Near field communication matching processing method and system for unmanned aerial vehicle in communication area without public network

Abstract

The invention provides an unmanned aerial vehicle near field communication matching processing method and system without a public network communication area, and relates to the technical field of intelligent control, and the method comprises the steps: generating a session key and an equipment authority topological graph according to a quantum random number; based on the equipment authority topological graph and the channel state tensor, communication capability matching is carried out on equipment in an unmanned aerial vehicle cluster, and a matching equipment pair set and a communication parameter mapping relation thereof are generated; performing joint coding on reconnaissance data, a control instruction and state information according to the session key, the equipment permission topological graph and the matching equipment pair set to generate a coding symbol stream; and according to the communication parameter mapping relation of the matching device to the set, performing directional transmission scheduling on the coding symbol stream to complete near field communication matching processing between the unmanned aerial vehicles. According to the invention, the cooperative operation efficiency and task success rate of the unmanned aerial vehicle rescue cluster in a public-network-free mountainous complex environment are improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, and in particular to a method and system for near-field communication matching and processing of unmanned aerial vehicles (UAVs) in areas without public network communication. Background Technology

[0002] The rescue command center dispatched a hybrid swarm of multi-rotor drones, including one command relay aircraft, three 3D modeling reconnaissance aircraft, and two supply delivery aircraft. The swarm was tasked with rapidly reconnaissance of the epicenter area, covering approximately 5 square kilometers, characterized by fragmented terrain and numerous landslides and collapses. The mission required the reconnaissance aircraft to transmit high-resolution orthophotos and thermal imaging data in real time to identify damaged buildings and sources of life, while the supply aircraft were required to accurately deliver first-aid kits according to instructions. Simultaneously, all drones needed to share their location and status information. The entire area had no public network signal, and the valleys, cliffs, and dense forests created an exceptionally complex electromagnetic propagation environment.

[0003] Traditional communication solutions employ simplified free-space models, which cannot cope with the unique "non-line-of-sight," "multipath deep fading," and "terrain shielding" effects of mountainous areas. For example, when reconnaissance aircraft fly into back slopes or canyons, the signal may drop sharply or even be interrupted due to mountain obstruction, resulting in the loss of critical disaster image data. Existing technologies lack the ability to perceive and model the three-dimensional spatial channel characteristics of rescue sites in real time, and cannot predict and avoid communication blind spots.

[0004] Emergency missions are highly phased (such as initial wide-area reconnaissance, mid-term focused search and rescue, and late-stage supply delivery), with drastic changes in the traffic and priorities of various UAVs at different stages. Traditional static or semi-static spectrum allocation methods cannot dynamically and globally optimize adjustments based on real-time mission requirements and channel changes. For example, when suspected signs of life are detected and all reconnaissance aircraft need to focus on the investigation, existing solutions cannot quickly and optimally reconfigure channels and time slots to ensure high-speed, low-latency transmission of multi-aircraft collaborative data streams in the area. Summary of the Invention

[0005] This invention provides a method and system for near-field communication matching processing of unmanned aerial vehicles (UAVs) in areas without public network communication, thereby improving the collaborative operation efficiency and mission success rate of UAV rescue clusters in complex mountainous environments without public networks.

[0006] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: Firstly, a method for near-field communication matching and processing of unmanned aerial vehicles (UAVs) in areas without public network communication, the method comprising: The near-field electromagnetic wave propagation environment is controlled to obtain three-dimensional channel characteristic data; the three-dimensional channel characteristic data is collected and a channel state tensor is generated. Based on the channel state tensor, a quantum annealing optimization model is constructed; according to the quantum annealing optimization model, the spectrum resource allocation problem is solved by simulating the quantum tunneling effect to obtain the carrier frequency, time slot offset and power control parameter set of each UAV. Based on the carrier frequency, time slot offset, and power control parameter set, an overlapping communication frame structure is constructed; based on the overlapping communication frame structure, a dynamic frame scheduling scheme is generated. Based on the dynamic frame scheduling scheme, a channel reciprocity feature sequence is extracted; a quantum random number is generated based on the channel reciprocity feature sequence; a session key and a device permission topology map are generated based on the quantum random number; based on the device permission topology map and the channel state tensor, the communication capabilities of devices in the UAV cluster are matched, and a set of matched device pairs and their communication parameter mapping relationship are generated. Based on the session key, device permission topology map, and matching device pair set, reconnaissance data, control commands, and status information are jointly encoded to generate an encoded symbol stream. Based on the mapping relationship of communication parameters of the matching device to the set, the coded symbol stream is scheduled for directional transmission to complete the near-field communication matching process between UAVs.

[0007] Secondly, the near-field communication matching and processing system for unmanned aerial vehicles (UAVs) in areas without public network communication includes: The control module is used to control the near-field electromagnetic wave propagation environment to obtain three-dimensional channel characteristic data; it collects three-dimensional channel characteristic data and generates a channel state tensor. The optimization module is used to construct a quantum annealing optimization model based on the channel state tensor; and to solve the spectrum resource allocation problem by simulating the quantum tunneling effect according to the quantum annealing optimization model, so as to obtain the carrier frequency, time slot offset and power control parameter set of each UAV. The module is used to construct an overlapping communication frame structure based on the carrier frequency, time slot offset, and power control parameter set; and to generate a dynamic frame scheduling scheme based on the overlapping communication frame structure. The matching module is used to extract the channel reciprocity feature sequence based on the dynamic frame scheduling scheme; generate quantum random numbers based on the channel reciprocity feature sequence; generate a session key and a device permission topology map based on the quantum random numbers; and perform communication capability matching on the devices in the UAV cluster based on the device permission topology map and the channel state tensor, generating a set of matched device pairs and their communication parameter mapping relationship. The processing module is used to jointly encode reconnaissance data, control commands, and status information based on the session key, device permission topology map, and matching device pair set to generate an encoded symbol stream; and to perform directional transmission scheduling of the encoded symbol stream according to the communication parameter mapping relationship of the matching device pair set to complete the near-field communication matching processing between UAVs.

[0008] The above-described solution of the present invention has at least the following beneficial effects: By employing proactive environmental perception and intelligent control, quantum-inspired global resource optimization, dynamic frame scheduling, physical layer intrinsic security generation, task-aware communication matching, and cross-layer joint coding transmission, the system systematically solves the core problems faced by UAV swarms in complex near-field environments without public network communication, such as unreliable communication, low efficiency, weak security, and poor adaptability. This enhances the collaborative operation efficiency and mission reliability of the swarm in scenarios such as emergency rescue. Attached Figure Description

[0009] Figure 1 This is a flowchart illustrating the near-field communication matching processing method for unmanned aerial vehicles (UAVs) in areas without public network communication, provided by an embodiment of the present invention. Figure 2 This is a schematic diagram of a near-field communication matching and processing system for unmanned aerial vehicles (UAVs) in a public network-free communication zone, provided by an embodiment of the present invention. Detailed Implementation

[0010] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0011] like Figure 1 As shown, an embodiment of the present invention proposes a near-field communication matching processing method for unmanned aerial vehicles (UAVs) in areas without public network communication. The method includes the following steps: Step 1: Control the near-field electromagnetic wave propagation environment to obtain three-dimensional channel characteristic data; collect the three-dimensional channel characteristic data and generate the channel state tensor; Step 2: Based on the channel state tensor, construct a quantum annealing optimization model; according to the quantum annealing optimization model, solve the spectrum resource allocation problem by simulating the quantum tunneling effect to obtain the carrier frequency, time slot offset and power control parameter set of each UAV. Step 3: Construct an overlapping communication frame structure based on the carrier frequency, time slot offset, and power control parameter set; generate a dynamic frame scheduling scheme based on the overlapping communication frame structure. Step 4: Based on the dynamic frame scheduling scheme, extract the channel reciprocity feature sequence; generate quantum random numbers according to the channel reciprocity feature sequence; generate session keys and device permission topology map according to the quantum random numbers; based on the device permission topology map and the channel state tensor, perform communication capability matching on the devices in the UAV cluster, and generate a set of matching device pairs and their communication parameter mapping relationship. Step 5: Based on the session key, device permission topology map, and matching device pair set, jointly encode the reconnaissance data, control commands, and status information to generate an encoded symbol stream; Step 6: Based on the mapping relationship of communication parameters of the matching device to the set, perform directional transmission scheduling of the encoded symbol stream to complete the near-field communication matching process between UAVs.

[0012] In this embodiment of the invention, by actively regulating the near-field electromagnetic environment and constructing a three-dimensional channel feature tensor, the system can accurately sense and adapt to complex communication environments. Combined with a quantum annealing optimization model for resource allocation, it can find global or near-global optimal solutions, thereby establishing stable and reliable communication links for UAV swarms in near-field regions with strong interference and significant multipath effects, greatly improving communication robustness. The design of dynamic frame scheduling and overlapping communication frame structures makes the communication signals random and dynamic in the time-frequency domain, making them difficult for external devices to continuously track and interfere with. Encryption using quantum random numbers generated based on channel reciprocity achieves physical layer security, significantly enhancing the communication's anti-interference and low-probability interception / detection capabilities. The quantum annealing optimization model can efficiently solve complex problems involving the joint allocation of spectrum, time slots, and power, avoiding the problem of traditional algorithms easily getting trapped in local optima. This allows for the maximization of limited spectrum resources, and the overlapping communication frame structure further improves spectrum efficiency, supporting efficient concurrent near-field communication in dense UAV swarms. By matching communication capabilities through device permission topology and channel status, appropriate communication parameters can be assigned to each pair of communication devices (the set of matched device pairs). This effectively reduces the drone's transmission power and extends the overall flight time of the cluster while ensuring communication quality. Quantum random numbers are generated using channel reciprocity, and session keys are produced accordingly, enabling one-time pad or high-security key distribution. This process is entirely based on the characteristics of the wireless channel itself, without relying on pre-shared keys or complex public key infrastructure, constructing a lightweight, intrinsic security system that effectively prevents eavesdropping and man-in-the-middle attacks. The entire processing flow (sensing, optimization, scheduling, matching, encryption, and transmission) is highly automated and does not rely on a central node or public network infrastructure. The generated device permission topology dynamically reflects the cluster's network relationships and capabilities. Even if some nodes fail or are dynamically added, the system can quickly re-match and schedule, demonstrating strong self-organization and resilience.

[0013] In a preferred embodiment of the present invention, step 1 involves regulating the near-field electromagnetic wave propagation environment to obtain three-dimensional channel characteristic data; acquiring the three-dimensional channel characteristic data and generating a channel state tensor, including: Step 1.1: Dynamically control the near-field electromagnetic wave propagation path in the area without public network communication by deploying a reconfigurable smart surface array to form multi-dimensional spatial beamforming; based on the multi-dimensional spatial beamforming, configure the beam pointing parameters of the UAV-borne phased array antenna, specifically including: Based on topographic survey data (including collapse points, landslides, and valley distribution) of a 5-square-kilometer area in the earthquake zone, RIS arrays were deployed at 3-5 key high points (such as partially collapsed highlands or stable cliff platforms) within the coverage area of ​​the command relay. Each RIS array consisted of no fewer than 100 electromagnetic control units, supporting continuous adjustment of the amplitude and phase of the incident electromagnetic waves. After deployment, initialization commands were sent to all RIS arrays via the command relay to complete time synchronization (based on BeiDou timing) and basic parameter configuration (the operating frequency band matched the UAV communication frequency band, and the initial reflection mode set to omnidirectional scattering).

[0014] The command relay receives the initial position information of each UAV in real time (acquired via onboard GPS / IMU) and, combined with a 3D terrain model, identifies line-of-sight obstruction areas (such as back slopes and canyons) in the current communication link. For these areas, the command relay sends control commands to the corresponding RIS array, adjusting the reflection phase of each unit in the array to construct an indirect communication path between the UAV, RIS, and the command relay. Simultaneously, the RIS array feeds back its real-time reflection status to the command relay, which calculates a multi-dimensional spatial beamforming scheme based on the multipath superposition principle. This scheme includes directional beam parameters for the command relay, three reconnaissance aircraft, and two delivery aircraft, ensuring that each UAV receives independent beam coverage.

[0015] The command relay sends the generated multi-dimensional spatial beamforming parameters (including beam pointing angle, beamwidth, and gain configuration) to the airborne phased array antenna control systems of each UAV. Based on the received parameters, the antenna system adjusts the phase difference between array elements to precisely point the antenna beam at the corresponding RIS array or the command relay (when line-of-sight is available). For example, for a reconnaissance aircraft flying into a canyon, its antenna beam will be configured to point towards the RIS array at the canyon entrance, ensuring non-line-of-sight communication via RIS reflection; while delivery aircraft in open areas will directly point their beam at the command relay to reduce transmission loss. After configuration, the UAV reports the antenna status to the command relay, forming a closed-loop confirmation.

[0016] Step 1.2: Based on the beam pointing parameters, the UAV performs multi-point scanning measurements in three-dimensional space to acquire channel impulse response data at different spatial locations, time slices, and frequency dimensions; the channel impulse response data is then processed using time-frequency analysis to extract three-dimensional channel feature data, specifically including: The command relay, combining mission requirements in the earthquake zone (focusing on landslide areas and areas with concentrated damaged buildings) with terrain data, planned differentiated 3D scanning paths for three 3D modeling reconnaissance aircraft. Each path divided the earthquake zone into 50m × 50m grid cells, with three measurement points at different altitudes (50m, 80m, and 120m, adapting to different building heights and terrain undulations) set within each grid cell, ensuring coverage of the entire 5 square kilometer area. Simultaneously, time slices (100ms measurement time per point, 50ms time interval between adjacent points) and frequency resources were allocated to each measurement point to avoid signal interference during multi-aircraft measurements.

[0017] Each reconnaissance aircraft flew along the planned path. Upon reaching each measurement point, it fixed its antenna attitude according to the beam pointing parameters configured in step 1.1 and initiated data acquisition. During acquisition, the command relay and RIS array transmitted known pilot signals (using the Zadoff-Chu sequence, which has good autocorrelation) to the UAV. The UAV simultaneously received the pilot and reflected signals, recording channel impulse response (CIR) data at different spatial locations (annotated in real time by the onboard GPS / IMU with an accuracy of 0.1m), different time slices (timestamped by BeiDou timing), and different frequency dimensions (acquired sequentially according to preset sub-channels). Simultaneously, the UAV transmitted the raw CIR data back to the command relay in real time to ensure no data loss.

[0018] The command relay preprocesses the received raw CIR data. First, it eliminates environmental noise through bandpass filtering (the filter frequency band matches the pilot signal frequency band), and then removes direct wave interference from multipath signals. Subsequently, time-frequency analysis is performed: in the time dimension, delay spread characteristics of the CIR data are extracted using a sliding window (window length 20ms) to reflect the temporal distribution of the multipath signal; in the frequency dimension, Fourier transform is performed on the CIR data for each time slice to obtain the frequency domain response, extracting Doppler frequency offset (reflecting the impact of UAV motion on the signal) and frequency-selective fading characteristics; in the spatial dimension, combined with the location information of different measurement points, the signal amplitude variation within the beam coverage area is analyzed to extract spatial attenuation characteristics. Finally, three-dimensional channel feature data, including key parameters such as delay spread, Doppler frequency offset, frequency fading coefficient, and spatial attenuation coefficient, are extracted from the time, frequency, and spatial dimensions.

[0019] Step 1.3: Based on the three-dimensional channel feature data, multi-dimensional information is collected through distributed sensing nodes. This multi-dimensional information includes spatial location information, timestamp information, and frequency domain response information. Spatiotemporal alignment calibration is performed on the multi-dimensional information to obtain calibrated multi-dimensional information, specifically including: Distributed sensor nodes are integrated on the command relay, each UAV, and the deployed RIS array to form a multi-source data acquisition network. Each sensor node synchronously collects three types of core information: first, spatial location information; the positions of the command relay and UAVs are determined by a combination of GPS / IMU (updating frequency 10Hz), and the position of the RIS array is determined by fixed coordinates (calibrated during deployment), while the attitude angles (pitch, roll, and heading) of each device are also recorded; second, timestamp information; all sensor nodes are synchronized based on the BeiDou time synchronization system to ensure a timestamp accuracy of 1ms for the collected data; and third, frequency domain response information; in addition to the frequency fading coefficient extracted in step 1.2, auxiliary parameters such as signal-to-noise ratio (SNR) and signal strength (RSSI) of each frequency sub-channel are supplemented and collected. The collected multi-dimensional information is correlated and labeled with the three-dimensional channel feature data from step 1.2 to form a dataset of channel features, location, time, and frequency domain assistance.

[0020] Spatiotemporal Deviation Detection and Calibration: First, time deviation calibration is performed. The relay extracts the timestamps of all data. For data with time offsets (e.g., timestamps lagging due to communication delays in some UAVs), linear interpolation is used to correct the timestamps, ensuring time synchronization of all data for the same measurement event. Next, spatial deviation calibration is performed. Combining the 3D terrain model with the attitude angle data of each device, the GPS positioning error of the UAV is corrected. For example, when the UAV is in a canyon, the GPS signal may be affected by multipath propagation, causing deviations. In this case, by matching the attitude angles with the terrain data, the positioning coordinates are adjusted to the actual flight position. For the reflected signal path of the RIS array, the signal propagation distance is recalculated based on its fixed coordinates and the UAV's position, correcting the path loss deviation in the spatial attenuation coefficient.

[0021] Consistency checks are performed on the calibrated multi-dimensional information. This involves comparing data collected by different devices within the same time and area (e.g., SNR data collected by command relay aircraft and reconnaissance aircraft on the same channel), and eliminating abnormal data with deviations exceeding a threshold (e.g., SNR deviation greater than 3dB). Simultaneously, considering the phased requirements of the earthquake zone mission, key data matching the current mission are retained. For example, in the initial wide-area reconnaissance phase, spatial attenuation and frequency response data over a wide coverage area are retained; in the mid-term focused search and rescue phase, channel characteristic data of areas with life-threatening heat sources are strengthened, ensuring the accuracy of the calibrated dataset.

