A low-altitude channel measurement method and device based on multi-machine dynamic topology and weakly correlated signals

Abstract

The application discloses a low-altitude channel measurement method and device based on multi-machine dynamic topology and weakly correlated signals, which comprises the following steps: determining the upper limit of parallel nodes and planning the transmission timing based on the dynamic range threshold, taking the weakly cross-correlated pseudo-random sequence as the basis, searching for the optimal measurement sequence by using the genetic algorithm, and forming a global scheduling scheme; performing the actual measurement according to the scheduling scheme, separating the target signal, suppressing the mutual interference through the time-domain sliding correlation operation, and obtaining the multipath channel characteristics and the original measurement data; for the original measurement data, removing the redundant information from the time domain, the time delay domain and the space domain by using the hardware parallel real-time processing algorithm, extracting the small-scale fading statistical characteristics, identifying the effective multipath components and completing the space domain sparse sampling, and reducing the data storage and backhaul bandwidth demand on the premise of reserving the key characteristics of the channel. The application can realize the low-altitude channel measurement of multiple unmanned aerial vehicles in parallel, and is suitable for the deployment and performance evaluation of the low-altitude communication network.

Description

Technical Field

[0001] This invention relates to the field of wireless channel measurement and characteristic analysis technology, specifically to a low-altitude channel measurement method and apparatus based on multi-machine dynamic topology and weakly correlated signals, particularly for channel measurement, interference suppression, and low-redundancy channel feature extraction for various types of low-altitude communication links such as air-to-air (A2A), air-to-ground (A2G), and ground-to-ground (G2G). Background Technology

[0002] The sixth-generation mobile communication system aims to build an integrated air-space-ground network to achieve seamless global coverage. As a key link connecting terrestrial and satellite networks, low-altitude networks play a vital role in providing communication services to airborne platforms and ground terminals. Unlike traditional terrestrial communication networks, the low-altitude communication environment exhibits significant three-dimensional wide-area spatial characteristics, including multiple communication links such as A2A, A2G, and G2G, and the radio wave propagation channel has obvious dynamic non-stationary characteristics.

[0003] Existing low-altitude channel measurement technologies still have significant limitations: First, in terms of measurement equipment, current low-altitude channel measurement devices are mainly point-to-point, making it difficult to quickly complete coordinated measurements of different altitudes and various link types such as A2A, A2G, and G2G in a wide area. Furthermore, multi-node measurement devices lack effective topology mechanisms and methods to suppress inter-node interference in measurement signals. Second, in terms of data processing, facing the high-density sampling requirements of highly dynamic low-altitude scenarios, existing methods for extracting channel measurement data suffer from high redundancy and poor real-time performance, making it difficult to achieve full-link measurement data storage and real-time analysis under limited storage resources and time constraints. Therefore, there is an urgent need for a low-altitude channel measurement device and method based on multi-machine dynamic topology and weakly correlated signals to solve the problems of low channel measurement efficiency and high data processing redundancy in wide-area, highly dynamic scenarios. Summary of the Invention

[0004] The purpose of this invention is to provide a low-altitude channel measurement method and apparatus based on multi-machine dynamic topology and weak correlation signals, which can effectively solve the shortcomings of existing technologies such as low efficiency of wide-area spatial channel measurement, severe self-interference and cross-interference during multi-node parallel measurement, high redundancy of measurement data leading to difficulties in backhaul and storage, and poor real-time performance of offline processing.

[0005] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows:

[0006] In a first aspect, the present invention discloses a low-altitude channel measurement method based on multi-machine dynamic topology and weakly correlated signals, the method comprising:

[0007] S1. For low-altitude, high-dynamic scenarios involving parallel launches of multiple UAVs and mutual interference between multiple nodes, a dynamic range threshold is preset. Based on the dynamic range threshold, the upper limit of parallel nodes is determined and the launch sequence of each node is planned to avoid temporal interference. A candidate pseudo-random sequence space is constructed based on a weakly cross-correlated pseudo-random sequence. A genetic algorithm is used to search for the optimal measurement sequence with the lowest cross-correlation and the corresponding parameter set in the candidate pseudo-random sequence space to suppress waveform domain interference. Through the coordination of launch sequence and optimal measurement sequence, a global scheduling scheme for multi-UAV measurement is formed.

