A method for unmanned aerial vehicle communication in complex terrains of remote mountainous areas

Abstract

The application relates to the technical field of mobile data communication, and discloses a UAV communication method for complex terrains in remote mountainous areas, which comprises the following steps: acquiring a real-time coordinate sequence of a UAV and digital elevation model data, and determining a terrain shielding state of a radio frequency propagation path; calculating a first Fresnel zone radius and terrain intrusion depth of the radio frequency propagation path; when the intrusion depth proportion reaches a switching threshold, a resource reservation handshake frame containing a deterministic switching timestamp is sent to a target takeover node by using a current link, a routing metric value is overwritten as infinity to reconstruct a communication topology, the application uses spatial position information to predict a link shielding trend, realizes routing prescheduling based on physical feedforward, completes resource pre-allocation before physical blocking occurs, avoids communication delay caused by ad hoc network routing re-convergence, and improves transmission continuity of inspection control signaling.

Description

Technical Field

[0001] This invention relates to a communication method for unmanned aerial vehicles (UAVs) in remote mountainous areas with complex terrain, and belongs to the field of mobile data communication technology. Background Technology

[0002] Current mobile ad hoc network technology offers flexibility in areas without fixed base station coverage. Through mutual forwarding among multiple mobile nodes, it enables long-distance data backhaul and control signaling interaction. The mainstream routing maintenance mechanism typically maintains the network topology based on link status feedback. This feedback mechanism mainly relies on real-time monitoring of received signal strength or link quality indicators. In inspection scenarios with complex geographical environments, such as remote mountainous areas, physical obstructions caused by sudden terrain changes often lead to instantaneous interruptions of radio frequency links. The existing feedback mechanism is essentially a delayed response mode, triggering route reconvergence only after sensing signal attenuation or packet loss timeout. This design creates communication blind spots when faced with precipitous signal drops, resulting in the loss of critical control commands.

[0003] Simply increasing transmission power or relay node density to address these risks not only increases system deployment costs and power consumption but may also trigger co-channel interference. Furthermore, it fails to fundamentally eliminate the non-line-of-sight propagation limitations caused by physical obstructions, a situation the industry generally accepts due to the uncertainty of latency under terrain interference. Increasing transmission power or relay node density further increases system deployment costs and power consumption, and cannot fundamentally eliminate the non-line-of-sight propagation limitations caused by physical obstructions. Besides hardware improvements, existing technologies also have limitations at the software logic level, such as communication protocols and topology control. Chinese invention patent CN120659122B discloses a method and system for communication and interaction of UAV swarms based on self-organizing networks. It optimizes the allocation of swarm resources through cluster head election and spectrum heat map construction. However, under conditions of strong electromagnetic interference and rugged terrain, the spectrum prediction model and A* algorithm path search are based on the assumption of a relatively stable electromagnetic environment, which makes it difficult to capture the instantaneous intrusion of micro-terrain into the Fresnel zone. Path switching still needs to be triggered after the bit error rate exceeds the limit. When the signal is blocked and instantly returns to zero, the passive remediation logic cannot eliminate the physical delay caused by the reconstruction of the routing protocol.

[0004] Therefore, how to utilize spatial geometric constraints to construct a feedforward route pre-scheduling mechanism to complete topology reconstruction and link resource handover before physical obstruction occurs is the technical problem to be solved by this invention. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: A UAV communication method for complex terrain in remote mountainous areas, the method comprising the following steps: Step S1: Obtain the real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence of the UAV and retrieve the digital elevation model data of the inspection corridor preset in the local memory. The real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence includes longitude, latitude, altitude, motion vector and carrier phase residual. Step S2: Determine the radio frequency beam cone envelope of the UAV relative to the associated relay node based on the real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence, and determine the terrain obstruction status of the radio frequency propagation path in combination with digital elevation model data. Step S3: Calculate the maximum radius of the first Fresnel zone of the radio frequency propagation path in a specific frequency band based on the terrain occlusion state, and extract the terrain intrusion depth corresponding to the terrain feature points along the radio frequency propagation path; Step S4: When the terrain intrusion depth is determined to be the proportion of the maximum radius of the first Fresnel zone to the preset handover threshold, a resource reservation handshake frame containing a deterministic handover timestamp, UAV identification and bandwidth resource requirement parameters is sent to the target takeover node using the current radio frequency link that has not yet been physically blocked. The preset handover threshold is dynamically mapped according to the signal-to-noise ratio of the current wireless environment. Step S5: After the resource reservation handshake frame is sent, the UAV overwrites the routing metric of the current radio frequency link to infinity in the network topology routing table to induce the ad hoc network routing protocol to start the neighbor node rediscovery mechanism and reconstruct the UAV's communication topology path.

[0006] Preferably, step S3 includes: determining the spatial straight-line distance between the UAV and the associated relay node based on the real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence, and calculating the three-dimensional coordinates of the center point of the radio frequency propagation path; retrieving digital elevation model data based on the three-dimensional coordinates of the center point, extracting the vertical offset of the highest obstruction point along the radio frequency propagation path relative to the spatial straight-line distance, and defining the vertical offset as the terrain intrusion depth.