[0022] Step 1.4 involves tensorizing the calibrated multi-dimensional information according to spatial, temporal, and frequency dimensions to construct a three-dimensional channel feature matrix. The three-dimensional channel feature matrix is ​​then normalized and its dimensions compressed to generate a channel state tensor representing the spatial correlation of the channel. Specifically, this includes: The three-dimensional structure of the tensor is determined, namely, the spatial dimension, the temporal dimension, and the frequency dimension. The spatial dimension is indexed by the spatial location combination of the UAV-RIS-command relay, and includes parameters such as UAV position coordinates, RIS array number, and spatial attenuation coefficient. The temporal dimension is indexed by the calibrated timestamp and includes time-varying features such as delay spread and Doppler frequency offset. The frequency dimension is indexed by the frequency sub-channel number and includes frequency domain parameters such as frequency fading coefficient, SNR, and RSSI. The calibrated multi-dimensional information is filled into the above three-dimensional structure to construct an initial three-dimensional channel feature matrix. Each element in the matrix corresponds to a complete set of channel features at a specific spatial location, time, and frequency. Based on this, tensor decomposition is performed on the normalized three-dimensional matrix using the PARAFAC (parallel factor) decomposition algorithm, which can decompose the three-dimensional tensor (denoted as X∈R) into its components. (I×J×K) Where I is the number of spatial location combinations, J is the number of time slices, and K is the number of frequency sub-channels, it is uniquely decomposed into the product of three independent low-dimensional factor matrices and a core tensor, as specifically implemented below: The first step is to initialize three factor matrices A∈R (I×F) B∈R (J×F) C∈R (K×F) , where F is the number of common factors, i.e. the compressed target dimension, which corresponds to the feature projection matrices of the three dimensions of space, time and frequency, respectively.

[0023] The second step involves iteratively optimizing the three factor matrices using the alternating least squares (ALS) method. Each time, two of the matrices are fixed, and the third matrix is ​​solved based on minimizing the error between the original tensor and the reconstructed tensor (using the mean square error MSE as the objective function). During the iteration process, a regularization term is used to avoid overfitting and ensure that the decomposition result reflects the true characteristics of the channel.

[0024] The third step involves determining the optimal number of common factors F through cross-validation. During validation, the dataset is divided into a training set (70%) and a test set (30%), with the criterion of minimizing the reconstruction error of the test set. The final value of F is typically 1 / 5 to 1 / 3 of the original dimension (e.g., compressing the spatial dimension from 500 position combinations to 100 common factors). After decomposition, the three low-dimensional factor matrices A, B, and C extract the channel correlation features in the spatial dimension, the dynamic change features in the time dimension, and the frequency domain response features in the frequency dimension, respectively, discarding redundant information in the original data.

[0025] Finally, the three low-dimensional factor matrices are multiplied again with the optimized core tensor to regenerate a three-dimensional channel state tensor. This tensor retains the three-dimensional spatial-temporal-frequency correlation and improves data processing efficiency through dimensionality compression. The generated channel state tensor can be transmitted in real time to the communication scheduling system of the UAV swarm. Based on the spatial correlation characteristics reflected by the tensor, the scheduling system can quickly identify the channel quality in different areas, allocate optimal spectrum resources for multi-UAV collaborative communication in key search and rescue areas, reduce data transmission latency, and ensure the stable transmission of disaster images and life detection data.

[0026] To eliminate the dimensional differences between parameters of different dimensions (e.g., spatial attenuation coefficient in dB, Doppler frequency offset in Hz), the initial three-dimensional channel feature matrix is ​​normalized. The Z-Score normalization method is used to calculate the mean and standard deviation of each dimension's parameters, converting the original parameters into standardized data with a mean of 0 and a standard deviation of 1. Considering the high dimensionality of the initial matrix (hundreds of location combinations in the spatial dimension, thousands of time slices in the temporal dimension, and hundreds of sub-channels in the frequency dimension), dimensionality compression is necessary to improve subsequent processing efficiency. Tensor decomposition algorithms (such as PARAFAC decomposition) are used to decompose the normalized three-dimensional matrix, extracting low-dimensional factor matrices that reflect the core characteristics of the channel. During the decomposition process, with the goal of preserving channel spatial correlation, cross-validation is used to determine the optimal compressed dimension (typically compressed to 1 / 5-1 / 3 of the original dimension) to ensure that the compressed data can still accurately represent the channel correlation characteristics between different spatial locations. Finally, the compressed low-dimensional factor matrix is ​​recombined to generate a channel state tensor that characterizes the spatial correlation of the channel. This tensor can be fed back to the communication scheduling system of the UAV swarm in real time, providing data support for dynamic spectrum allocation and communication link optimization.

[0027] In a preferred embodiment of the present invention, step 2 involves constructing a quantum annealing optimization model based on the channel state tensor; and solving the spectrum resource allocation problem by simulating the quantum tunneling effect according to the quantum annealing optimization model to obtain the carrier frequency, time slot offset, and power control parameter set for each UAV, including: Step 2.1, based on the spatial correlation and time-frequency characteristics in the channel state tensor, construct the objective function for spectrum resource allocation, including: Step 2.11, separating the spatial dimension component, time dimension component, and frequency dimension component from the channel state tensor; calculating the spatial isolation matrix between UAV nodes based on the spatial dimension component, extracting the channel coherence time series based on the time dimension component, and obtaining the frequency domain flatness index based on the frequency dimension component, specifically including: Based on the channel state tensor generated in step 1.4 (representing the three-dimensional correlation features of channel space-time-frequency), three independent dimensional components are separated from the factor matrix after tensor decomposition, and then the core parameters supporting the construction of the objective function are calculated respectively: Dimensional component separation involves extracting the spatial dimension factor matrix A, temporal dimension factor matrix B, and frequency dimension factor matrix C obtained from the decomposition in step 1.4 from the channel state tensor. Spatial dimension factor matrix A contains the correlation features of all UAV-RIS-command relay spatial location combinations; temporal dimension factor matrix B contains the channel dynamic change features of different time slices; and frequency dimension factor matrix C contains the response features of each frequency sub-channel. Through matrix transposition and feature alignment, it is ensured that the indices of the three components correspond one-to-one with the UAV number, timestamp, and sub-channel number. Spatial isolation matrix calculation involves calculating the spatial isolation between any two UAVs (including the command relay) based on the UAV position coordinates, RIS array deployment location, and spatial attenuation coefficient in spatial dimension factor matrix A, using a logarithmic distance path loss model combined with a beam orthogonality algorithm. Specifically, for each pair of UAVs (i, j), the total signal attenuation is first calculated based on their three-dimensional coordinates and the reflection path of the intermediate RIS array. Then, the orthogonality coefficient between the beams is calculated by combining the beam pointing angles of the phased array antennas on the two UAVs (higher orthogonality means less mutual interference). The total attenuation and orthogonality coefficient are weighted and summed with a weight of 0.6:0.4 to obtain the spatial isolation value of the UAV pair (i, j). Finally, a 6×6 spatial isolation matrix is ​​constructed (6 represents the total number of UAVs in the swarm, including 1 command relay aircraft, 3 reconnaissance aircraft, and 2 delivery aircraft). Channel coherence time series extraction involves extracting the Doppler frequency offset and delay spread characteristic parameters corresponding to each UAV from the time dimension factor matrix B, calculating the channel coherence time of each UAV in each time slice (i.e., the maximum time interval for the channel to remain stable), and arranging the coherence times of all time slices in timestamp order to form the channel coherence time series of each UAV. At the same time, a moving average filter (with a window length of 5 time slices) is used to eliminate abnormal fluctuations and ensure the smoothness of the sequence.

[0028] The frequency domain flatness index is obtained by calculating the variance of the fading coefficients of each frequency sub-channel and its five adjacent sub-channels for each frequency sub-channel in the frequency dimension factor matrix C (the smaller the variance, the flatter the frequency domain). The variance value is then normalized (the normalization range is 0-1, and the smaller the variance, the closer the flatness index is to 1) to obtain the frequency domain flatness index for each frequency sub-channel. The flatness indices of all sub-channels are arranged in frequency order to form a frequency domain flatness sequence.

[0029] Step 2.12: Calculate the normalized spatial orthogonality coefficients between each UAV pair based on the spatial isolation matrix; determine the normalized time window length based on the channel coherence time series; and divide the normalized subcarrier aggregation region based on the frequency domain flatness index, specifically including: Based on the spatial isolation matrix, channel coherence time series, and frequency domain flatness index obtained in step 2.11, normalized parameters are further calculated and aggregation regions are divided to provide a basis for index mapping: The calculation of the normalized spatial orthogonality coefficient involves statistically analyzing all elements in the spatial isolation matrix (i.e., the spatial isolation values ​​between UAV pairs) to determine the maximum and minimum values. A linear normalization method is then used to map each isolation value to the interval [0,1], yielding the normalized spatial orthogonality coefficient. The closer the coefficient is to 1, the better the spatial isolation between the two UAVs, and the lower the possibility of signal interference. The determination of the normalized time window length involves extracting the maximum value of the mean coherent time series of each UAV channel (denoted as Tmax). Using Tmax as a benchmark, the mean coherent time of each UAV is normalized (the normalized range is [0.5,1], ensuring the time window length is not less than the minimum task scheduling unit). Combining this with the task cycle of the UAV swarm (e.g., a task cycle of 10s for the wide-area reconnaissance phase and 5s for the key search and rescue phase), the normalized mean coherent time is multiplied by the task cycle to obtain the normalized time window length for each UAV (i.e., the minimum unit duration for time-domain scheduling). Normalized subcarrier aggregation region division involves traversing the frequency domain flatness sequence and filtering subchannels based on whether the flatness index is greater than a preset threshold (e.g., 0.8). Subchannels that continuously meet the threshold condition are divided into an aggregation unit. The number of subchannels in each aggregation unit is counted, and the normalization process is performed according to the proportion of the number of subchannels in the aggregation unit to the total number of subchannels to obtain the normalized subcarrier aggregation region (each region corresponds to a set of aggregateable subchannels, and the normalized value reflects the amount of subchannel resources in that region).

[0030] Step 2.13 maps the normalized spatial orthogonality coefficients to a spatial multiplexing gain index, the normalized time window length to a time-domain scheduling efficiency index, and the normalized subcarrier aggregation region to a spectrum utilization index, specifically including: The three normalized parameters obtained in step 2.12 are mapped to the core optimization indicators of the objective function, ensuring that the indicators are directly related to the core requirements of spectrum resource allocation: The spatial multiplexing gain index mapping establishes a positive correlation between the normalized spatial orthogonality coefficient and the spatial multiplexing gain. A coefficient of 1 represents the maximum spatial multiplexing gain (e.g., 1.5), while a coefficient of 0 represents the minimum (e.g., 0.5). Intermediate values ​​are calculated through linear interpolation to obtain the spatial multiplexing gain index for each UAV paired with other UAVs. This index reflects the spatial multiplexing efficiency when multiple aircraft simultaneously utilize spectrum resources. The temporal scheduling efficiency index mapping maps the normalized time window length to a temporal scheduling efficiency index. The closer the window length is to the mission cycle, the closer the scheduling efficiency index is to 1 (indicating less time-domain resource waste); the smaller the window length, the lower the scheduling efficiency index (requiring more handovers and reducing efficiency). During the mapping process, mission priority is considered (reconnaissance aircraft have higher priority than delivery aircraft, and their scheduling efficiency index weighting coefficient is increased by an additional 20%) to ensure the temporal resource guarantee for key missions. The spectrum utilization index mapping is a process of mapping the normalized value of the normalized subcarrier aggregation region to the spectrum utilization index. The more sub-channels and the higher the flatness of the aggregation region, the closer the spectrum utilization index is to 1 (aggregation can reduce the occupation of the protection bandwidth and improve the transmission rate per unit spectrum). The corresponding spectrum utilization index is calculated for each aggregation region to provide a basis for subsequent spectrum allocation. Step 2.14: Multiply the spatial reuse gain index, temporal scheduling efficiency index, and spectrum utilization index by their corresponding preset weighting coefficients and sum them to construct the basic objective function. In the basic objective function, set interference suppression constraints and power consumption constraints to form the objective function for spectrum resource allocation, specifically including: Based on the three core indicators obtained from step 2.13, a complete objective function for spectrum resource allocation is constructed by combining the constraint terms: The basic objective function is constructed by pre-setting weight coefficients based on the task requirements of the UAV swarm. The weight of the spatial reuse gain index is set to 0.4 (to adapt to the needs of multi-UAV collaborative communication), the weight of the temporal scheduling efficiency index is set to 0.3 (to adapt to the phased changes of the task), and the weight of the spectrum utilization rate index is set to 0.3 (to adapt to the efficient use of limited spectrum resources). The basic objective function is obtained by multiplying each of the three indices by their corresponding weight coefficients and then summing them. The core objective is to maximize the weighted sum of the three indices to achieve optimal multi-dimensional resource utilization efficiency.

[0031] Setting constraints involves adding disturbance suppression and power consumption constraints to the basic objective function: The interference suppression constraint term sets an interference power threshold (e.g., not exceeding -80dBm) between any two UAVs based on the spatial isolation matrix. By setting a penalty coefficient (which increases linearly with the difference between the interference value and the threshold when the interference exceeds the threshold), the objective function is ensured to avoid severe interference during the optimization process. The power consumption constraint term sets a maximum communication power threshold (e.g., 20dBm) for each UAV, taking into account the UAV's battery capacity (e.g., 4 hours of flight time for a reconnaissance aircraft and 3 hours for a delivery aircraft) and communication power requirements. When the power consumption exceeds the threshold, a negative correction factor is used to limit high-power-consumption resource allocation schemes.

[0032] The basic objective function is added to the two constraint terms to form the final spectrum resource allocation objective function. This function not only maximizes the efficiency of multi-dimensional resource utilization, but also effectively suppresses interference and controls power consumption.

[0033] The penalty coefficient increases linearly with the severity of the interference exceeding the threshold. It is also dynamically adjusted based on the spatial isolation of the drone to ensure that minor interference is only subject to moderate penalties, while severe interference is forcibly avoided. The specific value range and conditions are as follows: The penalty coefficient has a range of values ​​of [0.5, 5.0] (dimensionless, matching the weights of each indicator in the objective function (0.3~0.4) to avoid excessive or insufficient penalty), and the linear growth slope is 0.25 / dBm (i.e., for every 1dBm increase in the interference difference, the penalty coefficient increases by 0.25).

[0034] For example, using the interference power threshold (-80dBm) between any two drones as a benchmark, the interference difference ΔP is defined as: actual interference power - interference power threshold (ΔP > 0 indicates exceeding the threshold, ΔP ≤ 0 indicates not exceeding the threshold). Combined with the spatial isolation value in the spatial isolation matrix (denoted as S, ranging from [0,1], where S = 1 indicates complete isolation and S = 0 indicates no isolation), the penalty coefficient is determined for each scenario. Scenario 1: If the interference threshold is not exceeded (ΔP≤0), the penalty coefficient is 0.5 (minimum penalty, only used as a placeholder for basic constraints, and does not affect the core optimization of the objective function). Applicable conditions: The actual interference power is ≤-80dBm. Regardless of the spatial isolation level, it is considered acceptable interference, and only the minimum penalty is applied.

[0035] Scenario 2: Slightly exceeding the threshold (0 < ΔP ≤ 10dBm), the penalty coefficient is 0.5 + 0.25 × ΔP (linearly increasing, range: (0.5, 3.0]). Applicable conditions: actual interference power is between (-80dBm, -70dBm], and spatial isolation S ≥ 0.6 (the UAV has high beam orthogonality and the interference impact is small).

[0036] Scenario 3: For moderate exceedance of the threshold (10 < ΔP ≤ 20 dBm), the penalty coefficient is 3.0 + 0.25 × (ΔP - 10) (linearly increasing, range: (3.0, 5.0]). Applicable conditions: actual interference power is between (-70 dBm, -60 dBm], or spatial isolation is 0.3 ≤ S < 0.6 (the UAV has moderate beam orthogonality, and the interference effect is relatively obvious).

[0037] Scenario 4: Severely exceeding the threshold (ΔP > 20dBm), the penalty coefficient is forcibly set to 5.0 (maximum penalty, significantly reducing the objective function value of the resource allocation scheme, almost discarding it by the optimization model). Applicable conditions: Actual interference power > -60dBm, or spatial isolation S < 0.3 (low beam orthogonality of the UAV, severe interference significantly affects the communication link). Example: ΔP = 25dBm (actual interference power -55dBm), S = 0.2, penalty coefficient = 5.0; even if ΔP = 18dBm (actual interference power -62dBm) but S = 0.2, the value is still set to 5.0 according to this scenario.

[0038] The negative correction factor increases linearly with the magnitude of power exceeding the threshold and adapts to differences in drone endurance (delivery drones have shorter endurance and are more sensitive to power). It suppresses high-power schemes by reducing the objective function value. The specific value range and conditions are as follows: The negative correction factor has an overall range of [-0.8, 0] (dimensionless, matching the weights of each indicator in the objective function (0.3~0.4), with negative values ​​indicating a reduction in the objective function score). Its linear growth slope is -0.1 / dBm (meaning for every 1dBm increase in power exceeding the threshold, the absolute value of the correction factor increases by 0.1, resulting in a greater reduction in the objective function score). For reconnaissance aircraft (4-hour endurance), the correction factor ranges from [-0.6, 0] (higher endurance redundancy, moderate penalty); for delivery aircraft (3-hour endurance), the correction factor ranges from [-0.8, 0] (lower endurance redundancy, stricter penalty). For example, using the maximum communication power threshold (20dBm) for each drone as a benchmark, the power difference ΔPtx is defined as: actual communication power - maximum communication power threshold (ΔPtx > 0 indicates exceeding the threshold, ΔPtx ≤ 0 indicates not exceeding). The negative correction factor is then determined based on the drone type and scenario. If the power threshold is not exceeded (ΔPtx≤0), the correction factor is 0 (no negative correction, does not affect the objective function score); applicable condition: actual communication power ≤20dBm, regardless of UAV type, power consumption is considered reasonable. If the power threshold is slightly exceeded (0<ΔPtx≤6dBm), the correction factor for reconnaissance aircraft is 0-0.1×ΔPtx (linearly increasing, range: (0,-0.6]); applicable condition: actual communication power is between (20dBm, 26dBm], with less endurance pressure. For delivery aircraft, the correction factor is 0-0.13×ΔPtx (linearly increasing, range: (0,-0.78], approximately -0.8); applicable condition: actual communication power is between (20dBm, 26dBm], with greater endurance pressure and more severe penalties.