[0008] S2, according to the global scheduling scheme planned in step S1, the actual measurement task is executed. Combined with the optimal measurement sequence searched in step S1, the received superimposed signal is processed in real time using time-domain sliding correlation operation to separate the target signal and suppress mutual interference between nodes in order to obtain multipath channel characteristics and obtain the original measurement data.

[0009] S3. For the raw measurement data generated in step S2, a hardware parallel real-time processing algorithm is used to remove redundant information from three dimensions: time domain, time delay domain, and spatial domain. Small-scale fading statistical features are extracted, effective multipath components are identified, and spatial sparse sampling is completed. This reduces the data storage and backhaul bandwidth requirements while preserving key channel features.

[0010] Further, in step S1, the process of determining the upper limit of parallel nodes based on the dynamic range threshold includes the following steps:

[0011] To address the channel measurement accuracy requirements, hardware system performance, and scene interference characteristics in low-altitude, high-dynamic scenarios involving parallel launches of multiple UAVs and mutual interference between multiple nodes, a preset dynamic range threshold is established. According to the preset dynamic range threshold Determine the upper limit on the number of drones that can transmit signals simultaneously in parallel:

[0012]

[0013] In the formula, Indicates being received The dynamic range of the measured channel after interference from a single UAV is determined based on the background benchmark of multi-UAV mutual interference accumulation and the peak characteristics obtained by sequence correlation calculation, and is expressed as:

[0014]

[0015] In the formula, Indicates the first The measurement sequence and the first The sliding cross-correlation results of the measurement sequences , Indicates the first The results of the sliding autocorrelation of the measurement series.

[0016] Furthermore, in step S1, the planning process for the transmission timing of each node includes the following steps:

[0017] A1, Calculate the lower bound of the number of switching operations. ,make ;

[0018] A2, let Initialize in each switching time slot ;

[0019] A3, using seed set generate A set of cyclic shifters, from which to select... A set of non-mutually exclusive elements that satisfy... The set is placed greedily, prioritizing the scarcest time slots;

[0020] A4. If all measurement links are successfully traversed, output the required number of switching operations. And the launch sequence of each drone, end the process, otherwise... Return to step A1.

[0021] Further, in step S1, the process of obtaining the optimal measurement sequence includes the following steps:

[0022] B1, using a weakly cross-correlated pseudo-random sequence as the basis of the measurement signal, defines the first... Measurement sequence of a drone The mapping relationship between it and its generated parameters:

[0023]

[0024] In the formula, Indicates the sequence length. The generator function for the selected pseudo-random sequence; For the first The set of sequence generation parameters for each UAV includes at least the measurement sequence length and the number of measurement nodes;

[0025] Chromosome size is determined based on sequence length, and chromosome number is determined based on number of measurement nodes. The generation parameters for each UAV are then used. Chromosomes encoded as genetic algorithms are randomly generated and contain... The initial population of individuals;

[0026] B2, under the constraint of satisfying autocorrelation characteristics, find the optimal parameter set for each UAV. To minimize the maximum cross-correlation peak among all drone pairs, a minimax optimization problem is constructed:

[0027]

[0028] In the formula, Indicates generation The set of parameters required for a measurement sequence; Indicates the first The measurement sequence and the first The measurement sequence is in time delay The results of the sliding cross-correlation calculation are as follows; This indicates the cross-correlation amplitude;

[0029] B3 involves an intelligent search within the candidate pseudo-random sequence space, the specific process of which is as follows:

[0030] A set of measurement sequences is treated as an individual, and its fitness is evaluated using the maximum cross-correlation peak among the sequences. The smaller the cross-correlation peak, the higher the fitness. Survival selection is performed using a roulette wheel selection method to choose parent individuals. Single-point crossover and random mutation operations are performed on the population to update the sequence generation parameters and produce a new generation. This process is repeated iteratively until the maximum number of iterations or the fitness threshold is met, outputting the optimal combination of measurement sequences that satisfies the low cross-correlation constraint and its corresponding parameter set. .