[0007] Preferably, step S4 includes: obtaining the real-time rate of change of terrain intrusion depth relative to flight displacement, and adjusting the preset switching threshold based on the real-time rate of change; when the real-time rate of change shows a monotonically increasing trend, lowering the value of the preset switching threshold so that the timing of sending the resource reservation handshake frame is earlier than the time when the terrain physical occlusion occurs, thereby reserving system response redundancy time for the target takeover node to reallocate communication resources.

[0008] Preferably, a resource reservation handshake frame is sent to the target takeover node so that the target takeover node can parse the bandwidth resource requirement parameters and lock the orthogonal frequency division reusable subcarrier resource corresponding to the specific time slot in the local communication scheduling table at the start time indicated by the deterministic handover timestamp; wherein, the deterministic handover timestamp is determined by the UAV based on the current flight speed and terrain intrusion depth. It is generated in conjunction with the trend of terrain slope change along the radio frequency propagation path.

[0009] Preferably, after step S5, the method further includes: the UAV monitors the convergence status of the routing table in real time, and actively disconnects the wireless radio frequency link with the associated relay node according to the deterministic switching timestamp before the current radio frequency link is physically interrupted; the UAV synchronously accesses the reserved communication time slot of the target takeover node, and uses the locked orthogonal frequency division multiplexing subcarrier resources to restore service data transmission, thereby realizing lossless link handover at the self-organizing network level.

[0010] Preferably, the target takeover node is selected from candidate airport nodes deployed at different geographical coordinates along the UAV inspection path; the method also includes the following steps: calculating the terrain clearance of the radio frequency propagation path between each candidate airport node and the UAV, wherein the terrain clearance is represented by the reciprocal of the proportion of Fresnel interference in each candidate path; selecting the candidate airport node with the highest terrain clearance value as the target takeover node to optimize the link stability after the handover.

[0011] Preferably, the operation of overwriting the routing metric value to infinity modifies the metric parameters in the routing control packet, enabling the shortest path first algorithm within the ad hoc network to perceive that the current radio frequency link is unreachable. Before the signal interruption caused by physical obstruction occurs, this operation forces the network protocol layer to complete the calculation of alternative paths and the remapping of forwarding ports pointing to the target takeover node in advance, thereby eliminating the routing protocol convergence delay caused by terrain changes.

[0012] Preferably, the preset switching threshold is between 0.6 and 0.8. This value range is used to initiate the link handover process before the received signal strength indicator drops sharply when the energy center area of ​​the radio frequency propagation path is intruded by terrain to a extent that the proportion reaches 60% to 80%. This ensures the transmission reliability and channel continuity of critical control signaling when it travels through complex mountainous terrain conditions.

[0013] Preferably, the method further includes the following steps: real-time monitoring of the received signal strength indication of the current radio frequency link; when it is determined that the attenuation slope of the received signal strength indication exceeds a preset sudden change threshold within three consecutive sampling periods, and the terrain intrusion depth ratio has not yet reached the preset switching threshold, the routing metric value overwrite action in step S5 is forcibly executed to respond to sudden channel quality degradation caused by temporary ground moving obstacles not covered by the digital elevation model data.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. In UAV communication methods, by deeply correlating the real-time spatial three-dimensional coordinate sequence of the UAV with the pre-set digital elevation model data of the inspection corridor, the network layer routing decision path has the ability to perceive the physical obstruction of terrain in advance. This breaks the passive maintenance inertia of link degradation feedback triggering route re-convergence that has long existed in the field of mobile ad hoc networks. By calculating the terrain intrusion depth of the first Fresnel zone in the radio frequency propagation path in real time, and generating topology reconstruction feedforward instructions in advance in the healthy state before the physical blockage of the communication link, the routing convergence blind spot caused by terrain abrupt changes is eliminated, and deterministic link switching in complex geographical environments is realized, ensuring the absolute continuity of control signaling during the autonomous flight of the UAV in areas without signal.

[0015] 2. Construct a deterministic routing topology pre-scheduling mechanism based on spatial geometric and physical feedforward. Utilize the acquired spatial location information to pre-determine the interference trend of the terrain on the current radio frequency link. When the terrain intrusion depth is about to reach the interference threshold, the system proactively overwrites the routing metric of the current link to infinity in the routing table and sends a resource reservation handshake frame containing a deterministic handover timestamp to the target takeover node using the currently unblocked link. This allows the target takeover node to complete the pre-allocation of underlying communication time slots before a severe signal fading occurs at the physical layer. This "build first, then disconnect" routing handover logic bypasses the latency bottleneck of finding neighbor nodes after a link is disconnected, a common practice in the industry. It reduces the risk of underlying data packet loss due to physical obstruction in high-voltage inspection in mountainous areas to an objectively negligible level.

[0016] 3. Through the collaboration of the self-organizing network architecture and spatial pre-scheduling logic among multiple airport nodes, each UAV airport node not only undertakes terminal access functions, but also acts as an intelligent relay unit with spatial awareness capabilities. By utilizing the evolution curvature difference of terrain intrusion depth among nodes, the optimal terrain clearance node is dynamically locked from the candidate node group as the takeover target. This resource locking method based on physical space constraints enables robust reconstruction of cross-regional multi-hop transmission paths, enhances the underlying communication determinism of mobile ad hoc networks under the superposition of strong electromagnetic interference and extreme terrain physical obstruction, and provides closed-loop communication support with engineering realism for unmanned fully autonomous inspection in remote mountainous areas without network. Attached Figure Description

[0017] Figure 1 This is a flowchart of the physical feedforward routing pre-scheduling process of the present invention; Figure 2 This is a diagram of the deterministic link switching logic architecture of the present invention.