[0039] Severely exceeding the power threshold (ΔPtx > 6dBm); Reconnaissance aircraft, i.e., the correction factor is forcibly set to -0.6 (maximum negative correction, significantly reducing the objective function score); Applicable conditions: actual communication power > 26dBm, endurance pressure increases dramatically, strictly limited. Delivery aircraft, i.e., the correction factor is forcibly set to -0.8 (maximum negative correction, this option is almost abandoned); Applicable conditions: actual communication power > 26dBm, endurance cannot support long-term missions, high power allocation is strictly prohibited.

[0040] Based on the communication requirements and topology of the UAV swarm, a set of constraints for spectrum resource allocation is determined, specifically including: Combining the communication requirements and topology of the UAV swarm (a star topology with a command relay as the core and RIS as an auxiliary), the following set of constraints is determined: Bandwidth constraints are implemented based on the different operational needs of various UAVs. For example, three reconnaissance aircraft need to transmit high-resolution imagery and thermal imaging data, with a minimum bandwidth of 20MHz per aircraft; two delivery aircraft need to transmit control commands and location information, with a minimum bandwidth of 5MHz per aircraft; and the command relay aircraft needs to aggregate all data and issue commands, with a minimum bandwidth of 50MHz. Latency constraints ensure that the image data transmission latency for reconnaissance aircraft does not exceed 500ms, the command reception latency for delivery aircraft does not exceed 100ms, and the communication latency between the command relay aircraft and the RIS array does not exceed 50ms, ensuring real-time emergency response. Power constraints limit the communication power of each UAV to a preset threshold (e.g., 20dBm), and the total power consumption does not exceed 30% of the battery capacity per hour (to avoid affecting flight endurance). Interference constraints ensure that the co-channel interference power between any two UAVs does not exceed -80dBm, and the interference power between a UAV and the RIS array does not exceed -90dBm, guaranteeing communication link stability. Topology constraints mean that the spectrum resource allocation for all UAVs must be centered on the command relay, ensuring that the command relay can receive data from all UAVs and issue scheduling instructions. At the same time, the spectrum resource configuration of the RIS array must be synchronized with the UAVs within the coverage area.

[0041] Step 2.2: Convert the objective function and constraint set into the Ising model expression form to construct the Hamiltonian of the quantum annealing optimization model; based on the ground state energy distribution characteristics of the Hamiltonian, initialize the spin state and coupling strength parameters of the qubit, specifically including: Based on the objective function and constraint set obtained in step 2.1, this step constructs the Hamiltonian through the Ising model transformation and initializes the relevant quantum annealing parameters to provide model support for subsequent optimization solutions. The specific implementation process is as follows: The Ising model transformation between the objective function and constraints: The Ising model uses the spin state (+1 or -1) of qubits to represent binary decision variables for resource allocation (e.g., 1 indicates allocation of a certain spectrum resource, -1 indicates no allocation). First, the decision variables in the spectrum resource allocation problem (carrier frequency selection, time slot allocation, power level selection) are mapped to the spin state of qubits. For example, 50 available frequency sub-channels (divided into 1MHz intervals from 800MHz to 2.4GHz) correspond to 50 qubits, and the spin state of each qubit indicates whether the sub-channel is allocated to a certain drone. Similarly, 10 time slots (each time slot is 100ms) correspond to 10 qubits, and 3 power levels (10dBm, 15dBm, 20dBm) correspond to 2 qubits (represented by binary encoding). Subsequently, the objective function constructed in step 2.1 is converted into the energy term of the Ising model, and the set of constraints is converted into penalty terms (when the constraints are not met, the penalty terms will increase the system energy, prompting the optimization process to avoid such solutions), thus forming the objective expression of the Ising model.

[0042] Construction of the Hamiltonian in the quantum annealing optimization model: The Hamiltonian is the core of the quantum annealing model, used to describe the energy state of the system, and consists of transverse magnetic field terms and longitudinal magnetic field terms. The longitudinal magnetic field term, constructed based on the objective expression of the Ising model, reflects the energy distribution corresponding to the objective function and constraints. Its coefficients are determined by the weighting coefficients of the objective function and the penalty coefficients of the constraints. For example, the weighting coefficients corresponding to the spatial multiplexing gain index are directly mapped to the coupling strength of the corresponding qubit in the longitudinal magnetic field term. The transverse magnetic field term is used to simulate the quantum tunneling effect, helping the system escape local optima. Its initial strength is set according to the strictness of the constraints (the stricter the constraints, the stronger the initial transverse magnetic field). Adding the longitudinal and transverse magnetic field terms forms the complete Hamiltonian of the quantum annealing optimization model. The ground state energy of this Hamiltonian corresponds to the optimal solution for spectral resource allocation.

[0043] Quantum bit parameter initialization: Spin state initialization involves assigning an initial spin state (+1 or -1) to each qubit based on historical spectrum allocation data of the UAV swarm (random initialization is used if no historical data is available). For example, qubits corresponding to the core spectrum resources of the command relay (such as 1.8GHz-1.9GHz) are initialized to +1 (preferred allocation), while qubits corresponding to spectrum resources susceptible to interference are initialized to -1 (not allocated for the time being). Coupling strength parameter initialization involves calculating the coupling strength between each qubit based on the ground state energy distribution characteristics of the Hamiltonian. The coupling strength of qubits corresponding to resources with high correlation (such as sub-channels in the same aggregation region) is set to positive values ​​(to promote cooperative allocation), while the coupling strength of resources with low correlation is set to negative values ​​(to avoid conflict allocation). The absolute value range of the coupling strength is set to [0.1, 1.0] to ensure the stability of the quantum annealing process.

[0044] Step 2.3 involves dynamically adjusting the transverse magnetic field strength during quantum annealing by simulating the quantum tunneling effect. When the energy converges to a stable threshold, the final spin state configuration of the qubit is recorded, specifically including: This step is based on the quantum annealing optimization model constructed in step 2.2. By simulating the quantum tunneling effect, the parameters are dynamically adjusted until the system energy converges, thus obtaining the optimal spin state configuration. The specific implementation process is as follows: The quantum annealing process is initialized by setting the number of quantum annealing iterations (e.g., 1000 times), the energy convergence threshold (e.g., the energy change in 50 consecutive iterations is less than 10^-6), and the initial temperature (e.g., 100K, to simulate the thermal equilibrium state of the quantum system). The qubit spin state and coupling strength parameters initialized in step 2.2 are then input into the quantum annealing simulator to start the optimization process.

[0045] The core of quantum annealing is the dynamic adjustment of the transverse magnetic field strength, which achieves the transition from a quantum state to a classical state by weakening the transverse magnetic field strength. In the initial stage (the first 300 iterations), the transverse magnetic field strength remains at its maximum value (e.g., 10.0). The strong transverse magnetic field causes the spin state of the qubits to frequently flip, simulating the quantum tunneling effect, which helps to explore the entire solution space and avoid getting trapped in local optima. In the middle stage (301-800 iterations), the transverse magnetic field strength gradually decreases according to an exponential decay law (the decay coefficient is 0.99 / iteration). At this time, the quantum tunneling effect weakens, and the system begins to converge towards a lower energy state. In the later stage (801-1000 iterations), the transverse magnetic field strength drops to its minimum value (e.g., 0.1). The system is basically in a classical state, and the spin state tends to be stable, with only minor adjustments near local optima. Energy convergence monitoring and judgment involves recording the energy change curve based on the current energy value of the Hamiltonian during each iteration. When the energy change is less than a preset convergence threshold after 50 consecutive iterations, the system is considered to have converged to a stable state, and the corresponding energy value is the approximate ground state energy of the Hamiltonian. If convergence is not achieved after reaching the upper limit of iterations, the attenuation coefficient of the transverse magnetic field strength is adjusted appropriately (e.g., changed to 0.98 / iteration), and the annealing process is restarted until energy convergence. When the energy converges to the stable threshold, the quantum annealing process is stopped, and the final spin state configuration of all qubits is recorded (each qubit's spin state is +1 or -1). This configuration corresponds to the optimal or suboptimal solution to the spectrum resource allocation problem.

[0046] Step 2.4: Based on the final spin state configuration of the qubits, decode the spectrum resource allocation scheme corresponding to each UAV; map the spectrum resource allocation scheme into a specific carrier frequency allocation table, time slot offset configuration table, and power control parameter set to form a complete spectrum resource scheduling instruction set, specifically including: Based on the final spin state configuration of the qubits obtained in step 2.3, this step uses decoding mapping to form a specific resource allocation scheme and scheduling instruction set, providing an executable communication scheduling basis for the UAV swarm. The specific implementation process is as follows: Quantum bit spin state decoding: Based on the mapping relationship between qubit spin state and resource allocation decision variables established in step 2.2, the final spin state configuration is decoded: Carrier frequency decoding involves decoding the spin state of the qubits in the corresponding frequency sub-channel by +1 to indicate allocation to that sub-channel, and -1 to indicate non-allocation of that sub-channel. Combined with the normalized subcarrier aggregation region defined in step 2.12, the sub-channels decoded as allocated within the same aggregation region are integrated to form the carrier frequency set for each UAV (e.g., reconnaissance aircraft 1 is allocated 20 sub-channels from 1.2GHz to 1.22GHz to meet the 20MHz bandwidth requirement). Time slot offset decoding involves decoding the spin state of the qubits in the corresponding time slot by +1 to indicate occupation of that time slot, and -1 to indicate non-occupation of that time slot. Based on the normalized time window length obtained in step 2.12, consecutive time slot blocks are allocated to each UAV (e.g., delivery aircraft 1 is allocated time slots 2-3, corresponding to a duration of 200ms, meeting the low latency requirement), and the time slot offset for each UAV is calculated (i.e., the time difference between the first occupied time slot and the system's initial time slot, e.g., an offset of 100ms).

[0047] The power control parameter is decoded, which means decoding the combination of qubit spin states corresponding to the power level into a specific power value (e.g., spin state "+1, +1" is decoded as 20dBm, "+1, -1" is decoded as 15dBm, and "-1, +1" is decoded as 10dBm). Combined with the power constraint term in step 2.14, it is ensured that the decoded power value does not exceed the preset threshold. At the same time, the power value is adjusted according to the spatial isolation matrix (e.g., when the spatial isolation between UAVs is low, the power is appropriately reduced to suppress interference).

[0048] Resource allocation scheme mapping and table generation: The carrier frequency allocation table is generated by classifying UAVs by number (command relay, reconnaissance aircraft 1-3, delivery aircraft 1-2) and recording the carrier frequency range, number of sub-channels, and corresponding RIS array number for each UAV (e.g., reconnaissance aircraft 2 has a carrier frequency of 1.5GHz-1.52GHz, corresponding to RIS array 2). The table clearly indicates the priority of frequency resource usage (reconnaissance aircraft in key search and rescue areas have the highest priority). The time slot offset configuration table is generated by recording the time slot block number, time slot offset, time slot duration, and scheduling cycle for each UAV (e.g., command relay aircraft time slot blocks 0-4 have an offset of 0ms, a duration of 500ms, and a scheduling cycle of 1s), ensuring that the time slots of each UAV do not overlap and avoiding time domain conflicts. The power control parameter set is generated by recording the transmit power value, power adjustment step size (e.g., 5dB / step), maximum power threshold, and power optimization cycle (e.g., 10s / time) for each UAV, supporting dynamic power adjustment based on channel status.

[0049] A complete scheduling instruction set is formed by integrating the three tables mentioned above, classifying them according to UAV type and mission stage, and generating a complete spectrum resource scheduling instruction set. The instruction set includes frequency allocation instructions, time slot configuration instructions, and power control instructions. Each instruction is accompanied by an execution timestamp (based on BeiDou time synchronization, with an accuracy of 1ms) and verification feedback requirements (the UAV must report the execution result to the command relay after execution). Finally, the scheduling instruction set is distributed to each UAV and RIS array through the stable communication link established in step 1, ensuring that all devices execute the spectrum resource allocation scheme synchronously and achieving efficient scheduling of multi-aircraft collaborative communication.

[0050] In a preferred embodiment of the present invention, step 3 involves constructing an overlapping communication frame structure based on the carrier frequency, time slot offset, and power control parameter set; and generating a dynamic frame scheduling scheme based on the overlapping communication frame structure, including: Step 3.1: Based on the carrier frequency allocation table, allocate an independent subcarrier group to each UAV to obtain the subcarrier group allocation result; based on the time slot offset configuration table, set the frame start time offset for each UAV; based on the power control parameter set, configure the corresponding transmit power level for each UAV, specifically including: This step takes the carrier frequency allocation table, time slot offset configuration table, and power control parameter set output from step 2.4 as input to complete the basic resource anchoring for UAV communication, providing subcarrier-time-power three-dimensional parameter support for subsequent frame structure construction. The specific implementation process is as follows: Subcarrier group allocation involves extracting the bandwidth requirements (20MHz / unit for reconnaissance aircraft, 5MHz / unit for delivery aircraft, and 50MHz for command relay aircraft) of each UAV from the carrier frequency allocation table in step 2.4, as well as the carrier frequency range and subchannel flatness index. The division principle is based on frequency domain flatness ≥0.8, subcarrier spacing of 15kHz, and guard bandwidth of 500kHz to ensure stable frequency domain response within each subcarrier group and reduce intra-group interference.

[0051] Differentiated subcarrier group allocation, i.e., command relay aircraft: 333 subcarriers (including guard bandwidth) in the 1.8GHz-1.85GHz frequency band are allocated, divided into control channel subcarrier groups and data aggregation subcarrier groups. The control channel group accounts for 10% (33 subcarriers, used to issue dispatch commands), and the data aggregation group accounts for 90% (300 subcarriers, used to receive data from each UAV); reconnaissance aircraft (3 aircraft), each is allocated a continuous 20MHz frequency band (e.g., reconnaissance aircraft 1 is allocated 1.2GHz-1.22GHz), corresponding to 1333 subcarriers, divided into image data subcarrier groups + thermal imaging data subcarrier groups, in a ratio of 7:3 (image data requires higher bandwidth); delivery aircraft (2 aircraft), each is allocated a 5MHz frequency band (e.g., delivery aircraft 1 is allocated 1.5GHz-1.505GHz), corresponding to 333 subcarriers. A single group can meet the requirements of control commands and position feedback.

[0052] Assign a unique identifier to each subcarrier group (e.g., reconnaissance aircraft 2-image group-002), record the subcarrier number range, center frequency, bandwidth and corresponding RIS array coverage within the group, simulate adjacent channel interference between subcarrier groups using a spectrum analyzer, ensure inter-group isolation ≥30dB, and form the final subcarrier group allocation result table.

[0053] The frame start time offset setting, that is, combined with the time slot block information in the time slot offset configuration table in step 2.4, determines the core parameters of the communication frame. The frame period is set to 10ms (to adapt to the channel coherence time and reduce the impact of channel changes on frame transmission). Each frame contains 10 time slots (each time slot is 1ms). The frame header occupies 0.5ms, and the frame tail reserves 0.2ms as a guard interval to avoid inter-frame crosstalk. The offset differentiation calculation uses the frame start time of the command relay as the system reference time (t=0), and calculates the frame start offset according to the time slot offset requirements of each UAV: ​​For reconnaissance aircraft, due to the need to transmit high-resolution data, the frame start offset is set at 1ms, 3ms and 5ms to stagger the peak (e.g., reconnaissance aircraft 1 offsets by 1ms, reconnaissance aircraft 2 offsets by 3ms), to ensure that their data frames partially overlap in the time domain but do not conflict in the core data area; for delivery aircraft, the frame start offset is set at 2ms and 4ms to stagger the peak with the reconnaissance aircraft, while reserving a 0.3ms redundant offset to cope with time synchronization errors caused by changes in UAV flight attitude; all offsets are calibrated based on the Beidou time service system, and the synchronization accuracy is controlled within 10μs.

[0054] Offset configuration and feedback involves writing the frame start time offset into the communication module register of each UAV. After configuration, the UAV feeds back the synchronization status to the command relay via the control channel. If the synchronization error exceeds 50μs, the offset is readjusted until the requirements are met.

[0055] The transmit power level configuration involves extracting the power threshold (20dBm) and adjustment step size (5dB / step) from the power control parameter set in step 2.4, and combining it with the real-time signal-to-noise ratio (SNR) from the channel state tensor feedback in step 1.4 to establish an "SNR-power level" mapping table: Level 1 (10dBm, low power energy saving) when SNR ≥ 20dB; Level 2 (15dBm, balancing rate and power consumption) when 10dB ≤ SNR < 20dB; and Level 3 (20dBm, high power to ensure communication) when SNR < 10dB.

[0056] The configuration is differentiated by drone type: command relay drones are configured with a fixed power level of 2 (15dBm) to ensure coverage of all drones and stable power consumption; reconnaissance drones are configured to match the level according to the mapping table during normal reconnaissance, and are forcibly upgraded to level 3 when suspected signs of life are detected to ensure data transmission reliability; delivery drones are configured to primarily operate at level 1 or 2, and are downgraded to level 1 when approaching the delivery point (distance from the target ≤100m) to avoid excessive power interference with ground receiving equipment.

[0057] The power level parameters are written through the control interface of the UAV's onboard power amplifier. After configuration, the power calibration process is started, and the relay is instructed to receive the pilot signal of the UAV. The deviation between the actual received power and the theoretical value is measured. If the deviation exceeds 2dB, it is corrected by power adjustment command until the deviation is ≤1dB, thus forming the final transmit power level configuration result.