[0031] Step S3 further includes the following steps:

[0032] C1, Real-time extraction of time-domain equivalent fading features: Based on the original measurement data obtained in step S2, in a length of... Within the statistical window, the orthogonal fading waveform is denoted as and ,in, and represent the continuous sampled values ​​of the in-phase component (I-path) of the quadrature fading waveform within the statistical window. This represents the continuous sampled values ​​of the quadrature components (Q-path) of the quadrature fading waveform within the statistical window. This indicates the number of sampling points contained within the statistical window; a hardware parallel computing architecture is used to calculate the mean and variance of the orthogonal fading waveform as unbiased estimates. , ,in and These are the mean and standard deviation, respectively.

[0033] C2, Real-time extraction of effective multipath features in the time delay domain: For the original measurement data obtained in step S2, the effective multipath component extraction algorithm based on dynamic threshold is used to remove redundancy from the time delay domain information, and real-time redundancy removal of multipath in the time delay domain is completed to extract the effective multipath components of the channel impulse response.

[0034] C3: Real-time planning for sparse spatial sampling, which includes the following three steps:

[0035] C31: Calculate the power delay spectrum based on the effective multipath components of the channel impulse response extracted in real time in step C2:

[0036]

[0037] In the formula, express The spatial three-dimensional coordinates corresponding to the movement trajectory of the drone at any time. Indicates the effective multipath component. Indicates time delay;

[0038] C32: Based on the correlation coefficient of the power delay spectrum, the spatial sparsity boundary of the channel slice is determined in real time to achieve redundancy removal of spatial sampling locations and power delay spectrum correlation coefficient. Represented as:

[0039]

[0040] in, This represents the spatial displacement relative to the reference spatial position. Indicates offset along the preset measurement trajectory The candidate sampling positions obtained afterwards Indicates the reference spatial position In time delay The power delay spectrum below;

[0041] C33: Using correlation coefficient threshold Constraining the sparse sampling points in spatial location, the sampling points that satisfy the conditions are represented as:

[0042]

[0043] And update This is used to determine the next sparse sampling point;

[0044] C34: Using the updated reference position as the input for the next round of calculation, repeat steps C31 to C33 until all preset spatial nodes of the current channel slice have been traversed, thus completing the real-time planning of spatial sparse sampling for the current region.

[0045] In a second aspect, the present invention discloses a low-altitude channel measurement device based on multi-machine dynamic topology and weak correlation signals. The device includes a spatiotemporal reference unit (1-1), a core processing unit (1-2), a radio frequency front-end unit (1-3), and a data interaction and storage unit (1-4).

[0046] The spatiotemporal reference unit (1-1) provides the device with a unified high-precision time synchronization signal and spatial location information, and transmits them to the core processing unit (1-2). The core processing unit (1-2) generates measurement sequences according to the cooperative topology planning, controls the radio frequency front-end unit (1-3) to generate and transmit measurement signals, and also performs multi-domain real-time redundancy removal processing and small-scale fading characteristic analysis on the measurement signals received by the radio frequency front-end unit (1-3) after wireless channel transmission, and extracts effective channel feature data. The radio frequency front-end unit (1-3) amplifies and transmits the measurement signals with high power, and amplifies the received signals with low noise before transmitting them to the core processing unit (1-2). The data interaction and storage unit (1-4) stores the channel feature data output by the core processing unit (1-2) and performs scheduling command interaction between multiple nodes through the data transmission link.

[0047] Furthermore, the core processing unit (1-2) adopts a parallel processing architecture, which integrates a collaborative topology planning module (1-2-2), a measurement sequence generation module (1-2-3), a small-scale fading extraction module (1-2-1), and a multi-domain real-time deduplication module (1-2-4). The collaborative topology planning module (1-2-2) is connected to the measurement sequence generation module (1-2-3), and the output of the measurement sequence generation module (1-2-3) is connected to the transmit signal input of the RF front-end unit (1-3). The receive signal output of the RF front-end unit (1-3) is connected to a parallel splitting node, which is connected to the input of the small-scale fading extraction module (1-2-1) and the input of the multi-domain real-time deduplication module (1-2-4) to realize parallel splitting processing of the received signal.