[0018] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0020] A drone communication method for remote mountainous areas with complex terrain, the method includes the following steps: Step S1: Obtain the real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence of the UAV and retrieve the digital elevation model data of the inspection corridor preset in the local memory. The real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence includes longitude, latitude, altitude, motion vector and carrier phase residual. Step S2: Determine the radio frequency beam cone envelope of the UAV relative to the associated relay node based on the real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence, and determine the terrain obstruction status of the radio frequency propagation path in combination with digital elevation model data. Step S3: Calculate the maximum radius of the first Fresnel zone of the radio frequency propagation path in a specific frequency band based on the terrain occlusion state, and extract the terrain intrusion depth corresponding to the terrain feature points along the radio frequency propagation path; Step S4: When the terrain intrusion depth is determined to be the proportion of the maximum radius of the first Fresnel zone to the preset handover threshold, a resource reservation handshake frame containing a deterministic handover timestamp, UAV identification and bandwidth resource requirement parameters is sent to the target takeover node using the current radio frequency link that has not yet been physically blocked. The preset handover threshold is dynamically mapped according to the signal-to-noise ratio of the current wireless environment. Step S5: After the resource reservation handshake frame is sent, the UAV overwrites the routing metric of the current radio frequency link to infinity in the network topology routing table to induce the ad hoc network routing protocol to start the neighbor node rediscovery mechanism and reconstruct the UAV's communication topology path.

[0021] Preferably, step S3 includes: determining the spatial straight-line distance between the UAV and the associated relay node based on the real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence, and calculating the three-dimensional coordinates of the center point of the radio frequency propagation path; retrieving digital elevation model data based on the three-dimensional coordinates of the center point, extracting the vertical offset of the highest obstruction point along the radio frequency propagation path relative to the spatial straight-line distance, and defining the vertical offset as the terrain intrusion depth.

[0022] Preferably, step S4 includes: obtaining the real-time rate of change of terrain intrusion depth relative to flight displacement, and adjusting the preset switching threshold based on the real-time rate of change; when the real-time rate of change shows a monotonically increasing trend, lowering the value of the preset switching threshold so that the timing of sending the resource reservation handshake frame is earlier than the time when the terrain physical occlusion occurs, thereby reserving system response redundancy time for the target takeover node to reallocate communication resources.

[0023] Preferably, a resource reservation handshake frame is sent to the target takeover node so that the target takeover node can parse the bandwidth resource requirement parameters and lock the orthogonal frequency division reusable subcarrier resource corresponding to the specific time slot in the local communication scheduling table at the start time indicated by the deterministic handover timestamp; wherein, the deterministic handover timestamp is determined by the UAV based on the current flight speed and terrain intrusion depth. It is generated in conjunction with the trend of terrain slope change along the radio frequency propagation path.

[0024] Preferably, after step S5, the method further includes: the UAV monitors the convergence status of the routing table in real time, and actively disconnects the wireless radio frequency link with the associated relay node according to the deterministic switching timestamp before the current radio frequency link is physically interrupted; the UAV synchronously accesses the reserved communication time slot of the target takeover node, and uses the locked orthogonal frequency division multiplexing subcarrier resources to restore service data transmission, thereby realizing lossless link handover at the self-organizing network level.

[0025] Preferably, the target takeover node is selected from candidate airport nodes deployed at different geographical coordinates along the UAV inspection path; the method also includes the following steps: calculating the terrain clearance of the radio frequency propagation path between each candidate airport node and the UAV, wherein the terrain clearance is represented by the reciprocal of the proportion of Fresnel interference in each candidate path; selecting the candidate airport node with the highest terrain clearance value as the target takeover node to optimize the link stability after the handover.

[0026] Preferably, the operation of overwriting the routing metric value to infinity modifies the metric parameters in the routing control packet, enabling the shortest path first algorithm within the ad hoc network to perceive that the current radio frequency link is unreachable. Before the signal interruption caused by physical obstruction occurs, this operation forces the network protocol layer to complete the calculation of alternative paths and the remapping of forwarding ports pointing to the target takeover node in advance, thereby eliminating the routing protocol convergence delay caused by terrain changes.

[0027] Preferably, the preset switching threshold is between 0.6 and 0.8. This value range is used to initiate the link handover process before the received signal strength indicator drops sharply when the energy center area of ​​the radio frequency propagation path is intruded by terrain to a extent that the proportion reaches 60% to 80%. This ensures the transmission reliability and channel continuity of critical control signaling when it travels through complex mountainous terrain conditions.

[0028] Preferably, the method further includes the following steps: real-time monitoring of the received signal strength indication of the current radio frequency link; when it is determined that the attenuation slope of the received signal strength indication exceeds a preset sudden change threshold within three consecutive sampling periods, and the terrain intrusion depth ratio has not yet reached the preset switching threshold, the routing metric value overwrite action in step S5 is forcibly executed to respond to sudden channel quality degradation caused by temporary ground moving obstacles not covered by the digital elevation model data.