[0058] Step 3.2: Based on the subcarrier group allocation results and frame start time offset, construct a time-frequency two-dimensional overlapping frame structure; in the time-frequency two-dimensional overlapping frame structure, set up a control channel region, a data channel region, and a pilot channel region; configure power allocation ratios for different regions according to the transmit power level, specifically including: This step takes the subcarrier group allocation result from step 3.1, the frame start time offset, and the transmit power level as input to construct a frame structure that combines overlap characteristics and region division, and completes the power adaptation for each region. The specific implementation process is as follows: The time-frequency two-dimensional overlapping frame structure is constructed by using time and frequency as the two-dimensional coordinate axis. The time axis has a frame period of 10ms and is divided into 10 time slots (numbered 0-9) with 1ms intervals, and the frame start offset of each UAV is marked. The frequency axis is divided according to subcarrier groups, and the center frequency and bandwidth range of each UAV subcarrier group are marked (e.g., the center frequency of the image group of reconnaissance aircraft 2 is 1.3GHz and the bandwidth is 14MHz).

[0059] The overlap feature is implemented by using the difference in the frame start offset of each UAV to achieve time-domain overlap and using the frequency-domain isolation of subcarrier groups to achieve frequency reuse. For example, the frames of reconnaissance aircraft 1 (t=1ms-11ms) and delivery aircraft 1 (t=2ms-12ms) overlap in the time domain during the period t=2ms-11ms, but the subcarrier groups are located in the 1.2GHz and 1.5GHz frequency bands respectively, so there is no frequency domain conflict; the frames of the command relay aircraft (t=0-10ms) overlap with all UAV frames, and the signals are distinguished by dedicated subcarrier groups.

[0060] Frame structure visualization and verification involves generating a two-dimensional time-frequency overlapping frame structure heatmap, with the horizontal axis representing time (0-12ms) and the vertical axis representing frequency (0.8GHz-2.4GHz). Different colors are used to mark the frame coverage area of ​​each UAV. Verification is performed to check whether there is a simultaneous frequency conflict in the overlapping area (multiple UAVs occupying the same time-frequency resource). If such a conflict exists, the process returns to step 3.1 to adjust the offset or subcarrier group until there is no simultaneous frequency conflict.

[0061] The intra-frame channel region is divided according to the principle of control priority, data primary, and pilot auxiliary, with the total proportion satisfying a frame period of 10ms. Specifically, the control channel region occupies 5% (0.5ms) of the frame period, located at the frame header, and is used to transmit key control information such as frame synchronization signals, scheduling instructions, and status feedback; the data channel region occupies 85% (8.5ms) of the frame period, located in the middle of the frame, and is subdivided into sub-regions according to service type (such as the image sub-region and thermal imaging sub-region for reconnaissance aircraft); and the pilot channel region occupies 10% (1ms) of the frame period, distributed in the data channel region at a pilot insertion interval of once every 0.85ms, and is used for real-time channel estimation.

[0062] Subcarrier allocation in each region: Control channel region: 10% of the subcarriers in the subcarrier group are allocated to each UAV (e.g., 133 of the 1333 subcarriers of reconnaissance aircraft 1 are used for the control channel), using QPSK modulation (strong anti-interference capability); Data channel region: 80% of the subcarriers in the subcarrier group are allocated, with 64QAM modulation (high bandwidth) used for reconnaissance aircraft and 16QAM modulation (balancing rate and reliability) used for delivery aircraft; Pilot channel region: 10% of the subcarriers in the subcarrier group are allocated, using a fixed pilot sequence (Zadoff-Chu sequence) to ensure channel estimation accuracy.

[0063] The area boundary identification setting involves inserting a synchronization identifier sequence (such as a 16-bit fixed binary code 1100110011001100) at the beginning and end of each area. This is used by the UAV communication module to quickly identify area boundaries and avoid confusion between data and control signals.

[0064] Regional power allocation based on power level means that the transmit power level in step 3.1 is used as the total power benchmark, and the proportions are divided according to the principle of high priority for control channels, medium priority for pilot channels, and on-demand allocation for data channels. At the same time, the spatial isolation of the channel state tensor is adjusted, and the power ratio of control channels is increased for UAVs with low spatial isolation (which are more susceptible to interference).

[0065] Differentiated power allocation schemes: Level 1 (10dBm): 40% control channel (4dBm), 30% pilot channel (3dBm), and 30% data channel (3dBm), suitable for low-bandwidth, low-interference scenarios such as delivery aircraft; Level 2 (15dBm): 30% control channel (4.5dBm), 20% pilot channel (3dBm), and 50% data channel (7.5dBm), suitable for stable communication scenarios such as command relays; Level 3 (20dBm): 35% control channel (7dBm), 25% pilot channel (5dBm), and 40% data channel (8dBm), suitable for high-interference scenarios such as reconnaissance aircraft conducting key search and rescue operations.

[0066] The received power and signal-to-noise ratio of each region are calculated using a link budget tool to ensure that the control channel SNR is ≥15dB (to ensure reliable command reception), the data channel SNR is ≥10dB (to ensure data transmission rate), and the pilot channel SNR is ≥20dB (to ensure channel estimation accuracy). If these conditions are not met, the ratios are adjusted until they are met.

[0067] Step 3.3: Based on the aforementioned time-frequency two-dimensional overlapping frame structure, analyze the inter-frame interference patterns and channel occupancy to obtain the channel occupancy status and interference patterns; based on the channel occupancy status and interference patterns, determine the frame scheduling conflict detection rules and priority determination criteria, specifically including: Using the time-frequency two-dimensional overlapping frame structure from step 3.2 as input, inter-frame interference and channel occupancy status are analyzed, and conflict detection and priority rules are established to provide a basis for subsequent scheduling decisions. The specific implementation process is as follows: Inter-frame interference pattern and channel occupancy analysis involves using a command relay to collect interference data over 100 consecutive frame periods. Combined with the channel state tensor from step 1.4, the interference is categorized into three types based on its source and nature: co-channel interference (interference between overlapping areas of different UAV frames within the same subcarrier group, such as conflicts caused by subcarrier group configuration errors); adjacent-channel interference (interference between adjacent subcarrier groups, such as crosstalk between the 1.22GHz subcarrier group of reconnaissance aircraft 1 and the 1.23GHz subcarrier group of reconnaissance aircraft 2); and frame overlap interference (inter-frame interference that overlaps in the time domain but is isolated in the frequency domain, such as interference caused by power overflow between overlapping frames of reconnaissance aircraft 1 and delivery aircraft 1).

[0068] Interference pattern feature extraction involves extracting feature parameters for three types of interference: co-channel interference (interference power, conflict duration, and involved UAV pairs); adjacent-channel interference (interference power, frequency offset, and subcarrier spacing); and frame overlap interference (interference power, overlap duration, and power level difference). Interference patterns are then classified using clustering algorithms (such as K-Means) to form typical pattern libraries, including strong co-channel interference (>-70dBm), moderate adjacent-channel interference (-80dBm to -70dBm), and weak frame overlap interference (<-80dBm).

[0069] Channel occupancy quantitative analysis is performed based on a two-dimensional time-frequency overlapping frame structure heatmap. The occupancy rate of each time-frequency resource block (1ms×1MHz) is statistically analyzed and classified into high occupancy (>80%), medium occupancy (50%-80%), and low occupancy (<50%). At the same time, the channel utilization rate of each UAV (actual data transmission volume of the data channel / maximum channel capacity) is calculated to form a channel occupancy statistics table, marking high occupancy periods (e.g., t=3ms-7ms, multiple aircraft frames overlapping) and high utilization UAVs (e.g., reconnaissance aircraft 2, continuously transmitting image data).

[0070] The frame scheduling conflict detection rules are established, which combine interference mode and channel occupancy analysis to define three types of core conflicts: time-frequency conflict, which is that the same time-frequency resource block is occupied by multiple UAVs at the same time (the core cause of co-channel interference); power conflict, which is that the transmission power of a certain area exceeds the channel capacity, resulting in excessive interference to adjacent channel UAVs (the core cause of adjacent channel interference); and synchronization conflict, which is that the frame start time offset error of the UAV exceeds the guard interval, resulting in frame header overlap (the core cause of frame overlap interference).

[0071] Differentiated detection rules are established as follows: Time-frequency conflict detection, which monitors the occupancy status of each time-frequency resource block in real time. If a resource block is occupied by two or more UAVs at the same time, a conflict alarm is triggered. The detection accuracy is 10μs×100kHz. Power conflict detection, which collects the transmission power of each UAV in real time. If the power in a certain area exceeds 10% of the upper limit of the power allocation in that area, or the interference power in the adjacent frequency area exceeds -80dBm, a conflict alarm is triggered. Synchronization conflict detection, which calculates the deviation between the actual frame start time of the UAV and the preset offset by the received time difference of the frame header synchronization identifier sequence. If the deviation exceeds 0.2ms (the maximum value of the guard interval), a conflict alarm is triggered.

[0072] The conflict severity level is classified into three levels based on the degree of impact on communication: Level 1 (urgent), i.e., frequency conflict involving key data transmission of reconnaissance aircraft; Level 2 (important), i.e., power conflict or synchronization conflict affecting data transmission rate; and Level 3 (general), i.e., weak interference due to frame overlap, not affecting core services. Different levels correspond to different processing priorities.

[0073] Priority criteria were established, namely, combining the characteristics of UAV swarm rescue missions, establishing a three-dimensional priority system based on UAV type, operational urgency, and channel quality, with weights of 0.5, 0.3, and 0.2 for each dimension (type priority being the core). The priorities for each dimension are further subdivided: UAV type priority (weight 0.5): Command relay aircraft (Level 1, highest) > 3D modeling reconnaissance aircraft (Level 2) > Material delivery aircraft (Level 3, lowest); Operational urgency priority (weight 0.3): Life-threatening heat source detection data transmission (Level 1) > Landslide area image transmission (Level 2) > Routine reconnaissance data transmission (Level 3) > Delivery aircraft status feedback (Level 4) > Non-urgent command transmission (Level 5); Channel quality priority (weight 0.2): Excellent channel quality (SNR≥20dB, Level 1) > Good channel quality (10dB≤SNR<20dB, Level 2) > Poor channel quality (SNR<10dB, Level 3). UAVs with poor channel quality are given higher priority to avoid communication interruptions. The weighted summation method is used to calculate the real-time priority score of each UAV (score = type score × 0.5 + service score × 0.3 + channel score × 0.2, with a score difference of 2 points per level). The lower the priority score, the higher the priority. Therefore, the reconnaissance aircraft has a higher priority than the delivery aircraft.

[0074] Step 3.4: Construct a frame scheduling decision tree based on the frame scheduling conflict detection rules and priority determination criteria; based on the frame scheduling decision tree and combined with real-time channel state feedback information, generate an adaptive dynamic frame scheduling scheme. The dynamic frame scheduling scheme includes a frame scheduling timing table, power adjustment instructions, and conflict avoidance strategies, specifically including: Using the conflict detection rules (including conflict types and levels) and priority determination criteria (three-dimensional weighted system) from step 3.3 as input, a hierarchical decision tree model is constructed. The key lies in ensuring the accuracy and uniqueness of the decision output through the design of multi-dimensional decision nodes and differentiated scheduling strategies. Specifically, this includes: A four-level progressive decision node design is adopted, with each node being directly related to the technical output of the preceding step, ensuring the coherence of the decision logic and the effectiveness of constraints. The specific levels are as follows: The first level (root node), namely the conflict existence determination node, takes the real-time frame scheduling conflict monitoring results as input and outputs two branches: conflict present / no conflict. This node directly calls the conflict detection rules in step 3.3, using the time-frequency resource block occupancy status, power threshold, and synchronization deviation as monitoring indicators to achieve rapid initial judgment of conflicts.

[0075] The second level, namely the conflict level determination node, only applies to branches with conflicts. It takes the interference power of the conflict or the type of business involved as input and outputs three branches: Level 1 (urgent), Level 2 (important), and Level 3 (general). The level division directly follows the conflict level standard in step 3.3. Among them, Level 1 conflict specifically refers to the time-frequency conflict involving the transmission of life heat source detection data, which is the highest priority scenario.

[0076] The third level is the priority score determination node. The input is the real-time priority score of the UAV involved in the conflict (calculated according to the three-dimensional weighted method in step 3.3, with a score range of 1.0-5.0, and the lower the score, the higher the priority). The output has two branches: score ≤ 2.0 (high priority) / score > 2.0 (low priority). This node is the core embodiment of the technical goal of business priority protection, ensuring that the communication resources of high-priority UAVs (such as reconnaissance aircraft) are satisfied first.

[0077] The fourth level is the channel quality assessment node. It takes the real-time signal-to-noise ratio (SNR) of high-priority UAVs as input and outputs three branches: SNR≥20dB (excellent), 10dB≤SNR<20dB (good), and SNR<10dB (poor). This node provides a basis for channel adaptation and avoids communication interruptions caused by fixed scheduling in poor channel scenarios.

[0078] The scheduling strategy for the leaf nodes of the decision tree is defined. Each decision path's terminal (leaf node) corresponds to a unique scheduling strategy to ensure the executability of the decision output. The strategies are as follows (covering typical scenarios): Path 1 (conflict exists, Level 1, score ≤ 2.0, arbitrary channel quality): High-priority UAVs (such as reconnaissance aircraft transmitting life detection data) maintain the current frame parameters (subcarrier group, frame offset), while low-priority UAVs (such as delivery aircraft) immediately execute a combined strategy of frame offset, 1ms adjustment and power level downgrade by 1 level to ensure that high-priority services are uninterrupted.

[0079] Path 2 (conflict exists, Level 2, score > 2.0, SNR < 10dB): All conflicting UAVs simultaneously execute the strategy of shifting the subcarrier group to the protection bandwidth by 1MHz and extending the frame protection interval to 0.3ms, which avoids conflicts and improves transmission reliability in harsh channel scenarios.

[0080] Path 3 (no conflict, no grade, no score, and arbitrary channel quality): Maintain the current frame parameters and dynamically adjust the data channel modulation scheme only according to the channel quality (good, adjust to 64QAM; good, adjust to 16QAM; poor, adjust to QPSK) to achieve a balance between not wasting resources and stable communication.

[0081] To ensure the practicality of the decision-making model, a historical conflict dataset and cross-validation method were used to optimize the decision tree: 100 sets of measured conflict data from the earthquake zone (including different terrains and business types) were selected as the training set, and 20 sets of data were selected as the test set. The objective functions were a conflict resolution success rate of ≥95% and a scheduling delay of ≤1ms. The decision logic was optimized by adjusting the node judgment threshold (such as the priority score threshold). The final decision tree model has a unique mapping relationship between input and output.

[0082] After each UAV completes one frame of data transmission (i.e., every 10ms, synchronized with the frame period), it uploads 3D feedback data to the command relay via the control channel. The dimension selection corresponds one-to-one with the decision tree node requirements to avoid redundancy. Scheduling-related dimensions: frame transmission status (success / failure / retransmission count), current service type (e.g., life detection-reconnaissance aircraft 1); channel quality dimensions: real-time SNR, bit error rate (BER), channel impulse response; device status dimensions: UAV remaining battery power, flight attitude (affecting antenna beam pointing, indirectly related to channel quality).

[0083] The command relay performs two steps on the feedback data: cleaning, removing outlier data with BER > 10%, and calibrating SNR through the channel state tensor in step 1.4 (correcting measurement bias caused by terrain occlusion); standardization: converting multi-dimensional data into structured data of UAV ID - real-time priority score - channel quality level - service status, directly adapting to the input interface of the decision tree, and ensuring that the latency of data transmission to the decision tree is ≤ 0.5ms.

[0084] Based on the output of the decision tree and standardized feedback data, a complete scheme is generated, including a frame scheduling timing table, power adjustment commands, and conflict avoidance strategies. The technical characteristics of each component are clearly defined and interrelated. Specifically: Using a timeline and UAV ID as a two-dimensional index, the core parameters of each UAV are clearly defined within the next 100 frame cycles (1 second, balancing real-time performance and planning). The key technical fields in the table include: Basic parameters: frame start time (accurate to 10μs), frame end time, subcarrier group number (related to the allocation result in step 3.1).

[0085] Priority Identification: High-priority services are marked for emergency protection, and 5% of spare time and frequency resources (such as 1.9GHz-1.905GHz subcarrier groups) are reserved.

[0086] Modulation method: The modulation type of the data channel is marked according to the channel quality (e.g., "Reconnaissance Aircraft 2-64QAM").

[0087] Power adjustment commands (energy consumption and communication balance): generated based on decision tree strategy and equipment status data. The command format adopts a dual mode of machine-readable code and human-readable instructions. Core technical features include: Command parameters: target power level (levels 1-3, associated with step 3.1), power ratio of each channel region (control / data / pilot, associated with step 3.2), execution timestamp (accurate to the start of the frame). If the drone's remaining battery power is ≤20%, the command will forcibly impose a power level cap constraint of Level 2; if SNR <10dB, the command will indicate an emergency permission allowing temporary power exceeding the threshold by 10% (≤22dBm). For potential conflicts identified by the decision tree (discovered through time-series table simulations, such as frame overlap between reconnaissance aircraft 3 and delivery aircraft 2 at t=50ms), a layered avoidance strategy will be formulated, with the following technical priorities: Level 1 Avoidance (Priority): Time-domain avoidance, which adjusts the frame start offset of low-priority UAVs (e.g., increasing the delivery aircraft 2 from 4ms to 6ms), without wasting frequency domain resources; Level 2 Avoidance (Backup): Frequency-domain avoidance, which shifts the subcarrier groups of high-priority UAVs towards the protection bandwidth (e.g., shifting the reconnaissance aircraft 3 by 1MHz) to avoid interference; Level 3 Avoidance (Emergency): Power avoidance, which reduces the power level of low-priority UAVs by 1 level, and is only activated when time-domain and frequency-domain avoidance are not feasible.

