizes and publishes updated routing and time slot tables; member nodes complete re-entry into the cluster and link reconstruction; the new cluster head initiates a reconnection to the ground command node, restoring task service continuity. During this process, the system prioritizes the recovery of control plane links before gradually restoring data plane throughput.

[0101] This invention provides a networking communication method for a self-organizing UAV telemetry and control system based on a dynamic cluster structure. Through a three-level linkage closed-loop architecture of perception, prediction, and protection, it achieves a fundamental leap from passive response to proactive prevention. By predicting link breakage time and link status level through link hold time, it proactively issues warnings and establishes backup paths before link breakage, transforming route maintenance from post-reconstruction to pre-switching, achieving millisecond-level switching. This solves the problems of frequent link breakage and long route reconstruction delays under highly dynamic topologies, avoiding the safety hazards of UAVs being out of contact during the "link breakage-pathfinding-recovery" period in traditional solutions. This method is applicable to multi-UAV collaborative tasks in areas with no or weak infrastructure, such as mountainous areas, sea areas, and disaster sites, including but not limited to swarm collaborative search, formation flight, and emergency communication relay scenarios.

[0102] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A networking communication method of a dynamic cluster structure-based unmanned aerial vehicle (UAV) measurement and control system, characterized in that, include: After multiple drones are powered on, they enter a listening-competition state. Each node broadcasts its own status information through Hello messages, dynamically elects a cluster head based on the node status information, and pre-stores candidate cluster heads. Each node collects motion vectors by periodically broadcasting Hello messages. Based on the received motion vectors of neighboring nodes, each node calculates the link hold time and determines the link status level according to the link hold time. When the link hold time is lower than the warning threshold, a routing warning is triggered in advance and a backup path is established. When the primary path fails, the node switches to the backup path. A control plane queue and a data plane queue are established at the link layer to monitor channel quality in real time. The modulation and coding scheme and the time slot resource allocation ratio between the control plane queue and the data plane queue are dynamically adjusted according to the channel quality. When the channel quality is lower than a preset threshold, the time slot resources are forcibly allocated to the control plane queue first.

2. The method according to claim 1, characterized in that, The node status information includes normalized remaining power, number of neighboring nodes, and location stability indicators. The node status information dynamically elects cluster heads and pre-stores candidate cluster heads, specifically including: Each node calculates its election weight using a comprehensive weighting formula. The node with the highest weight is elected as the primary cluster head, and the node with the second highest weight is stored as a candidate cluster head. The formula for calculating the comprehensive weight W is as follows: ； Where α, β, and γ are preset weighting coefficients for remaining power, relative position stability, and number of neighboring nodes, respectively; X is the remaining power; S is the relative position stability; and Z is the number of neighboring nodes. The relative position stability S is calculated as follows: A three-dimensional position history sequence of length N is maintained for each node. The relative stability index is obtained by calculating the displacement increment and its standard deviation σ between adjacent sampling points. The formula for calculating relative position stability is: S = 1 / (1+σ).

3. The method according to claim 1, characterized in that, The specific formula for calculating the link hold time is as follows: ； in, R is the link hold time, and R is the effective communication radius. This is the current distance. , These are the velocity vectors of nodes A and B, respectively.

4. The method according to claim 1, characterized in that, The specific methods for establishing the backup path include: When the link hold-up time is lower than the warning threshold but higher than the handover threshold, the system enters a warning state. The node that calculates the link hold-up time sends a warning message to the source node. Upon receiving the message, the source node sends a probe message to the alternative next hop. Based on the response, the source node selects the node with the smallest number of hops and the highest RSSI as the backup next hop and sends a backup path confirmation message to confirm the establishment of the backup path. The source node and the backup next hop exchange keep-alive messages during the cycle to ensure the effectiveness of the backup path. When the link hold-up time is lower than the handover threshold and the backup path has not been established or has failed, an emergency state is entered. The source node broadcasts a Fast RREQ message. Neighbor nodes that receive the message and can reach the target node reply with a Fast RREP message. The source node sets the first replying node it receives as the backup next hop.

5. The method according to claim 4, characterized in that, The specific method for switching to the backup path when the primary path fails includes: when the link hold time drops to 0 or the primary path is actually broken, the FPGA of the source node modifies the routing table entry at the link layer, switches the next hop of the target node to the backup next hop, and after the switch is completed, sends a path release message to the original next hop to release resources, and the backup path becomes the new primary path.

6. The method according to claim 1, characterized in that, The control plane queue includes heartbeat packets, remote control commands, and telemetry keyframes, while the data plane queue includes video streams and sensor batch data.

7. The method according to claim 1, characterized in that, The specific steps of dynamically adjusting the modulation and coding scheme and the time slot resource allocation ratio between the control plane queue and the data plane queue based on channel quality include: When the signal quality is higher than the first threshold, a high modulation method is adopted, and time slot resources are mainly allocated to the data plane queue. When the channel quality is below the second threshold, it falls back to the robust modulation scheme and forces the time slot resources to be preferentially allocated to the control plane queue, suspending or reducing the code rate of the data plane queue. If the channel quality is between the first and second thresholds, the current modulation scheme remains unchanged.

8. A self-organizing network UAV telemetry and control system based on a dynamic cluster structure, characterized in that, include: Radio frequency front-end module, baseband processing unit, flight control interface unit, positioning and timing module, and network control and scheduling unit. The radio frequency front-end module is used for the transmission and reception of multi-band signals; The baseband processing unit includes a heterogeneous computing architecture of FPGA and ARM, where the FPGA is used to process physical layer data and the ARM is used to process protocol stack data. The flight control interface unit is used to connect to the UAV's flight control computer and receive remote control commands and telemetry data feedback. The positioning and timing module is used to acquire the location information of the UAV and provide clock synchronization for the ad hoc network. The network control and scheduling unit performs the method as described in any one of claims 1-7.

9. The system according to claim 8, characterized in that, The radio frequency front-end module includes a power amplifier, a low-noise amplifier, and a programmable filter connected in sequence.

10. The system according to claim 9, characterized in that, The flight control interface unit includes a UART interface and a CAN bus interface, and the positioning and timing module includes a real-time dynamic differential global positioning module for positioning.

Images