[0029] Example 1: In the inspection of power transmission lines located in remote valleys and subject to high-voltage electromagnetic interference, a UAV communication method for complex terrain in remote mountainous areas is used as the topology control logic of a mobile ad hoc network. It operates within a communication architecture consisting of a UAV and multiple alternative airport nodes. The UAV flies at a fixed cruising speed along the rugged terrain marked by a given inspection corridor digital elevation model. Due to the alternating physical obstructions of ridges and transmission towers, the radio frequency propagation path experiences transient blockages. The passive routing convergence mechanism, triggered by drops in received signal strength, causes control signaling transmission gaps of 100ms to 300ms, compromising the safety of the UAV's beyond-line-of-sight flight. The UAV's control module periodically acquires a real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence, including longitude, latitude, altitude, motion vector, and carrier phase residual. Based on this sequence, the control module calculates the spatial straight-line distance between the UAV and the currently associated relay node. It then retrieves the locally stored inspection corridor digital elevation model data, extracts the vertical offset of the highest obstruction point along the radio frequency propagation path relative to the spatial straight-line distance, and defines it as the terrain intrusion depth. The control module calculates the maximum radius of the first Fresnel zone of the radio frequency propagation path based on the electromagnetic wave diffusion law of a specific transmission frequency band. The control module calculates the terrain intrusion depth. When the real-time rate of change of the drone's flight displacement shows a monotonically increasing trend relative to the actual flight displacement, the control module lowers the preset switching threshold. The value is used to reserve system response redundancy time for the takeover node to reallocate communication resources, when determining the terrain intrusion depth. Occupying the largest radius of the first Fresnel zone The proportion reaches the preset switching threshold When the underlying radio frequency channel has not yet experienced substantial attenuation, the control module sends a resource reservation handshake frame to the target takeover node using the current radio frequency link. Based on terrain clearance, the target takeover node is pre-selected. The resource reservation handshake frame includes a deterministic handover timestamp, UAV identification, and bandwidth resource requirement parameters. The target takeover node receives and parses the requirement parameters and performs a locking operation on the orthogonal frequency division reused subcarrier resources corresponding to the specific time slot. In practice, after parsing the requirement parameters, the target takeover node extracts the currently available orthogonal frequency division reused subcarrier mask sequence at its Media Access Control (MAC) layer, avoiding communication frequencies already occupied by neighboring nodes. It broadcasts a resource reservation declaration frame with a specific timestamp through distributed coordination functions or reserved control channels. In its own resource scheduling table, it forcibly marks the selected subcarrier block parameters and corresponding time periods as non-preemptible. Thus, in a self-organizing network architecture without a central base station, the pre-allocation and anti-collision locking of dedicated time slots are completed through underlying signaling handshake. The start time of the specific time slot is determined by the target takeover node based on its current flight speed and terrain intrusion depth. Based on the collaborative calculation of terrain slope change trends, the above process completes the temporal pre-allocation of underlying communication resources before physical blockage, using three-dimensional spatial location feedforward.

[0030] After sending the resource reservation handshake frame, the UAV overwrites the routing metric of the current radio frequency link to infinity in its network topology routing table. This modification of the metric parameter triggers the shortest path first algorithm inside the ad hoc network to detect that the current link is unreachable. Based on this, the network protocol layer calculates alternative paths in advance and directs the data traffic to the forwarding port of the target takeover node. To achieve this cross-layer guidance action, the network protocol stack inside the UAV has a built-in cross-layer communication callback interface. When the network layer routing protocol detects that the next-hop metric of the original associated relay node has been overwritten and updated, it immediately triggers an asynchronous interrupt signal with hardware priority to the lower layer.Upon receiving the interruption, the data link layer forcibly refreshes its local MAC address forwarding table, directly replacing the underlying physical transmission target address that originally pointed to the old associated relay node with the MAC address of the target takeover node. This allows pending service data in the buffer to skip the regular address resolution waiting period and directly overflow to the new port through the underlying hardware interface, completing physical frame encapsulation. The UAV monitors the convergence status of the routing table in real time. When the deterministic handover timestamp arrives, the UAV actively disconnects the radio frequency link with the associated relay node and simultaneously accesses the reserved communication time slot of the target takeover node. The UAV uses locked orthogonal frequency division multiplexing subcarrier resources to restore service data transmission. The aforementioned metric covers... The writing mechanism transforms the evolutionary constraints of the spatial geometric and physical environment into direct routing control parameters for the telecommunications network layer, avoiding network protocol convergence delays and maintaining data transmission continuity under high-frequency obstruction conditions. In the dynamically evolving terrain obstruction perception scenario, the construction of the radio frequency beam cone envelope is used to define the retrieval spatial boundary of the terrain obstruction state. The control module retrieves the UAV's current longitude, latitude, altitude, and carrier phase residual, and determines the main lobe radiation direction of the radio frequency antenna by combining it with the current motion vector. When performing this direction determination action, the UAV simultaneously calls the onboard inertial measurement unit to obtain real-time body attitude data including yaw angle, pitch angle, and roll angle, and calculates the three-dimensional motion vector and the actual nose position. The system detects and compensates for deviations in sideslip and attitude angles between physical orientations using a servo gimbal mechanism. This ensures that the mechanical normal of the communication antenna remains in a controlled, responsive state, parallel to its actual spatial flight vector. This guarantees that the main lobe direction derived from the trajectory is accurately equivalent to the actual radiation path of the radio frequency beam in physical space. The control module establishes a conical geometric model with a preset angle along the main lobe radiation direction, using the UAV's phase center as the vertex. This conical geometric model is defined as the radio frequency beam cone envelope. The computation module only retrieves digital elevation model data for the inspection corridor within the geographical area covered by the radio frequency beam cone envelope. This limited retrieval mechanism enables the determination of terrain occlusion status. The system reduces the dimensionality of the global elevation matrix search to a cone projection comparison of the local line-of-sight path. Before performing the above-mentioned regional retrieval, the system calculates the three-dimensional coordinates of the RF center point formed by the spatial straight lines at the transmitting and receiving ends as the geometric guidance reference axis. Based on this guidance reference axis, the system generates the bottom boundary of the cone envelope according to the lateral divergence expansion of the antenna divergence angle. This combines the one-dimensional center guidance logic at the point scale with the three-dimensional array screening logic at the regional scale, ensuring that each line-of-sight profile line derived from the center point is strictly constrained within the three-dimensional solid interference space of the beam envelope. By detecting whether there are any ground elevation values ​​within the RF beam cone envelope that exceed the line-of-sight limit, the control module outputs the terrain obstruction status criterion of the RF propagation path.