[0088] Construct a communication scenario for the earthquake zone (including terrain masking and multipath fading models), and verify the core indicators of the input scheduling scheme: conflict resolution success rate ≥ 95%; service transmission rate (reconnaissance aircraft ≥ 100 Mbps, delivery aircraft ≥ 10 Mbps); scheduling delay ≤ 1 ms; if the indicators do not meet the standards, return to the decision tree to adjust the node strategy and regenerate the scheme to ensure that the output scheme is practical. The verified scheme is transmitted through the encrypted control channel (using AES-128 encryption) of the command relay. Each UAV completes parameter configuration within one frame period (10ms) after receiving the signal. To avoid synchronization errors, a dual synchronization mechanism of BeiDou time synchronization and frame header synchronization is adopted to ensure that the frame parameter adjustments of all UAVs take effect at the same time (time deviation ≤10μs). After the UAV executes the scheme, it immediately returns the execution result code (e.g., 00 - success, 01 - power over-limit) through the feedback channel. The command relay has a built-in emergency dispatch module. If it receives a failure code such as 01, it immediately triggers the backup strategy (e.g., activates the reserved backup subcarrier group) and generates a new adjustment command within 0.5ms to achieve a rapid closed loop of fault-repair and ensure the continuity of communication for earthquake relief.

[0089] In a preferred embodiment of the present invention, step 4, based on the dynamic frame scheduling scheme, extracts a channel reciprocity feature sequence; generates a quantum random number based on the channel reciprocity feature sequence; and generates a session key and device permission topology map based on the quantum random number, including: Step 4.1: Based on the frame scheduling timing table in the dynamic frame scheduling scheme, configure a bidirectional channel sounding time slot; within the bidirectional channel sounding time slot, control the UAV to perform reciprocal channel measurements between the pairs to acquire uplink channel response data and downlink channel response data, specifically including: Using the frame scheduling timing table from step 3.4 as input, uplink / downlink channel data supporting reciprocity analysis is obtained through precise time slot configuration and bidirectional channel synchronization measurement. The specific implementation process is as follows: Based on the frame start offset, time slot occupancy range, and service priority of each UAV in the frame scheduling time sequence table, the detection time slot is configured. That is, the time slot segment of core business (such as the data transmission of life detection of reconnaissance aircraft) is prohibited from being inserted into the detection time slot. The frame protection interval (0.2ms), low traffic time slot (such as the gap of delivery aircraft status feedback) or reserved time slot (the safe configuration time slot marked in the time sequence table) are selected first to ensure that the detection and communication services do not conflict with each other.

[0090] The core parameters of the detection time slots are uniformly set as follows: single time slot duration is 0.2ms (matching the channel coherence time to avoid measurement failure due to channel abrupt changes), time slot interval is 10ms (consistent with the frame period to ensure periodic monitoring), and each detection period contains 2 consecutive sub-time slots (sub-time slot 1: uplink detection, sub-time slot 2: downlink detection, with an interval of 0.05ms for transmit / receive switching); in the frequency domain, 5% of the subcarriers in each UAV subcarrier group are allocated to the detection time slots (e.g., 67 of the 1333 subcarriers of reconnaissance aircraft 1 are used for detection), and orthogonal frequency division multiplexing (OFDM) mode is adopted, with a subcarrier interval of 15kHz to avoid frequency domain interference with the data / control channel.

[0091] Based on the frame scheduling timing table, a detection time slot configuration table is generated for each UAV, marking parameters such as the detection time slot start time (accurate to 10μs), subcarrier number, and transmit / receive switching time. For example, the detection time slot of reconnaissance aircraft 1 is t=1.8ms-2.0ms (sub-time slot 1: 1.8ms-1.9ms, sub-time slot 2: 1.95ms-2.05ms), and the subcarrier number is 1-67. The configuration table is sent to each UAV through an encrypted control channel. After receiving it, the UAV writes the parameters into the communication module register to complete the time slot configuration initialization.

[0092] Based on the star topology of the UAV swarm (with the command relay as the core), the measurement link types are determined as follows: point-to-point links between the command relay and each UAV (6 in total, including 1 command aircraft, 3 reconnaissance aircraft, and 2 delivery aircraft); links between UAVs of the same type (such as reconnaissance aircraft 1 and reconnaissance aircraft 2, a total of 3 links) to ensure coverage of all communication scenarios; during measurement, the measurement is performed in pairs of UAVs (such as command aircraft-reconnaissance aircraft 1 pair) according to priority (core links take precedence over collaborative links).

[0093] A unified triggering mechanism is adopted to achieve measurement synchronization. 10μs before the start of the detection time slot, the command relay sends a synchronization trigger signal (using a 16-bit binary synchronization code 1010101011001100) to the UAVs participating in the measurement. After receiving the synchronization signal, the UAVs send pilot signals to the command in sub-time slot 1 (uplink) and to the command in sub-time slot 2 (downlink). The pilot sequence adopts the Zadoff-Chu sequence (length 256) with good orthogonality to ensure that there is no mutual interference between uplink and downlink pilots.

[0094] Uplink and downlink channel response data acquisition consists of two parts: uplink data acquisition, where the command unit receives the pilot signal sent by the UAV in sub-time slot 1, calculates the uplink channel response data based on the least squares algorithm, including parameters such as channel amplitude, phase, and delay spread, and simultaneously records the acquisition timestamp and UAV position information; and downlink data acquisition, where the UAV receives the pilot signal sent by the command unit in sub-time slot 2, calculates the downlink channel response data using the least squares algorithm, and similarly records the timestamp and position information. The acquired data is stored in a temporary buffer and uploaded to the command unit once every 10ms detection cycle.

[0095] After receiving uplink and downlink data, the command aircraft first checks the consistency of timestamps (the deviation between uplink and downlink data timestamps must be ≤5μs) and the stability of the position (the change in the position of the UAV during the measurement period must be ≤0.5m). If these conditions are not met, the data is marked as invalid, triggering a retest in the next detection cycle. Valid data is associated with the corresponding UAV pair identifier and detection time slot information to form a structured dataset of UAV pair-uplink and downlink channel response-spatiotemporal parameters.

[0096] Step 4.2: Perform reciprocity verification processing on the uplink channel response data and downlink channel response data, and calculate the channel reciprocity error index; when the channel reciprocity error index is less than a preset threshold, extract the channel amplitude fluctuation sequence and phase jitter sequence, and merge them to generate a channel reciprocity feature sequence, specifically including: First, the uplink and downlink channel response data are preprocessed to ensure spatiotemporal consistency between the two types of data. Based on the timestamps and location information recorded in step 4.1, the uplink and downlink data of the same UAV pair within the same detection time slot are precisely aligned. Abnormal data with timestamp deviations exceeding 5 μs or UAV position changes greater than 0.5 m during the measurement period are removed to avoid distortion of reciprocity judgment caused by equipment movement or synchronization errors. For the aligned valid data, the core response parameters of the uplink and downlink channels are extracted, including channel amplitude, phase, delay spread, and frequency fading coefficient, forming paired parameter sets.

[0097] Subsequently, channel reciprocity verification is performed to verify the uplink and downlink symmetry of the wireless channel within a short time interval (10ms probe time slot). This is specifically achieved by calculating two types of reciprocity error indices: amplitude reciprocity error, which is the ratio of the absolute difference between the uplink and downlink channel amplitude parameters to their average value; and phase reciprocity error, which is the ratio of the difference between the uplink and downlink channel phase parameters (modulo 2π) to π. The two error indices are weighted and summed with a weight ratio of 0.6:0.4 to obtain a comprehensive channel reciprocity error index. This index directly reflects the degree of symmetry between the uplink and downlink channels; a smaller value indicates better reciprocity.

[0098] The determination of the preset reciprocity error threshold needs to take into account the complex channel environment characteristics of the seismic zone. The threshold is set to 0.15 (i.e., 15%). This threshold can tolerate slight reciprocity deviations caused by multipath effects in mountainous areas, while effectively eliminating severe non-reciprocal data caused by terrain obstruction and sudden interference. When the calculated channel reciprocity error index is less than 0.15, the current channel is determined to meet the reciprocity requirements, and then valid feature sequences are extracted; if the error index exceeds the threshold, the data set is marked as invalid, triggering the retesting process for the next detection cycle.

[0099] In the feature sequence extraction stage, for uplink and downlink channel response data that meet the reciprocity requirement, channel amplitude fluctuation sequences and phase jitter sequences are extracted in chronological order (unit: probe time slot). The amplitude fluctuation sequence is calculated using the channel amplitude parameter difference within consecutive probe periods, reflecting the random variation characteristics of the channel amplitude; the phase jitter sequence is calculated using the channel phase parameter difference (modulo 2π) within consecutive probe periods, reflecting the random jitter characteristics of the channel phase. The two sequences are merged point-by-point according to timestamp order to form a channel reciprocity feature sequence that combines random amplitude and phase characteristics. This sequence is generated entirely based on the physical layer characteristics of the wireless channel, possessing inherent randomness and unpredictability.

[0100] Step 4.3: Generate an original random bit stream based on the channel reciprocity characteristic sequence; post-process the original random bit stream to generate quantum random numbers that conform to the quantum randomness standard, specifically including: First, a raw random bitstream is generated by converting the channel reciprocity characteristic sequence in the analog domain into a binary bitstream in the digital domain. For the amplitude fluctuation sequence, an adaptive hierarchical quantization strategy is adopted. Based on the statistical distribution characteristics of the sequence (determined based on the sequence data of the first 100 detection periods), the amplitude fluctuation values ​​are divided into 2^n equally probable intervals (n≥1, n=1 in this embodiment, i.e., binary quantization). Each interval corresponds to 1 binary bit (0 or 1). For example, fluctuation values ​​less than the mean are mapped to 0, and fluctuation values ​​greater than or equal to the mean are mapped to 1. For the phase jitter sequence, a similar binary quantization method is used, dividing the phase jitter values ​​(0~2π) into two equally probable intervals, mapped to 0 and 1 respectively. The binary numbers of the two quantized sequences are alternately concatenated in timestamp order to form the raw random bitstream. During the concatenation process, interleaving is used to avoid the influence of correlation of a single sequence, ensuring the initial randomness of the bitstream.

[0101] The original random bitstream then undergoes multi-stage post-processing to eliminate potential statistical biases and correlations, ensuring that the final random numbers meet quantum randomness standards (such as the NISTSP800-22 random number test standard). The first stage performs debiasing using an XOR equalization method, grouping the original bitstream into consecutive two-bit segments. If two groups of bits are identical (00 or 11), they are discarded; if they are different (01 or 10), the first bit is retained as a valid bit. This process eliminates potential 0 / 1 imbalance biases in the bitstream (such as quantization bias caused by slow channel fading). The second stage performs correlation elimination using a sliding window deduplication method. A sliding window of length 8 is set, and the bit sequences within the window are checked for duplicates. If four or more consecutive identical bits are found, the duplicates are removed, and the non-repeating bits are retained, eliminating potential short-term correlations in the sequence. The third stage performs compression using Huffman coding to compress the debiased and decorrelated bitstream, preserving core random characteristics and removing redundant information to ensure the compactness of the random numbers.

[0102] Finally, randomness verification is performed. The post-processed bitstream is input into a preset randomness verification system, which includes several core verification indicators such as uniformity verification, independence verification, and run-length verification (covering key verification items in the NISTSP800-22 standard). If the bitstream passes all verification indicators (pass rate ≥ 99% for each test), it is determined to meet the quantum randomness standard and is used as the final quantum random number output. If it fails the verification, it returns to the debiasing stage to re-optimize parameters (such as adjusting the sliding window length) until a quantum random number that meets the requirements is generated. This generation process is entirely based on the physical layer reciprocity characteristics of the wireless channel, without relying on external random sources or complex algorithms, possessing inherent security characteristics. Moreover, the generation rate is synchronized with the detection period (10ms / set of random numbers), which can meet the security requirements of real-time communication in UAV swarms.

[0103] Step 4.4: Based on the quantum random number, generate a session key; according to the topology and task roles of the UAV cluster, and in conjunction with the session key, generate a device permission topology diagram. This device permission topology diagram defines the communication permission levels and data access ranges between devices, specifically including: Based on the quantum random numbers generated in step 4.3, a lightweight and highly secure key system and dynamic permission topology are constructed to support secure collaborative communication of UAV swarms. The specific implementation process is as follows: The session key generation phase uses quantum random numbers as the core, and key parameters are designed in conjunction with the communication security requirements of the drone swarm. Based on the transmission security requirements of emergency rescue scenarios, the session key length is set to 256 bits (meeting the AES-256 encryption standard). A continuous 256 bits of quantum random numbers are used as the basic key material. If the quantum random number length is insufficient, multiple quantum random numbers are generated and concatenated to ensure the key length meets the standard. To enhance the key's dynamism and anti-cracking capabilities, a one-cycle key update mechanism is adopted. The key validity period is synchronized with the scheduling cycle of the dynamic frame scheduling scheme (1 second in this embodiment). After each scheduling cycle, the session key is regenerated based on new quantum random numbers, avoiding the security risks caused by using the same key for an extended period. After the key is generated, it is distributed through the stable communication link (beamforming link based on RIS array) established in step 1. During the distribution process, quantum random number encryption is used for transmission (i.e., the session key of the current period is encrypted using the session key of the previous period) to ensure that the key is not eavesdropped or tampered with during transmission. After each UAV receives the key, it verifies the integrity of the key through the key verification code (generated based on the first 32 bits of quantum random number). After the verification is successful, it is stored in the secure cache area for subsequent data encryption and decryption.

[0104] During the device permission topology generation phase, a hierarchical and dynamic permission system needs to be constructed by combining the topology structure, task roles, and session key associations of the UAV cluster. First, the core topology of the UAV cluster is defined (with the command relay as the central node, 3 reconnaissance aircraft and 2 delivery aircraft as edge nodes, and the RIS array as auxiliary communication nodes). Based on this topology, communication permission levels are divided into three levels: Level 1 (core permission) is the point-to-point communication permission between the command relay and each UAV, supporting bidirectional transmission of all types of data (reconnaissance data, control commands, status information, etc.); Level 2 (cooperative permission) is the communication permission between UAVs of the same type (such as between reconnaissance aircraft), supporting only the transmission of data related to cooperative operations (such as location synchronization information, area coverage complementarity information); Level 3 (basic permission) is the communication permission between the delivery aircraft and the reconnaissance aircraft, supporting only unidirectional transmission of necessary information (such as the reconnaissance aircraft transmitting target location information to the delivery aircraft).

[0105] Each access level is assigned a unique access identifier based on session keys. Level 1 access is associated with a global session key (shared by all devices), Level 2 access is associated with a session key specific to the same type of device (e.g., a key specific to a reconnaissance aircraft), and Level 3 access is associated with a one-way access key (e.g., a key from a reconnaissance aircraft to a delivery aircraft), ensuring that data access at different access levels is strictly controlled. Simultaneously, the data access scope of each device is clearly defined in the topology diagram. For example, the command relay can access the transmission data of all devices, the reconnaissance aircraft can only access its own and collaborating UAVs' data, and the delivery aircraft can only access the target data of the command relay and the corresponding reconnaissance aircraft, preventing data leakage.

[0106] The device permission topology map is dynamically updated. When the topology of the drone swarm changes (e.g., a reconnaissance drone leaves the swarm due to insufficient endurance, or a new rescue drone joins) or the mission phase changes (e.g., from wide-area reconnaissance to focused search and rescue), the session key and permission identifier are regenerated based on new quantum random numbers, and the permission levels and access scopes in the topology map are updated synchronously. The updated topology map is distributed to all devices via a command relay, and each device adjusts its communication permission configuration according to the topology map, ensuring that the swarm can maintain a secure and orderly communication order even in dynamically changing scenarios. This topology map achieves both fine-grained control of communication permissions and ensures the security of the permission system through the dynamic association of quantum random numbers.

[0107] In a preferred embodiment of the present invention, based on the device permission topology map and the channel state tensor, communication capability matching is performed on devices in the UAV swarm to generate a set of matched device pairs and their communication parameter mapping relationship, including: Step 4.5: Extract the communication permission level, data access range, and task role identifier of each UAV device from the device permission topology diagram; extract the channel quality indicators, spatial correlation coefficient, and delay spread parameters between each UAV pair from the channel state tensor. Specifically, this includes: extracting device communication control parameters from the device permission topology diagram. The extraction of communication permission levels requires clearly defining the permission hierarchy of each device (Level 1 core permission, Level 2 collaborative permission, Level 3 basic permission). For example, the command relay aircraft has Level 1 permission with all UAVs, reconnaissance aircraft have Level 2 permission, and reconnaissance aircraft have Level 3 permission with delivery aircraft. The system records the communication direction (one-way / two-way) corresponding to the permissions. Data access scope extraction must clearly define the data types that each device is allowed to transmit. For example, the command relay can access all reconnaissance data, control commands, and status information; the reconnaissance aircraft can only access the location synchronization data of the cooperating UAVs and its own reconnaissance data; and the delivery aircraft can only receive delivery commands from the command relay and target location data from the reconnaissance aircraft. Task role identification extraction must label the core functions of each device. For example, the command relay—data aggregation and command issuance; reconnaissance aircraft 1—3D modeling and life detection; and delivery aircraft 2—precise delivery of first-aid kits, ensuring subsequent matching aligns with mission requirements.

[0108] Subsequently, channel adaptability parameters are extracted from the channel state tensor. Channel quality index extraction requires the use of the time-frequency-space three-dimensional features of the tensor to screen core parameters reflecting link stability, including real-time signal-to-noise ratio (SNR), bit error rate (BER), signal strength, and frequency fading coefficient. Among these, SNR and BER directly determine communication reliability, while the frequency fading coefficient reflects the degree of multipath effect. Spatial correlation coefficient extraction requires the use of the spatial dimension components of the tensor to calculate the channel spatial correlation between any two UAVs. A higher coefficient indicates a higher similarity in the channel environment between the two UAVs, making them more suitable for collaborative communication or spectrum reuse. Delay spread parameter extraction requires the use of the time dimension components of the tensor to obtain the signal transmission delay and delay fluctuation range between each UAV pair. This parameter directly affects the matching priority of services with high real-time requirements (such as life detection data transmission).