[0031] Example 2: In the case of power line inspection in a remote mountainous area accompanied by high-voltage electromagnetic interference, a mobile ad hoc network including a test drone and three alternative airport relay nodes was built using a hardware-in-the-loop physics simulation platform. This simulation platform integrates a three-dimensional ray tracing core model and actively superimposes Gaussian white noise with a signal-to-noise ratio of 15dB at the RF output, mixing in multipath fading interference caused by vegetation scattering to simulate the actual channel attenuation process. To construct a problem intensity gradient comparison system, the relative flight speed variables of the test drone were set to 10m / s, 15.5m / s, and 25m / s, representing the time constraint of terrain physical occlusion approximation, and a preset switching threshold was established. Associated resource reservation overhead and link maintenance stability, when a preset switching threshold is reached. When the value approaches the lower limit of 0.3, high-frequency handover causes network topology oscillations and consumption of orthogonal frequency division multiplexing subcarrier resources. When the preset handover threshold is reached... When the upper limit of 0.8 is exceeded, the underlying radio frequency channel experiences substantial attenuation within the convergence period of the shortest path first algorithm, and the probability of communication interruption increases. This is based on the maximum radius of the first Fresnel zone. Based on the physical laws of diffusion with carrier frequency variation, and with a transmission frequency band of 2.4 GHz and a transceiver node spacing of 3000 m, the experiment limited the preset handover threshold to the range of 0.65 to 0.75. Within this range, the system maintains a time redundancy of no less than 120.5 ms for the target takeover node to lock onto a specific time slot. The preset handover threshold is dynamically mapped based on the current wireless environment signal-to-noise ratio (SNR). The process includes the control module extracting the measured SNR of the current wireless RF link. When the measured SNR is greater than or equal to 20 dB, the control module calls the 0.75 preset handover threshold δ nominal reference value. When the measured SNR is between 10 dB and 20 dB, a compensation of 0.05 is subtracted from the nominal reference value. When the measured SNR is less than 10 dB, the control module subtracts a compensation of 0.15 from the nominal reference value and sets this as the preset handover threshold for the current period. The module adjusts the preset switching threshold δ in a stepwise manner based on the measured signal-to-noise ratio (SNR) parameter. Under conditions of increased interference, it triggers the resource reservation handshake frame transmission logic in advance. The threshold segmentation judgment and compensation amount extraction are based on the deduction and setting of the log-normal shadow fading margin model. When the SNR is 20dB or above, the physical layer has sufficient redundancy space to resist multipath fading. Maintaining the nominal threshold can avoid network oscillations caused by false blockages. When the SNR drops to the 10dB to 20dB range or below 10dB, the demodulation degradation threshold of the RF receiver is significantly compressed. At this time, the dimensionless compensation amount of 0.05 or 0.15 deducted respectively is physically equivalent to forcibly shifting the predicted occurrence point of complete line-of-sight blockage forward by one or three standard control data packets' maximum retransmission communication time windows in the spatial geometric dimension, in order to offset the transmission vulnerability caused by the deterioration of the noise floor.