[0109] All extracted parameters need to be processed in a structured manner, using the device pairs as the basic index (such as command relay aircraft-reconnaissance aircraft 1, reconnaissance aircraft 1-reconnaissance aircraft 2, etc.), to construct a multi-dimensional parameter set of device pair-communication authority-task role-channel quality-spatial correlation-delay spread, to ensure that the control requirements and channel characteristics of each device pair correspond one-to-one.

[0110] Step 4.6: Construct a device communication demand matrix based on communication permission levels and task role identifiers; construct a channel matching degree evaluation matrix based on channel quality indicators, spatial correlation coefficients, and delay spread parameters; and weight and fuse the device communication demand matrix and the channel matching degree evaluation matrix to generate a comprehensive matching score matrix. Specifically, this includes: first, constructing the device communication demand matrix, the core of which is to convert the device's permission level and task role into a quantified communication demand intensity. The rows and columns of the matrix represent devices in the UAV swarm (including command relays), and the elements in the matrix represent the communication demand score (range 0-10) for the corresponding device pair. The scoring rules combine two dimensions: permission level and task role. For equipment pairs with Level 1 permission (command center and various devices) and involved in core tasks (e.g., command center issuing dispatch instructions to reconnaissance aircraft, reconnaissance aircraft uploading life detection data to command center), the required score is 8-10 points. For equipment pairs with Level 2 permission (collaboration between similar types of devices, such as between reconnaissance aircraft) and involved in collaborative tasks (e.g., location synchronization, complementary coverage areas), the required score is 5-7 points. For equipment pairs with Level 3 permission (one-way communication between different types of devices, such as reconnaissance aircraft transmitting target location to delivery aircraft) and involved in basic tasks, the required score is 3-4 points. Equipment pairs with no communication permission or no task association (e.g., between delivery aircraft) receive a required score of 0 points. After the matrix is ​​constructed, it needs to be dynamically adjusted based on the current task stage (e.g., increasing the required score for reconnaissance aircraft and command center during the key search and rescue stage) to ensure that the quantified requirements align with the actual task scenario.

[0111] Secondly, a channel matching evaluation matrix is ​​constructed. The core of this matrix is ​​to convert parameters such as channel quality, spatial correlation, and delay spread into quantified channel adaptation scores (ranging from 0 to 10). The rows and columns of the matrix are consistent with the device communication requirement matrix, and the elements are the channel adaptation scores for the corresponding device pairs. The scoring calculation uses a multi-parameter weighted summation: channel quality accounts for 0.5, with signal-to-noise ratio (SNR) and bit error rate (BER) dominating the score; higher SNR and lower BER result in a higher score. Spatial correlation coefficient accounts for 0.3; a higher coefficient indicates better channel environment coordination between the device pairs, leading to a higher adaptation score. Delay spread parameter accounts for 0.2; smaller delay and narrower fluctuation range result in a higher score. Especially for services with high real-time requirements, the weight of the delay parameter can be temporarily increased to 0.3. For example, a command aircraft and a reconnaissance aircraft have a high SNR, strong spatial correlation, and low delay, resulting in a channel adaptation score of 9; a reconnaissance aircraft and a delivery aircraft have severe channel attenuation and large delay fluctuations, resulting in a channel adaptation score of 3.

[0112] Finally, a weighted fusion of the two matrices is performed to generate a comprehensive matching score matrix. The fusion weights must balance task priority and communication reliability. The weight of the device communication demand matrix is ​​set to 0.4 (ensuring priority matching for core tasks), and the weight of the channel matching evaluation matrix is ​​set to 0.6 (ensuring matching results adapt to the channel environment and improve communication stability). During fusion calculation, the scores of corresponding device pairs in the two matrices are multiplied by their respective weights and then summed to obtain the comprehensive matching score for that device pair (range 0-10). For example, if a device pair has a demand score of 9 and a channel adaptation score of 8, its comprehensive matching score is 9 × 0.4 + 8 × 0.6 = 8.4. After fusion, the comprehensive matching score matrix is ​​normalized to ensure that the scores of all device pairs are distributed within the 0-10 range, facilitating subsequent pairing and selection. Step 4.7: Based on the comprehensive matching score matrix, pair up the drone devices to obtain a set of successfully matched device pairs, specifically including: All device pairs in the comprehensive matching scoring matrix are sorted from high to low according to their comprehensive scores, with higher-scoring device pairs having higher pairing priority. For device pairs with the same score, priority is given to those involving first-level authority (command and control equipment) or core tasks (life detection, precision delivery). If there are still ties, priority is given to those device pairs with higher channel adaptation scores to ensure that core services and high-quality channel resources are matched first.

[0113] Then, a pairwise pairing screening is performed, starting with the highest priority device pair, and the feasibility of pairing is confirmed in sequence: if neither device in the pair is occupied by other pairings, and the overall score of the pair is not lower than the preset pairing threshold (the threshold is set to 4 points in combination with the seismic zone channel environment, and the communication reliability of device pairs with a score lower than this value is difficult to guarantee), then the pair is determined to be a successfully matched pair; if one of the devices is occupied, the pair is skipped and the screening continues downwards; if the overall score of a pair is higher than the threshold but there is a device occupation conflict, and the pair involves core tasks (such as communication between command aircraft and reconnaissance aircraft), the overall score of the paired devices is re-evaluated. If the score of the paired devices is lower than that of the current pair, the original pairing can be terminated, and the matching of core task device pairs can be prioritized.

[0114] The command relay aircraft must be paired with each UAV individually (mandatory pairing at Level 1 clearance) to ensure comprehensive command issuance and data aggregation; reconnaissance aircraft must complete at least one collaborative pairing to ensure complementary area coverage; delivery aircraft must be paired with the command aircraft and the corresponding reconnaissance aircraft in the target area to ensure accurate reception of delivery commands. After pairing, a set of successfully matched equipment pairs is formed.

[0115] Step 4.8: Map communication parameters for the successfully matched device pairs, assigning each matched device pair a corresponding modulation and coding scheme, retransmission mechanism parameters, and flow control threshold. Bind the modulation and coding scheme, retransmission mechanism parameters, and flow control threshold to the matched device pairs to generate a communication parameter mapping table containing device identifiers, matching weights, and communication parameters. Specifically, this includes: differentiated allocation of communication parameters, with core parameters including the modulation and coding scheme, retransmission mechanism parameters, and flow control threshold. The allocation rules are closely related to the device pair's task type, communication permissions, and channel quality. For equipment pairs (such as command aircraft-reconnaissance aircraft) with high overall scores (≥7 points), excellent channel quality (signal-to-noise ratio ≥20dB), and transmitting high-bandwidth data (such as high-resolution images and thermal imaging data), high-order modulation and coding schemes (64QAM and Turbo coding) are adopted to improve transmission rate. For equipment pairs (such as reconnaissance aircraft-delivery aircraft) with medium overall scores (4-6 points), average channel quality (10dB≤SNR<20dB), and transmitting medium-bandwidth data (such as coordinated location information and delivery commands), mid-order modulation and coding schemes (16QAM and convolutional coding) are adopted to balance rate and reliability. For equipment pairs with poor channel quality (SNR<10dB) but need to ensure the transmission of core commands, low-order modulation and coding schemes (QPSK and Hamming coding) are adopted to prioritize communication reliability.

[0116] For equipment pairs with high real-time requirements (such as command aircraft issuing delivery instructions to delivery aircraft), a stop-and-wait retransmission mechanism is adopted, with a retransmission timeout of 50ms and a maximum number of retransmissions of 3, to avoid excessive retransmissions leading to latency accumulation. For equipment pairs with non-real-time requirements but high data integrity requirements (such as reconnaissance aircraft uploading historical reconnaissance data to command aircraft), a sliding window retransmission mechanism is adopted, with a window size of 8, a retransmission timeout of 100ms, and a maximum number of retransmissions of 5, to improve data transmission integrity. For equipment pairs with stable channel quality, the maximum number of retransmissions can be appropriately reduced to decrease resource consumption.

[0117] For equipment pairs between command aircraft and reconnaissance aircraft, the uplink flow control threshold is set to 100Mbps (to accommodate high-resolution image transmission), and the downlink flow control threshold is set to 20Mbps (to accommodate dispatch command transmission). For equipment pairs between reconnaissance aircraft, the flow control threshold is set to 10Mbps (to accommodate lightweight data such as position synchronization). For equipment pairs between command aircraft and delivery aircraft, the downlink flow control threshold is set to 5Mbps (to accommodate delivery commands), and the uplink flow control threshold is set to 2Mbps (to accommodate delivery status feedback).

[0118] After parameter allocation is complete, the parameters are bound to device pairs, generating a communication parameter mapping table. This table contains core fields: device pair identifier, matching weight (i.e., comprehensive matching score), modulation and coding scheme type, retransmission mechanism parameters (retransmission method, timeout, maximum number of retransmissions), flow control threshold (uplink / downlink), and parameter effective time (synchronized with the dynamic frame scheduling scheme). Once the mapping table is generated, it is distributed to all matched device pairs via the secure channel of the command relay. Each device adjusts its own communication module parameters according to the configuration in the table, ensuring parameter consistency and coordination between paired devices.

[0119] In a preferred embodiment of the present invention, step 5, based on the session key, the device permission topology map, and the set of matching device pairs, jointly encodes the reconnaissance data, control commands, and status information to generate an encoded symbol stream, including: Step 5.1: Based on the data access scope in the device permission topology diagram, classify and categorize the reconnaissance data, control commands, and status information to obtain categorized data; then, divide the categorized data into high-security, medium-security, and low-security levels according to their security levels, specifically including: Based on the data access scope in the device permission topology diagram, reconnaissance data, control commands, and status information are categorized and segmented. For reconnaissance data, it is segmented according to the data acquisition subject and purpose into high-resolution orthophoto data, thermal imaging life detection data, and 3D modeling data of landslide areas. For control commands, it is segmented according to command priority and execution object into precise positioning commands, UAV trajectory adjustment commands, communication parameter configuration commands, and cluster collaborative scheduling commands. For status information, it is segmented according to the feedback subject into UAV remaining battery / attitude information, RIS array reflection status information, and channel quality feedback information. During the classification process, a unique identifier is added to each type of data to determine the data generating device, receiving device, and transmission direction, ensuring that data flow completely matches the access scope of the permission topology diagram and preventing unauthorized data transmission.

[0120] Subsequently, based on the data's security sensitivity and task priority, the categorized data was divided into three security levels. High-security data is defined as data types that impact core rescue effectiveness and whose leakage would lead to serious consequences. This includes thermal imaging life detection data, precise delivery instructions for first-aid kits, and core dispatch instructions for drone swarms. This type of data is directly related to life safety and mission success or failure, requiring the highest level of security protection. Medium-security data is defined as data types that support collaborative operations and whose leakage would affect operational efficiency. This includes high-resolution orthophoto data, 3D modeling data of collapsed areas, drone trajectory adjustment instructions, and drone attitude / battery information. This type of data is the foundation of swarm collaboration and requires a balance between security and transmission efficiency. Low-security data is defined as data types that assist in monitoring and whose leakage has no critical impact. This includes RIS array reflection status information, channel quality feedback information, and non-emergency communication parameter query instructions. This type of data prioritizes transmission efficiency and has relatively lower security requirements.

[0121] After the security level classification is completed, a security level label is added to each type of data and bound to a data identification label to form a structured data list of data identification, generating device, receiving device, and security level. The list is simultaneously distributed to the command relay and all matching device pairs to ensure that the data level can be quickly identified in subsequent encryption and encoding processes, providing a basis for differentiated processing and preventing unauthorized devices from accessing high-security data.

[0122] Step 5.2: Based on the session key, encrypt high-security data, medium-security data, and low-security data respectively to obtain first encrypted data, second encrypted data, and third encrypted data. Specifically, this includes: extracting the session key and key update mechanism generated in step 4.4, and adapting different encryption strengths and key usage strategies according to the data security level. For high-security data, a 256-bit AES encryption algorithm (Advanced Encryption Standard) is used. For each frame of high-security data transmitted (such as a single frame of thermal imaging life detection data), the session key is updated based on the latest generated quantum random number. The key update frequency is synchronized with the scheduling cycle of the dynamic frame scheduling scheme (1 second / time) to ensure the transmission security of highly sensitive data. For medium-security data, a 128-bit AES encryption algorithm is used, with a key update frequency of 5 seconds / time, balancing security and computational resource consumption, and adapting to the efficient transmission of large-volume data such as high-resolution images. For low-security data, a 64-bit AES encryption algorithm is used, with a key update frequency of 10 seconds / time, minimizing the computational load of the UAV while ensuring basic security, and avoiding affecting flight endurance.

[0123] The encryption process is based on a structured data list. The command relay unit acts as the encryption control core, centrally encrypting various types of data: after receiving raw data uploaded by each UAV, it calls the corresponding encryption algorithm and the currently valid session key according to the data's security level label, encrypting the data frame by frame. After encryption, an encryption identifier (including encryption algorithm type, key version number, and encryption timestamp) is added to the encrypted data. The key version number corresponds one-to-one with the generation batch of the quantum random number, facilitating decryption by the receiving end by matching the version number and calling the correct key. The first encrypted data (high security level), the second encrypted data (medium security level), and the third encrypted data (low security level) are stored in independent caches to avoid confusion between data of different security levels.

[0124] Step 5.3: Based on the communication capabilities and task roles of the devices in the matching device pair set, forward error correction codes are assigned to high-density, medium-density, and low-density data respectively to obtain the first, second, and third error correction codes. Specifically, this includes extracting the communication capability parameters (including channel quality indicators, spatial correlation coefficients, and delay spread parameters) and task role identifiers of the devices in the matching device pair set, and constructing an error correction code adaptation decision table based on data level, communication capability, and task role. For communication capability assessment, excellent channel quality (signal-to-noise ratio ≥ 20dB, bit error rate < 10) is considered. -5 For equipment pairs with high coding efficiency, error correction codes with good coding efficiency should be prioritized; for equipment pairs with average channel quality (10dB ≤ signal-to-noise ratio < 20dB, 10...), the error correction codes with high coding efficiency should be prioritized. -5 ≤Bit error rate<10 -3 For equipment pairs with poor channel quality (signal-to-noise ratio < 10dB, bit error rate ≥ 10), select error-correcting codes that balance coding efficiency and error correction capability;-3 For devices with strong error correction capabilities, select error correction codes.

[0125] Differentiated error correction code allocation is implemented based on a decision table: For high-density data (first-level encrypted data), considering its mission coreness and data integrity requirements, a high-order forward error correction code with strong error correction capabilities is allocated, specifically a Turbo code (coding rate 1 / 2). This coding scheme can effectively correct burst errors and random errors in mountainous channels with multipath fading and severe signal obstruction, ensuring lossless transmission of life detection data and accurate delivery commands. For medium-density data (second-level encrypted data), balancing transmission efficiency and error correction requirements, a convolutional code (coding rate 3 / 4) is allocated. This coding scheme has moderate computational complexity and is adaptable to high resolution. Rapid transmission of large volumes of data, such as images, is achieved while simultaneously meeting the integrity requirements of collaborative operation data. For low-density data (third-level encrypted data), Hamming codes with basic error correction capabilities (coding rate 7 / 8) are allocated with transmission efficiency as the core. This encoding scheme can quickly process auxiliary monitoring data without excessively consuming UAV computing resources and communication bandwidth. After the error correction codes are allocated, corresponding error correction code parameters (including encoding type, coding rate, and code length) are bound to each type of encrypted data, and the communication capability identifiers of the matching device pairs are associated to ensure that the receiving end can quickly call the corresponding decoding algorithm according to the channel characteristics of the device pair, thereby improving decoding efficiency.

[0126] Step 5.4: Interleave and fuse the first encrypted data, the second encrypted data, and the third encrypted data with the corresponding first error correction code, the second error correction code, and the third error correction code to form a layered coded data block; perform time-weighted sorting on the layered coded data block and convert it into a coded symbol stream, wherein the coded symbol stream includes data symbols, control symbols, and synchronization symbols, specifically including: The first encrypted data and the first error correction code, the second encrypted data and the second error correction code, and the third encrypted data and the third error correction code are interleaved and fused respectively. The interleaving operation adopts a block interleaving method, rearranging consecutive encrypted data blocks according to a preset interleaving matrix, and then combining them with the corresponding error correction code symbols according to the interval pattern of data symbols and error correction symbols to form hierarchical coded data blocks. For example, high-density hierarchical coded data blocks are combined with one thermal imaging data symbol and one Turbo error correction symbol, medium-density data blocks are combined with two image data symbols and one convolutional error correction symbol, and low-density data blocks are combined with four state feedback symbols and one Hamming error correction symbol, ensuring that the error correction code accurately covers the corresponding encrypted data and improving error recovery capabilities.

[0127] Subsequently, the three hierarchical coded data blocks were time-weighted and sorted. High-density data blocks (including life detection data and precision delivery commands) were assigned a weight of 0.5, prioritizing their allocation of transmission time resources to ensure the core data was transmitted first. Medium-density data blocks (including imagery and modeling data) were assigned a weight of 0.3, following closely behind the high-density data transmission. Low-density data blocks (including status feedback and auxiliary information) were assigned a weight of 0.2, placing them last in the time sequence. During the sorting process, combined with the time slot configuration of the dynamic frame scheduling scheme, high-density data blocks were mapped to time slots with optimal channel quality (such as unobstructed, low-interference line-of-sight time slots), while medium and low-density data blocks were sequentially mapped to the remaining time slots, ensuring the core data was transmitted under optimal channel conditions.