[0032] After platform startup, the control module continuously acquires real-time dynamic carrier phase differential positioning three-dimensional coordinate sequences at a sampling frequency of 10Hz. In the control group, which does not include spatial geometric and physical location parameters, the system relies on the received signal strength dropping to -85dBm to trigger route convergence. Test data shows that when dealing with sudden obstruction at 25m / s, the original electromagnetic wave signal is interfered with by 15dB of background noise. The route table reconstruction time in the control group reaches 350.5ms, causing high-frequency frame loss of inspection image data. The experimental group synchronously receives the original coordinate sequence containing interference noise, inputs it into the Kalman filter algorithm for processing, and the control module extracts the vertical offset of the highest obstruction point along the RF propagation path relative to the straight-line distance in space based on the filtered coordinate sequence, and calculates the current terrain intrusion depth. When the test drone approached the simulated ridge at a speed of 15.5 m / s, the control module of the test group calculated the maximum radius of the first Fresnel zone. The measured terrain intrusion depth was 9.6m. As the flight displacement increased, the actual terrain intrusion depth also increased. The terrain intrusion depth shows a monotonically increasing trend. When it reaches 6.8m, this value accounts for the maximum radius of the first Fresnel zone. The ratio reached 0.71. This judgment action was triggered before the received signal strength dropped to the attenuation critical point. The control module sent a resource reservation handshake frame containing a deterministic switching timestamp and bandwidth resource requirement parameters to the target takeover node through the current radio frequency link. In the partially missing control group with missing route metric value overwrite operation, the network protocol layer maintained the original route direction after sending the handshake frame, causing an average data buffer backlog of 180.2ms. In the experimental group, the route metric value of the current radio frequency link was overwritten to infinity after sending the resource reservation handshake frame. This overwrite parameter drove the shortest path first algorithm to converge the alternative path within 45.4ms.

[0033] The operational records of the problem intensity gradient control system show that, at occlusion approach velocities of 10 m / s and 15.5 m / s, the control signaling transmission gap time in the experimental group remained between 0 ms and 5.5 ms. A performance inflection point occurred when the physical occlusion approach velocity increased to 35 m / s, exceeding the system design setting. The shortened spatial reservation time triggered handshake frame conflicts, increasing the signaling transmission gap time to 85.2 ms. The aforementioned nonlinear measured data trajectory indicates that the calculated terrain intrusion depth... With the maximum radius of the first Fresnel zone The spatial geometric proportions are used to transform physical location features into network layer metrics within a specific speed boundary range, overwriting control commands, reducing routing convergence delay caused by background noise and passive signal attenuation, and maintaining the communication topology consistency of mobile data communication networks in complex terrain environments.

[0034] Example 3: In the inspection of power transmission lines in high-altitude and complex terrain, the computing module on the UAV is limited by hardware computing power and cannot retrieve massive spatial elevation data in real time. The mechanism of route switching triggered by a drop in received signal strength can cause signaling convergence delay when there is sudden ridge obstruction. The control module acquires a real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence containing longitude, latitude, and altitude. To avoid the large spatial matrix comparison calculation load generated by globally traversing the digital elevation model data, the control module extracts the current positioning coordinates of the UAV and the node coordinates of the associated relay nodes, and constructs a three-dimensional spatial system. A linear analytical equation is established connecting the two points. The control module generates a continuous sequence of three-dimensional sampling points along this linear analytical equation with a spatial step size of 10m. It then retrieves the surface entity elevation values ​​that match the latitude and longitude of each point in the locally stored digital elevation model data of the inspection corridor. The control module calculates the vertical difference between the original elevation value of each sampling point and the corresponding surface entity elevation value. From all output vertical differences, it selects the point with the largest positive upward extreme value, determines this point as the highest occlusion point, and labels its corresponding value as the terrain intrusion depth. Calculate the maximum radius of the first Fresnel zone. With terrain intrusion depth Next, the control module extracts the normal flight velocity component of the UAV in the direction approaching the highest obstruction point, calls up continuous adjacent elevation data to calculate the local slope sine of the current terrain, and sets the maximum radius of the first Fresnel zone as... With terrain intrusion depth The difference is used as the numerator, and the product of the normal flight velocity component and the local slope sine is used as the denominator to calculate the remaining line-of-sight dwell time of the UAV when it flies over the obstruction critical surface.

[0035] To transform 3D terrain features into explicit computational input, the computation module, after extracting the highest obstruction point, extracts feedforward reference points at equal intervals along the projection direction of the motion vector onto the horizontal plane. A virtual 2D collision slope is constructed using the elevation tangent of the highest obstruction point and the reference points. The sine of the angle between this virtual slope and the standard ground plane is calculated as the local slope sine. Simultaneously, the overall spatial motion vector of the UAV is projected onto a 3D normal axis orthogonal to this virtual slope. The modulus of this orthogonal projection ray is extracted and established as the normal flight velocity component. The control module deducts the fixed signaling interaction time consumed by locking the orthogonal frequency division multiplexing subcarrier resources from the remaining line-of-sight dwell time. Based on the obtained time difference, a deterministic handover timestamp for the target takeover node is generated. Spatial discrete sampling difference filtering and kinematic analysis are then performed. The dimensional transformation logic converts 3D terrain feature matching and time resource scheduling into a linear computation process. The control module uses this to output the time nodes of the underlying communication resources pre-allocated under the condition of limited computing power. When calculating the deterministic switching timestamp, the computation module performs boundary monitoring on the absolute value of the product of the normal flight velocity component and the local slope sine value. The computation module obtains the product of the normal flight velocity component and the local slope sine value, compares the absolute value with the preset motion threshold constant of 0.1 meters per second. When the absolute value of the product is less than the preset motion threshold constant, the computation module determines that the current radio frequency propagation path does not show a rapid obstruction approach, stops the deterministic switching timestamp generation instruction, maintains the existing network topology routing metric value in the current sampling period, and waits for the absolute value of the product to be greater than or equal to the preset motion threshold constant in the subsequent sampling period before resuming the calculation.

[0036] Example 4: When the system is deployed in a mountainous area to be inspected, the coordinate system alignment procedure is used to match the locally stored digital elevation model data of the inspection corridor with the real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence. The UAV stays at a ground reference point with known spatial coordinates. The control module reads the initial coordinates output by the real-time dynamic carrier phase differential positioning and retrieves the nominal altitude value of the corresponding geographical location in the digital elevation model data of the inspection corridor. The control module calculates the static system deviation between the initial coordinates and the nominal altitude value and uses it as a coordinate translation vector to compensate for the spatial straight-line distance in the subsequent solution. The corrected spatial analytical equation makes the extracted terrain intrusion depth... It represents the physical relative positional relationship between the drone and the ground surface entity.

[0037] In a radio electromagnetic environment with dynamic fluctuations, the signaling delay calibration process is used to determine the fixed signaling interaction duration required to generate a deterministic handover timestamp. Under line-of-sight link conditions, the control module sends a resource reservation handshake frame of preset data length to the target takeover node, while monitoring the delay distribution from the completion of transmission to the receipt of the resource lock confirmation acknowledgment. The control module calculates the average response duration of at least 50 handshake tests and adds twice the standard deviation to the average to establish the fixed signaling interaction duration. The control module writes the fixed signaling interaction duration into the memory register as the deduction benchmark for calculating the remaining line-of-sight dwell time. When the UAV approaches the obstruction critical surface with its normal flight velocity component, the deterministic handover timestamp calculated using the calibrated fixed signaling interaction duration drives the target takeover node to lock the orthogonal frequency division multiplexing subcarrier resources on time. The node communication topology completes the underlying link handover at the node where physical line-of-sight blockage occurs.

[0038] In multi-relay node collaborative deployment scenarios, the target takeover node selection procedure is used to determine the terrain clearance index of the communication link. The computing module obtains the three-dimensional coordinates of the currently idle candidate relay nodes and establishes a spatial scanning path between the current coordinates of the UAV and the coordinates of each candidate relay node. The control module extracts terrain undulation parameters from the digital elevation model data of the inspection corridor along the spatial scanning path, and defines the maximum bulge height relative to the reference line-of-sight plane in the terrain undulation parameters as the path occlusion extreme value. The control module uses the maximum radius of the first Fresnel zone... Subtract the extreme value of path occlusion to obtain the path clearance margin, and then use the path clearance margin to determine the maximum radius in the first Fresnel zone. The proportion in the data determines the terrain clearance index. The computing module selects the node with the highest terrain clearance index from multiple candidate relay nodes as the target takeover node to ensure radio frequency transmission within the reserved communication time slot.

[0039] Example 5: When the UAV high-altitude inspection communication system is in factory testing condition, the system calls the offline optimization procedure to determine the preset switching threshold based on the terrain intrusion depth. The dynamic mapping relationship of real-time change rate is used. The computing module reads the historical three-dimensional coordinate sequence of the UAV approaching physical obstacles along the flight path in a typical mountainous scene, and extracts the terrain intrusion depth calculated at two consecutive sampling periods. The calculation module calculates the terrain intrusion depth at two different time points. The difference is divided by a fixed sampling time interval to calculate the terrain intrusion depth. The real-time rate of change is calculated, and the calculation module constructs a linear compensation algorithm. The preset switching threshold is set to be equal to the pre-stored basic threshold constant minus the product of the gain coefficient and the real-time rate of change.

[0040] To calibrate the specific value of the gain coefficient, testers input a historical 3D coordinate sequence into the hardware-in-the-loop physical simulation platform, setting the optimization range of the gain coefficient to 0.01 to 0.10, and the optimization step size to 0.01. The system ran the topology control logic of the mobile ad hoc network under each step size parameter, synchronously recording the increase in system response time and communication link packet loss rate due to threshold changes. Test data indicated that when the gain coefficient was below 0.03, the reduction in the preset handover threshold was insufficient to compensate for the channel degradation caused by high-speed terrain approach, resulting in a service data packet loss rate greater than 5%. When the gain coefficient was above 0.08, the preset handover threshold... Prematurely triggering the routing metric overwrite command results in an invalid network topology switch and additional consumption of orthogonal frequency division multiplexing subcarrier resources. Based on the principle of minimizing packet loss, the computation module fixes the gain coefficient to 0.05. At this point, the network maintains continuous transmission of service data and minimizes signaling overhead. The offline optimization procedure presets the switching threshold. The dynamic adjustment strategy is transformed into a clear linear computation benchmark, establishing specific underlying quantization parameters for the route pre-scheduling mechanism under complex terrain.