[0128] Finally, the sorted hierarchical encoded data blocks are converted into an encoded symbol stream. The symbol stream's structure is designed to adapt to the transmission characteristics of the UAV communication module, containing three core symbols: synchronization symbols (5%), employing the orthogonal Zadoff-Chu sequence, used by the receiver to achieve frame synchronization and time slot alignment, solving the synchronization deviation problem caused by channel delay fluctuations in mountainous areas; control symbols (10%), containing control information such as data security level identifiers, error correction code parameters, and interleaving matrix indexes, ensuring that the receiver can quickly parse the encoding rules of the symbol stream; and data symbols (85%), composed of interleaved and fused encrypted data symbols and error correction symbols, arranged sequentially according to the hierarchical sorting result. After the symbol stream is generated, it is modulated using the modulation scheme (such as 64QAM, 16QAM, etc.) in the communication parameter mapping table of the matching device pair to form the final transmittable baseband signal.

[0129] In a preferred embodiment of the present invention, step 6, which involves scheduling the coded symbol stream for targeted transmission based on the communication parameter mapping relationship of the matching device to the set, includes: Step 6.1: Based on the communication parameter mapping table, extract the modulation and coding scheme, retransmission mechanism parameters, and flow control thresholds for each matched device pair; configure symbol mapping rules according to the modulation and coding scheme, set automatic retransmission request strategies according to the retransmission mechanism parameters, and determine the data buffer size according to the flow control thresholds. Specifically, this includes: accurately extracting the core communication parameters of each matched device pair from the communication parameter mapping table, including modulation and coding schemes (such as 64QAM and Turbo coding, 16QAM and convolutional coding, etc.), retransmission mechanism parameters (retransmission mode, timeout time, maximum number of retransmissions), and flow control thresholds (uplink / downlink bandwidth thresholds). During the extraction process, the device pair identifier is used as an index to associate and bind the parameters with the corresponding task roles (such as core data transmission, collaborative operation, auxiliary monitoring) and channel quality indicators (signal-to-noise ratio, bit error rate) to ensure the pertinence of parameter configuration.

[0130] Symbol mapping rules are configured for modulation and coding schemes: Based on the signal constellation characteristics of different modulation types, differentiated symbol mapping logic is formulated for various modulation and coding schemes. For high-density data transmission equipment pairs using 64QAM (such as command relays and reconnaissance aircraft), the encoded binary data is mapped to corresponding constellation point symbols in groups of 6 bits, balancing transmission rate and anti-interference. For medium-density data equipment pairs using 16QAM modulation, mapping is performed in groups of 4 bits to optimize the transmission efficiency of medium-bandwidth data. For low-density data equipment pairs using QPSK modulation, mapping is performed in groups of 2 bits to prioritize correct demodulation of signals in weak channel environments. Symbol synchronization identifiers are embedded synchronously in the mapping rules to ensure that the receiver can quickly identify the modulation type and complete symbol demodulation, adapting to the time-varying characteristics of mountainous channels.

[0131] Automatic retransmission request strategies are set according to retransmission mechanism parameters: For real-time task equipment pairs using a stop-and-wait retransmission mechanism (such as command relays and delivery aircraft), the retransmission timeout is strictly matched with the time slot length of dynamic frame scheduling (such as 50ms), and the maximum number of retransmissions is set to 3 to avoid the cumulative retransmission delay affecting the timeliness of delivery commands; for non-real-time data equipment pairs using a sliding window retransmission mechanism (such as reconnaissance aircraft and command relays), retransmission trigger conditions are set according to the window size (such as 8 data blocks). When the number of unacknowledged data blocks in the window exceeds 3, batch retransmission is initiated, with a timeout of 100ms and a maximum number of retransmissions of 5 to ensure the integrity of large-volume data such as high-resolution images; retransmission priorities are set synchronously in the retransmission strategy, with retransmission requests for high-density data taking precedence over medium- and low-density data to avoid core data loss due to retransmission queuing.

[0132] The data buffer size is determined based on the flow control threshold: Combining the uplink / downlink flow thresholds of each device pair (e.g., an uplink threshold of 100Mbps for command and reconnaissance aircraft), the receiving and transmitting ring buffers are configured in a ratio of buffer capacity = flow threshold × 1.2, reserving 20% ​​redundancy to cope with sudden flow fluctuations in mountainous channels. For high-density data transmission device pairs, a dual-buffer alternating storage mechanism (separate receiving and processing buffers) is adopted to avoid conflicts between data reception and demodulation processing. For low-density data device pairs, a single-buffer configuration is used to save UAV storage resources, with the buffer size strictly controlled within 10% of the device's storage capacity, balancing data caching requirements and battery life. After all configurations are completed, the symbol mapping rules, retransmission strategy parameters, and buffer configuration instructions are sent to the corresponding matched device pairs via the control channel to ensure synchronous effectiveness at both ends.

[0133] Step 6.2: Calculate the beamforming weight matrix based on the spatial location information of the matching device pair and the spatial dimension component in the channel state tensor; based on the beamforming weight matrix, adjust the radiation mode of the UAV-borne phased array antenna to form a directional beam, specifically including: The system extracts real-time spatial location information of the transmitting end (e.g., reconnaissance aircraft) and receiving end (e.g., command relay aircraft) of the matching equipment. This information is generated by the airborne GPS / IMU integrated positioning system and includes three-dimensional coordinates (accuracy 0.1m), pitch / roll / heading attitude angles (accuracy 0.01°), and motion velocity vectors. The update frequency is 10Hz to adapt to the dynamic flight state of the UAV. From the spatial dimension components of the channel state tensor, the system filters the core parameters between the transmitting end and the receiving end, including the spatial attenuation coefficient (reflecting the signal loss degree of line-of-sight / non-line-of-sight links), multipath propagation path identifier (distinguishing between direct waves and RIS reflected wave paths), and the spatial correlation coefficient of adjacent equipment pairs (to determine the risk of beam interference overlap). The system retrieves the reflection status data of the RIS arrays in the corresponding coverage area, including the unit activation status of each RIS array, the current reflection phase configuration, and the link availability marker (e.g., unobstructed - available, partially obstructed - degraded).

[0134] All collected data must be timestamped to ensure that the spatiotemporal reference of the transmitter location, receiver location, channel parameters, and RIS status data at the same time is consistent. For location data in terrain-obstructed areas, the data is corrected in conjunction with the 3D terrain model in step 1.1. For example, when a drone flies into a canyon, the coordinate deviation caused by GPS signal drift is corrected by the terrain matching algorithm to ensure that the spatial location parameters are completely matched with the actual communication scenario.

[0135] Based on the preprocessed multi-source data, the communication link type of the matching device pair is determined, and the target path calculated by the weight matrix is ​​filtered. Combining the 3D terrain model and the spatial coordinates of the transmitter and receiver, the line-of-sight accessibility of the direct communication link is first determined: if there are no terrain obstructions (such as landslides or mountains) within the line of connection between the two, and the spatial attenuation coefficient is less than 80dB (adapting to the signal transmission characteristics of the emergency communication frequency band), then the direct link is determined as the main transmission path; if there are line-of-sight obstructions or the direct link attenuation coefficient is greater than 80dB, then the RIS auxiliary link filtering process is initiated.

[0136] RIS-assisted link selection prioritizes minimizing total path loss and maximizing interference orthogonality. The specific steps include: identifying 1-2 RIS arrays covering the area based on the spatial location of the transmitter and receiver (e.g., RIS at the canyon entrance, highland RIS); calculating the total path attenuation from transmitter to RIS array to receiver, i.e., the sum of the attenuation from transmitter to RIS and from RIS to receiver; eliminating invalid paths with a total attenuation greater than 100dB based on the reflection efficiency parameters fed back by the RIS arrays; assessing the interference risk of the remaining valid paths using the spatial correlation coefficients of adjacent device pairs, and selecting paths with a correlation coefficient less than 0.3 with other device pairs as target reflection links. If multiple valid reflection paths exist, the path with fewer RIS arrays and fewer reflection links is prioritized to reduce signal transmission delay.

[0137] Based on the characteristics of the target transmission path (direct link or RIS-assisted link) and combined with the interference suppression requirements, the beamforming weight matrix is ​​calculated in a hierarchical manner. The dimension of this matrix is ​​consistent with the number of elements of the airborne phased array antenna (e.g., a 128-element antenna corresponds to a 128×1 weight matrix). Each matrix element corresponds to the amplitude and phase control parameters of an antenna element.

[0138] The first layer calculates the basic gain weight to ensure the main lobe of the beam is accurately pointed towards the target direction. For direct links, the beam pointing reference is the direction of the line connecting the transmitter and receiver, and the main lobe gain requirement is determined by combining the spatial attenuation coefficient: for every 10dB increase in the attenuation coefficient, the main lobe gain increases by 3dB, but the maximum gain does not exceed the antenna hardware limit (e.g., 30dB). By adjusting the phase difference of each array element, the radiated signals of the antenna elements are made to be in phase and superimposed in the direction of the receiver. For example, the phase of the array element on the left leads that of the array element on the right, ensuring that the main lobe of the beam points to the receiver on the right. For RIS-assisted links, the beam pointing reference is divided into two stages: the first stage points to the incident area of ​​the RIS array, and the second stage combines the reflection phase parameters of the RIS array and calculates the link phase compensation to ensure that the signal reflected by the RIS can accurately point to the receiver. At the same time, the main lobe gain is adjusted according to the reflection efficiency of the RIS array. For example, when the RIS reflection efficiency is 80%, the main lobe gain is increased by 5dB compared to the direct link to compensate for reflection loss.

[0139] The second layer optimizes the interference suppression weights to reduce the interference of beam sidelobes to adjacent devices. It extracts the spatial location and channel quality parameters of adjacent device pairs from step 4.5 to identify potential interference targets affected by the current beam (such as another nearby reconnaissance aircraft). The azimuth and elevation angle ranges of the interference targets are marked using spatial correlation coefficients. Based on the basic gain weights, the amplitude attenuation of antenna elements pointing towards the interference target is adjusted; for example, the amplitude of the corresponding element is reduced by 50%, so that the gain of the beam sidelobes in the direction of the interference target is reduced by more than 15dB, ensuring that the interference power of the sidelobes to the interference target does not exceed -90dBm. Simultaneously, the element parameters in the main lobe direction remain unchanged to avoid affecting the signal strength of the core communication link.

[0140] The third layer is power constraint weight correction, ensuring that the weight matrix meets the equipment power control requirements. The transmit power level of the transmitting end in step 3.1 is extracted (e.g., level 1 10dBm, level 3 20dBm), and the total power amplitude of the weight matrix is ​​matched and calibrated with the transmit power level. If the calculated total power amplitude exceeds the threshold corresponding to the current level, the amplitude parameters of all array elements are attenuated proportionally to ensure that the total power does not exceed the limit. For matched equipment pairs for high-density data transmission (e.g., reconnaissance aircraft-command relay), within the allowable power range, the main lobe gain is prioritized, with only minor adjustments to the sidelobe array element parameters to balance communication quality and power consumption.

[0141] Based on the dynamic changes in UAV flight and channel environment, the beamforming weight matrix is ​​adjusted in real time. The attitude changes of the transmitting UAV are monitored in real time by the attitude sensor of the onboard IMU. If the pitch or heading angle changes by more than 0.5°, the phase parameters of the weight matrix are corrected according to the attitude change. For example, when the UAV's heading angle deflects 3° to the right, the phase of all array elements is adjusted synchronously, causing the main lobe of the beam to deflect 3° to the right simultaneously, thus offsetting the beam offset caused by the attitude change. Combined with the real-time update data of the channel state tensor (updated every 10ms), if the spatial attenuation coefficient of the target link suddenly increases, the amplitude of the array element corresponding to the main lobe gain is immediately increased to ensure stable signal strength at the receiving end.

[0142] The beamforming weighting matrix is ​​analyzed. Each element in the matrix corresponds to an element of the phased array antenna, and each element contains two sets of core parameters: amplitude coefficient and phase coefficient. During the analysis, the amplitude coefficient is standardized to a power control ratio (range 0-1); the phase coefficient is standardized to a recognizable angle value (range 0-360°). After analysis, a structured parameter table of element number-amplitude ratio-phase angle is generated according to the physical arrangement order of the antenna elements (e.g., linear arrangement or area array arrangement) to avoid mismapping of element parameters.

[0143] Based on a structured parameter table, digital control signals for each array element are generated. The phase control signal uses 16-bit binary encoding, directly mapping the phase angle from 0 to 360° (each bit corresponds to 0.014°, meeting the 0.1° accuracy requirement); the amplitude control signal uses 8-bit binary encoding, mapping the power ratio from 0 to 1 (each bit corresponds to a ratio accuracy of 0.0039). The control signals are transmitted via a high-speed serial peripheral interface (SPI) at a rate of 10 Mbps to ensure that all array element control signals are transmitted within 1 ms, avoiding array element coordination deviations caused by transmission delays. During transmission, the timestamp of the BeiDou time synchronization system is used as the synchronization reference, and a synchronization flag is added to the control signal frame header to ensure that all array element control signals are received and executed synchronously, with a deviation not exceeding 10 μs.

[0144] After receiving the digital control signal, the phase difference of the array element output signal is adjusted according to the encoded value of the phase control signal; the amplification factor is adjusted according to the encoded value of the amplitude control signal to ensure that the output power of the array element conforms to the standardized power ratio. For example, an amplitude command of 0.6 corresponds to an amplification factor of 60% of the reference value, ensuring that the beam main lobe gain meets the design requirements (e.g., 25dB). During the hardware driving process, the parameter adjustment response time of each array element does not exceed 50μs, ensuring rapid beamforming and adapting to the time slot requirements of dynamic frame scheduling.

[0145] After beam generation, the physical characteristics of the beam are acquired in real time, including the main lobe pointing angle, beamwidth, main lobe gain, and sidelobe suppression ratio. By receiving a preset pilot signal (the Zadoff-Chu sequence consistent with step 4.1), the actual pointing of the main lobe is calculated and compared with the designed pointing (such as the receiver position or the incident direction of the RIS array). If the deviation exceeds 0.5°, the phase parameters of the corresponding array elements are fine-tuned. This conversion process does not rely on external control nodes or public network support, and the parameter adjustment accuracy and hardware response speed are adapted to the dynamic communication needs in environments without public networks, ensuring that the directional beam can quickly adapt to link changes and maintain a stable signal transmission path.

[0146] Step 6.3: The encoded symbol stream is segmented according to the matching device pair set to generate sub-symbol streams corresponding to each matching device pair. Based on the symbol mapping rules, each sub-symbol stream is mapped to a physical resource block to form a scheduling transmission unit. Specifically, this includes segmenting the encoded symbol stream generated in Step 5.4 based on the communication parameters (bandwidth requirements, flow control thresholds, modulation and coding schemes) of the matching device pair set. The segmentation logic is centered on the specific requirements of each device pair: the symbol stream length is allocated according to the bandwidth threshold of each device pair. For example, the bandwidth of the command relay aircraft and reconnaissance aircraft device pair is 20MHz, and the allocated sub-symbol stream length accounts for 40% of the total symbol stream (adapting to high-resolution image transmission requirements); the bandwidth of the delivery aircraft and command aircraft device pair is 5MHz, and the allocated sub-symbol stream length accounts for 10% of the total symbol stream (adapting to control command transmission). During the segmentation process, unique identification information is added to each sub-symbol stream, including the sending device ID, receiving device ID, data security level, symbol stream length, and checksum, to ensure that the receiving end can accurately identify and receive the target data and avoid data confusion between multiple device pairs.

[0147] The segmented sub-symbol streams require preprocessing optimization: For high-density data sub-symbol streams, an additional frame synchronization identifier (a 16-bit fixed sequence) is added to improve the synchronization speed and anti-interference capability of the receiver; for large-volume medium-density data sub-symbol streams, they are divided into blocks according to the size of the physical resource blocks (e.g., each block contains 1024 symbols) to facilitate subsequent mapping and retransmission management; for low-density data sub-symbol streams, the preprocessing process is simplified, retaining only the core identifier information to improve processing efficiency. The preprocessed sub-symbol streams are stored in the dedicated buffer of the corresponding device pair, awaiting resource mapping.

[0148] Subsequently, based on the symbol mapping rules and the time-frequency resource planning of the dynamic frame scheduling scheme, each sub-symbol stream is mapped to a Physical Resource Block (PRB). The division of the Physical Resource Block is based on the carrier frequency allocation table in step 2.4 and the time-frequency two-dimensional overlapping frame structure in step 3.2. Each resource block corresponds to a fixed time-frequency resource unit (e.g., 1ms × 1MHz), and the channel quality level (excellent / good / poor) of the resource block is marked. The mapping logic follows the principle of adapting high-quality resources to high-priority data: sub-symbol streams of high-density data are preferentially mapped to resource blocks with excellent channel quality (SNR≥20dB), and the resource blocks are allocated continuously to reduce handover losses; sub-symbol streams of medium-density data are mapped to resource blocks with good channel quality (10dB≤SNR<20dB), and non-contiguous allocation can be used to improve resource utilization; sub-symbol streams of low-density data are mapped to the remaining resource blocks to adapt to their low transmission requirements.

[0149] The dynamic frame scheduling scheme uses a time-frequency resource occupancy table to monitor the allocation status of each resource block in real time, ensuring that the same resource block is mapped to only one device pair's sub-symbol stream within the same time slot. If a resource conflict occurs during the mapping of a device pair's sub-symbol stream, the resource block allocation of the lower-priority device pair is adjusted first to ensure the transmission needs of the higher-priority device pair. After mapping, a mapping table of sub-symbol stream identifier, physical resource block number, time-frequency coordinates, and channel quality level is generated and distributed to the transmitting device and the command relay for subsequent transmission scheduling and link monitoring. Each mapped resource block is combined to form an independent scheduling transmission unit, which contains complete sub-symbol stream blocks, identification information, and checksums, and has independent transmission and demodulation capabilities.

[0150] Step 6.4: Based on the frame scheduling timing table in the dynamic frame scheduling scheme, allocate transmission time slots for each scheduling transmission unit; combine the directional beam to send the scheduling transmission unit within the allocated transmission time slots to achieve directional transmission scheduling. Specifically, this includes: extracting the frame scheduling timing table in the dynamic frame scheduling scheme, which clarifies the time slot allocation priority, available time slot segments, and time slot lengths for each device pair (based on the decision tree optimization results of step 3.4). Time slot allocation follows the core principle of "priority tiering and staggered timing": Level 1 priority equipment pairs (high-density data transmission between command and reconnaissance aircraft) are allocated to core time slots with the best channel quality and least interference (such as the first 30% of time slots within a frame period), and the time slot length is adapted to the size of the sub-symbol stream blocks (such as a 500ms time slot corresponding to large volume image data); Level 2 priority equipment pairs (coordinated data transmission between reconnaissance aircraft) are allocated to secondary core time slots (the middle 40% of time slots within a frame period), and the time slot length is adjusted as needed; Level 3 priority equipment pairs (low-density data transmission between delivery aircraft and command aircraft) are allocated to edge time slots (the last 30% of time slots within a frame period), ensuring that core tasks have priority access to high-quality timing resources.

[0151] When allocating specific transmission time slots to each scheduling transmission unit, synchronous planning must be performed in conjunction with the time-frequency coordinates of the physical resource blocks: ensuring that the time slot allocation of the scheduling transmission unit is consistent with the time dimension of the corresponding physical resource block to avoid time-frequency resource mismatch; at the same time, based on the spatial location and beam pointing of the matching equipment pair, the beam coverage time slots of other equipment pairs are avoided to reduce cross-interference between directional beams. For example, the cooperative transmission time slots of two adjacent reconnaissance aircraft need to be staggered by more than 50ms, and the beam pointing angle should be greater than 15° to ensure that they do not interfere with each other; after the time slot allocation is completed, a transmission scheduling table of "scheduling transmission unit - transmission time slot - physical resource block - beam parameters" is generated and sent to each transmitting end device, and the global scheduling status of the command relay is updated synchronously.

[0152] The transmission execution phase is centered on directional beams and uses a transmission scheduling table to initiate data transmission: the transmitting device activates the directional beams within a preset time slot based on the start time of the transmission time slot (based on BeiDou time synchronization with an accuracy of 10μs), and loads the symbol stream of the scheduled transmission unit onto the corresponding time-frequency resources according to the physical resource block mapping relationship; during transmission, the update data of the channel state tensor is monitored in real time, and if the channel quality suddenly deteriorates (such as a sharp drop in SNR), the fast retransmission strategy in the retransmission mechanism parameters is immediately activated to complete the retransmission in the current time slot or the next backup time slot to avoid data loss; for links involving RIS arrays, the reflection status of the RIS array is monitored synchronously during transmission to ensure the stability of the beam reflection path.

[0153] Based on preset information in the transmission scheduling table, the receiving device pre-adjusts its beam direction (coordinating with the transmitting beam) and receives scheduled transmission units within the corresponding time slot. It filters target data using sub-symbol stream identifiers, demodulates and decodes the data using symbol mapping rules, and ultimately recovers the original data. After reception, it sends a reception confirmation signal (including reception status and bit error rate) back to the transmitting end. The transmitting end updates its buffer status based on the confirmation signal and directs the relay to synchronously update the global transmission status, forming a closed-loop scheduling mechanism of "allocation-transmission-feedback-adjustment". The entire process does not rely on public network infrastructure. Through self-organized time slot allocation and directional beam coordination, it enables efficient, low-interference parallel transmission by multiple devices in complex mountainous environments, ensuring the continuous operation of emergency rescue missions.

[0154] like Figure 2 As shown, embodiments of the present invention also propose a near-field communication matching and processing system for unmanned aerial vehicles (UAVs) in areas without public network communication, comprising: The control module is used to control the near-field electromagnetic wave propagation environment to obtain three-dimensional channel characteristic data; it collects three-dimensional channel characteristic data and generates a channel state tensor. The optimization module is used to construct a quantum annealing optimization model based on the channel state tensor; and to solve the spectrum resource allocation problem by simulating the quantum tunneling effect according to the quantum annealing optimization model, so as to obtain the carrier frequency, time slot offset and power control parameter set of each UAV. The module is used to construct an overlapping communication frame structure based on the carrier frequency, time slot offset, and power control parameter set; and to generate a dynamic frame scheduling scheme based on the overlapping communication frame structure. The matching module is used to extract the channel reciprocity feature sequence based on the dynamic frame scheduling scheme; generate quantum random numbers based on the channel reciprocity feature sequence; generate a session key and a device permission topology map based on the quantum random numbers; and perform communication capability matching on the devices in the UAV cluster based on the device permission topology map and the channel state tensor, generating a set of matched device pairs and their communication parameter mapping relationship. The processing module is used to jointly encode reconnaissance data, control commands, and status information based on the session key, device permission topology map, and matching device pair set to generate an encoded symbol stream; and to perform directional transmission scheduling of the encoded symbol stream according to the communication parameter mapping relationship of the matching device pair set to complete the near-field communication matching processing between UAVs.

[0155] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for near-field communication matching and processing of unmanned aerial vehicles (UAVs) in areas without public network communication, characterized in that, The method includes: The near-field electromagnetic wave propagation environment is controlled to obtain three-dimensional channel characteristic data; the three-dimensional channel characteristic data is collected and a channel state tensor is generated. Based on the channel state tensor, a quantum annealing optimization model is constructed; according to the quantum annealing optimization model, the spectrum resource allocation problem is solved by simulating the quantum tunneling effect to obtain the carrier frequency, time slot offset and power control parameter set of each UAV. Based on the carrier frequency, time slot offset, and power control parameter set, an overlapping communication frame structure is constructed; based on the overlapping communication frame structure, a dynamic frame scheduling scheme is generated. Based on the dynamic frame scheduling scheme, a channel reciprocity feature sequence is extracted; a quantum random number is generated based on the channel reciprocity feature sequence; a session key and a device permission topology map are generated based on the quantum random number; based on the device permission topology map and the channel state tensor, the communication capabilities of devices in the UAV cluster are matched, and a set of matched device pairs and their communication parameter mapping relationship are generated. Based on the session key, device permission topology map, and matching device pair set, reconnaissance data, control commands, and status information are jointly encoded to generate an encoded symbol stream. Based on the mapping relationship of communication parameters of the matching device to the set, the coded symbol stream is scheduled for directional transmission to complete the near-field communication matching process between UAVs.

2. The near-field communication matching processing method for UAVs in areas without public network communication as described in claim 1, characterized in that, The near-field electromagnetic wave propagation environment is modulated to obtain three-dimensional channel characteristic data; Acquire three-dimensional channel feature data and generate a channel state tensor, including: By deploying a reconfigurable smart surface array, the propagation path of near-field electromagnetic waves in the area without public network communication is dynamically controlled to form multi-dimensional spatial beamforming; based on the multi-dimensional spatial beamforming, the beam pointing parameters of the UAV airborne phased array antenna are configured. Based on the beam pointing parameters, the UAV performs multi-point scanning measurements in three-dimensional space to acquire channel impulse response data at different spatial locations, time slices, and frequency dimensions; the channel impulse response data is then processed by time-frequency analysis to extract three-dimensional channel feature data. Based on the aforementioned three-dimensional channel feature data, multi-dimensional information is collected through distributed sensing nodes. This multi-dimensional information includes spatial location information, timestamp information, and frequency domain response information. Spatiotemporal alignment calibration is then performed on the multi-dimensional information to obtain calibrated multi-dimensional information. The calibrated multidimensional information is tensorized and organized according to spatial, temporal and frequency dimensions to construct a three-dimensional channel feature matrix. The three-dimensional channel feature matrix is ​​then normalized and dimensionally compressed to generate a channel state tensor that characterizes the spatial correlation of the channel.

3. The near-field communication matching processing method for UAVs in areas without public network communication as described in claim 2, characterized in that, Based on the channel state tensor, a quantum annealing optimization model is constructed. According to the quantum annealing optimization model, the spectrum resource allocation problem is solved by simulating the quantum tunneling effect to obtain the carrier frequency, time slot offset, and power control parameter set for each UAV, including: Based on the spatial correlation and time-frequency characteristics in the channel state tensor, an objective function for spectrum resource allocation is constructed; according to the communication requirements and topology of the UAV swarm, a set of constraints for spectrum resource allocation is determined. The objective function and set of constraints are converted into the Ising model expression form to construct the Hamiltonian of the quantum annealing optimization model; the spin state and coupling strength parameters of the qubit are initialized according to the ground state energy distribution characteristics of the Hamiltonian. By simulating the quantum tunneling effect, the transverse magnetic field strength is dynamically adjusted during the quantum annealing process. When the energy converges to the stable threshold, the final spin state configuration of the qubit is recorded. Based on the final spin state configuration of the qubits, the spectrum resource allocation scheme corresponding to each UAV is decoded; the spectrum resource allocation scheme is mapped into a specific carrier frequency allocation table, time slot offset configuration table and power control parameter set to form a complete spectrum resource scheduling instruction set.

4. The near-field communication matching processing method for unmanned aerial vehicles in areas without public network communication as described in claim 3, characterized in that, Based on the spatial correlation and time-frequency characteristics in the channel state tensor, an objective function for spectrum resource allocation is constructed, including: Spatial dimension components, temporal dimension components, and frequency dimension components are separated from the channel state tensor; the spatial isolation matrix between UAV nodes is calculated based on the spatial dimension components, the channel coherence time series is extracted based on the temporal dimension components, and the frequency domain flatness index is obtained based on the frequency dimension components. Based on the spatial isolation matrix, calculate the normalized spatial orthogonality coefficients between each UAV pair; based on the channel coherence time series, determine the normalized time window length; based on the frequency domain flatness index, divide the normalized subcarrier aggregation region. The normalized spatial orthogonality coefficients are mapped to the spatial multiplexing gain index, the normalized time window length is mapped to the time-domain scheduling efficiency index, and the normalized subcarrier aggregation area is mapped to the spectrum utilization index. The basic objective function is constructed by multiplying the spatial reuse gain index, the time-domain scheduling efficiency index, and the spectrum utilization index with their corresponding preset weight coefficients and then summing them. Interference suppression constraints and power consumption constraints are set in the basic objective function to form the objective function for spectrum resource allocation.

5. The near-field communication matching processing method for UAVs in areas without public network communication as described in claim 4, characterized in that, Based on the carrier frequency, time slot offset, and power control parameter set, an overlapping communication frame structure is constructed. Based on the overlapping communication frame structure, a dynamic frame scheduling scheme is generated, including: Based on the carrier frequency allocation table, an independent subcarrier group is allocated to each UAV to obtain the subcarrier group allocation result; based on the time slot offset configuration table, a frame start time offset is set for each UAV; based on the power control parameter set, a corresponding transmit power level is configured for each UAV. Based on the subcarrier group allocation results and frame start time offset, a time-frequency two-dimensional overlapping frame structure is constructed; in the time-frequency two-dimensional overlapping frame structure, a control channel region, a data channel region, and a pilot channel region are set; according to the transmit power level, a power allocation ratio is configured for different regions; Based on the aforementioned time-frequency two-dimensional overlapping frame structure, inter-frame interference patterns and channel occupancy are analyzed to obtain channel occupancy and interference patterns; based on channel occupancy and interference patterns, frame scheduling conflict detection rules and priority determination criteria are determined. A frame scheduling decision tree is constructed based on the frame scheduling conflict detection rules and priority determination criteria. Based on the frame scheduling decision tree and combined with real-time channel state feedback information, an adaptive dynamic frame scheduling scheme is generated. The dynamic frame scheduling scheme includes a frame scheduling timing table, power adjustment instructions, and conflict avoidance strategies.

6. The near-field communication matching processing method for UAVs in areas without public network communication as described in claim 5, characterized in that, Based on the dynamic frame scheduling scheme, the channel reciprocity feature sequence is extracted; Quantum random numbers are generated based on the channel reciprocity characteristic sequence; Based on quantum random numbers, a session key and device permission topology diagram are generated, including: Based on the frame scheduling time table in the dynamic frame scheduling scheme, a bidirectional channel probing time slot is configured; within the bidirectional channel probing time slot, the UAV is controlled to perform reciprocal channel measurements between the pairs to obtain uplink channel response data and downlink channel response data; The uplink channel response data and downlink channel response data are subjected to reciprocity verification processing, and the channel reciprocity error index is calculated. When the channel reciprocity error index is less than a preset threshold, the channel amplitude fluctuation sequence and phase jitter sequence are extracted and merged to generate a channel reciprocity feature sequence. Based on the channel reciprocity characteristic sequence, an original random bit stream is generated; the original random bit stream is then post-processed to generate quantum random numbers that conform to the quantum randomness standard. Based on the quantum random number, a session key is generated; according to the topology and task roles of the UAV cluster, a device permission topology map is generated in combination with the session key, and the device permission topology map defines the communication permission level and data access range between each device.

7. The near-field communication matching processing method for unmanned aerial vehicles in areas without public network communication as described in claim 6, characterized in that, Based on the device permission topology map and channel state tensor, communication capabilities are matched among devices in the UAV swarm, generating a set of matched device pairs and their communication parameter mapping relationships, including: The communication permission level, data access range, and task role identifier of each UAV device are extracted from the device permission topology diagram; the channel quality index, spatial correlation coefficient, and delay spread parameter between each UAV pair are extracted from the channel state tensor. Based on communication permission levels and task role identifiers, a device communication requirement matrix is ​​constructed; based on channel quality indicators, spatial correlation coefficients, and delay spread parameters, a channel matching degree evaluation matrix is ​​constructed; the device communication requirement matrix and the channel matching degree evaluation matrix are weighted and fused to generate a comprehensive matching score matrix. Based on the comprehensive matching score matrix, the drone devices are paired up in pairs to obtain a set of successfully matched device pairs; For the set of successfully matched device pairs, communication parameters are mapped, and a corresponding modulation and coding scheme, retransmission mechanism parameters, and flow control threshold are assigned to each matched device pair. The modulation and coding scheme, retransmission mechanism parameters, and flow control threshold are then bound to the matched device pairs to generate a communication parameter mapping table containing device identifiers, matching weights, and communication parameters.

8. The near-field communication matching processing method for unmanned aerial vehicles in areas without public network communication as described in claim 7, characterized in that, Based on the session key, device permission topology map, and matching device pair set, reconnaissance data, control commands, and status information are jointly encoded to generate an encoded symbol stream, including: Based on the data access scope in the device permission topology diagram, the reconnaissance data, control commands, and status information are classified and categorized to obtain classified data; the classified data is then divided into high-security, medium-security, and low-security data according to their security levels. Based on the session key, high-security data, medium-security data, and low-security data are encrypted respectively to obtain first encrypted data, second encrypted data, and third encrypted data; Based on the communication capabilities and task roles of the matching devices in the set, forward error correction codes are assigned to high-security data, medium-security data, and low-security data respectively to obtain the first error correction code, the second error correction code, and the third error correction code. The first encrypted data, the second encrypted data, and the third encrypted data are interleaved and fused with the corresponding first error correction code, the second error correction code, and the third error correction code to form a layered coded data block; the layered coded data block is then time-weighted and sorted, and converted into a coded symbol stream, which includes data symbols, control symbols, and synchronization symbols.

9. The near-field communication matching processing method for unmanned aerial vehicles in areas without public network communication as described in claim 8, characterized in that, Based on the mapping relationship of communication parameters of the matching device to the set, the coded symbol stream is scheduled for targeted transmission, including: Based on the communication parameter mapping table, the modulation and coding scheme, retransmission mechanism parameters, and flow control threshold of each matching device pair are extracted; symbol mapping rules are configured according to the modulation and coding scheme, automatic retransmission request strategy is set according to the retransmission mechanism parameters, and data buffer size is determined according to the flow control threshold. Based on the spatial location information of the matching device pair and the spatial dimension component in the channel state tensor, a beamforming weight matrix is ​​calculated; based on the beamforming weight matrix, the radiation mode of the UAV-borne phased array antenna is adjusted to form a directional beam. The encoded symbol stream is divided according to the set of matching devices to generate sub-symbol streams corresponding to each matching device pair; according to the symbol mapping rules, each sub-symbol stream is mapped to a physical resource block to form a scheduling transmission unit. Based on the frame scheduling timing table in the dynamic frame scheduling scheme, a transmission time slot is allocated to each scheduling transmission unit; combined with the directional beam, the scheduling transmission unit is sent within the allocated transmission time slot to achieve directional transmission scheduling.

10. A near-field communication matching and processing system for unmanned aerial vehicles (UAVs) in areas without public network communication, wherein the system implements the method as described in any one of claims 1 to 9, characterized in that, include: The control module is used to control the near-field electromagnetic wave propagation environment in order to obtain three-dimensional channel characteristic data; Collect three-dimensional channel feature data and generate a channel state tensor; The optimization module is used to construct a quantum annealing optimization model based on the channel state tensor; and to solve the spectrum resource allocation problem by simulating the quantum tunneling effect according to the quantum annealing optimization model, so as to obtain the carrier frequency, time slot offset and power control parameter set of each UAV. The module is used to construct an overlapping communication frame structure based on the carrier frequency, time slot offset, and power control parameter set; and to generate a dynamic frame scheduling scheme based on the overlapping communication frame structure. The matching module is used to extract the channel reciprocity feature sequence based on the dynamic frame scheduling scheme; Quantum random numbers are generated based on the channel reciprocity characteristic sequence; Based on quantum random numbers, a session key and a device permission topology map are generated; based on the device permission topology map and the channel state tensor, the communication capabilities of devices in the UAV swarm are matched to generate a set of matched device pairs and their communication parameter mapping relationship. The processing module is used to jointly encode reconnaissance data, control commands, and status information based on the session key, device permission topology map, and matching device pair set to generate an encoded symbol stream; and to perform directional transmission scheduling of the encoded symbol stream according to the communication parameter mapping relationship of the matching device pair set to complete the near-field communication matching processing between UAVs.

Images