[0041] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0042] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A communication method for unmanned aerial vehicles (UAVs) in remote mountainous areas with complex terrain, characterized in that, The method includes the following steps: Step S1: Obtain the real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence of the UAV and retrieve the digital elevation model data of the inspection corridor preset in the local memory. The real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence includes longitude, latitude, altitude, motion vector and carrier phase residual. Step S2: Determine the radio frequency beam cone envelope of the UAV relative to the associated relay node based on the real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence, and determine the terrain obstruction status of the radio frequency propagation path in combination with digital elevation model data. Step S3: Calculate the maximum radius of the first Fresnel zone of the radio frequency propagation path in a specific frequency band based on the terrain occlusion state, and extract the terrain intrusion depth corresponding to the terrain feature points along the radio frequency propagation path; Step S4: When the terrain intrusion depth is determined to be the proportion of the maximum radius of the first Fresnel zone to the preset handover threshold, a resource reservation handshake frame containing a deterministic handover timestamp, UAV identification and bandwidth resource requirement parameters is sent to the target takeover node using the current radio frequency link that has not yet been physically blocked. The preset handover threshold is dynamically mapped according to the signal-to-noise ratio of the current wireless environment. Step S5: After the resource reservation handshake frame is sent, the UAV overwrites the routing metric of the current radio frequency link to infinity in the network topology routing table to induce the ad hoc network routing protocol to start the neighbor node rediscovery mechanism and reconstruct the UAV's communication topology path.

2. The UAV communication method for complex terrain in remote mountainous areas according to claim 1, characterized in that, Step S3 includes: determining the spatial straight-line distance between the UAV and the associated relay node based on the real-time dynamic carrier phase differential positioning three-dimensional coordinate sequence, and calculating the three-dimensional coordinates of the center point of the radio frequency propagation path; retrieving digital elevation model data based on the three-dimensional coordinates of the center point, extracting the vertical offset of the highest obstruction point along the radio frequency propagation path relative to the spatial straight-line distance, and defining the vertical offset as the terrain intrusion depth.

3. The UAV communication method for complex terrain in remote mountainous areas according to claim 1, characterized in that, Step S4 includes: obtaining the real-time rate of change of terrain intrusion depth relative to flight displacement, and adjusting the preset switching threshold based on the real-time rate of change; when the real-time rate of change shows a monotonically increasing trend, lowering the value of the preset switching threshold so that the timing of sending the resource reservation handshake frame is earlier than the time when the terrain physical occlusion occurs.

4. The UAV communication method for complex terrain in remote mountainous areas according to claim 1, characterized in that, The resource reservation handshake frame is sent to the target takeover node so that the target takeover node can parse the bandwidth resource requirement parameters and lock the orthogonal frequency division reuse subcarrier resources corresponding to the specific time slot in the local communication scheduling table at the start time indicated by the deterministic handover timestamp. The deterministic handover timestamp is generated by the UAV based on the current flight speed, terrain intrusion depth and terrain slope change trend along the radio frequency propagation path.

5. The UAV communication method for complex terrain in remote mountainous areas according to claim 1, characterized in that, Step S5 and beyond also includes: the UAV monitors the convergence status of the routing table in real time, and actively disconnects the wireless radio frequency link with the associated relay node according to the deterministic switching timestamp before the current radio frequency link is physically interrupted; the UAV synchronously accesses the reserved communication time slot of the target takeover node, and uses the locked orthogonal frequency division multiplexing subcarrier resources to restore service data transmission, thereby realizing lossless link handover with a build-then-disconnect approach at the self-organizing network level.

6. The UAV communication method for complex terrain in remote mountainous areas according to claim 1, characterized in that, The target takeover node is selected from candidate airport nodes deployed at different geographical coordinates along the UAV inspection path; the method also includes the following steps: calculating the terrain clearance of the radio frequency propagation path between each candidate airport node and the UAV, wherein the terrain clearance is represented by the reciprocal of the proportion of Fresnel interference in each candidate path; selecting the candidate airport node with the highest terrain clearance value as the target takeover node to optimize the link stability after the handover.

7. The UAV communication method for complex terrain in remote mountainous areas according to claim 1, characterized in that, The operation of overwriting the routing metric value to infinity modifies the metric parameters in the routing control packet, enabling the shortest path first algorithm within the ad hoc network to perceive that the current radio frequency link is unreachable. This operation forces the network protocol layer to complete the calculation of alternative paths and the remapping of forwarding ports to the target takeover node in advance before signal interruption caused by physical obstruction occurs, thus eliminating the routing protocol convergence delay caused by terrain changes.

8. The UAV communication method for complex terrain in remote mountainous areas according to claim 1, characterized in that, The preset switching threshold is between 0.6 and 0.

8. This value range is used to initiate the link handover process before the received signal strength indicator drops sharply when the energy center area of ​​the radio frequency propagation path is intruded by terrain to a extent that it reaches 60% to 80%. This ensures the transmission reliability and channel continuity of critical control signaling when it travels through complex mountainous terrain conditions.

9. A UAV communication method for complex terrain in remote mountainous areas according to claim 1, characterized in that, The method also includes the following steps: real-time monitoring of the received signal strength indication of the current radio frequency link; when it is determined that the attenuation slope of the received signal strength indication exceeds the preset sudden change threshold within three consecutive sampling periods, and the terrain intrusion depth ratio has not yet reached the preset switching threshold, the routing metric value overwrite action in step S5 is forcibly executed to respond to sudden channel quality degradation caused by temporary ground moving obstacles not covered by digital elevation model data.

Images