[0048] Furthermore, the RF front-end unit (1-3) includes a measurement high-power amplifier, a measurement low-noise amplifier, a measurement transmitting antenna, and a measurement receiving antenna; the input terminal of the measurement high-power amplifier is connected to the analog signal output port of the core processing unit (1-2), and its output terminal is connected to the measurement transmitting antenna; the measurement receiving antenna is connected to the input terminal of the measurement low-noise amplifier, and the output terminal of the measurement low-noise amplifier is connected to the analog signal input port of the core processing unit (1-2); the control interface of the core processing unit (1-2) is connected to the enable terminal of the amplifier or an RF switch to control the physical on / off state of the transceiver link.

[0049] Furthermore, the data processing unit (1-4) includes a high-speed storage module (1-4-1), a data transmission communication module (1-4-2), a dedicated data transmission RF front-end, and a data transmission omnidirectional antenna. The high-speed storage module (1-4-1) is connected to the data output end of the core processing unit (1-2) via a high-speed data bus. The baseband interface of the data transmission communication module (1-4-2) is connected to the core processing unit (1-2), and its RF interface is connected to the dedicated data transmission RF front-end. The dedicated data transmission RF front-end includes a data transmission high-power amplifier and a data transmission low-noise amplifier. Both its output and input ends are connected to the data transmission omnidirectional antenna to construct a wireless command and data interaction channel independent of the measurement link.

[0050] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0051] First, the low-altitude channel measurement method and apparatus of the present invention, based on multi-machine dynamic topology and weakly correlated signals, overcomes the limitations of traditional point-to-point measurement systems in terms of spatial coverage. Through the collaborative work of multiple nodes, it can simultaneously acquire measured data of multiple types of links, such as air-to-air, air-to-ground, and ground-to-ground, significantly expanding the measurement range and efficiency, and enabling rapid acquisition of wide-area multi-link channel characteristics in complex low-altitude environments.

[0052] Secondly, the low-altitude channel measurement method and apparatus based on multi-machine dynamic topology and weakly correlated signals of this invention solves the problems of interference and data redundancy in multi-machine parallel measurement. By combining dynamic topology planning and weakly correlated measurement signals, self-interference and mutual interference between nodes are suppressed from both timing control and waveform design aspects. At the same time, by combining hardware parallel processing algorithms, channel features are extracted from three dimensions: time domain, time delay domain, and spatial domain, and redundant data is eliminated in real time. While ensuring the complete expression of channel characteristics, the data storage pressure and subsequent data processing burden are significantly reduced. Attached Figure Description

[0053] Figure 1 This is a structural diagram of the low-altitude channel measurement device based on multi-machine dynamic topology and weak correlation signals according to the present invention.

[0054] Figure 2 The flowchart shows the measurement signal optimization process based on the genetic algorithm.

[0055] Figure 3 Hardware architecture diagram for equivalent fading features extraction;

[0056] Figure 4 A comparison chart showing the actual small-scale fading amplitude and the reconstruction results;

[0057] Figure 5 The diagram shows the channel response and effective multipath extraction results. Detailed Implementation

[0058] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0059] The hardware architecture of this embodiment is as follows: Figure 1 As shown, each node includes a spatiotemporal reference unit (1-1), a core processing unit (1-2), an RF front-end unit (1-3), and a data interaction and storage unit (1-4), which work together to achieve the following steps. Specifically, the spatiotemporal reference unit (1-1) provides a unified, high-precision time synchronization signal and spatial location information for the entire system; the core processing unit (1-2) is responsible for running collaborative topology planning, measurement sequence generation, and feature extraction and analysis algorithms for received signals; the RF front-end unit (1-3) is responsible for high-power transmission amplification and low-noise reception amplification of the measurement signals, and is controlled by the core processing unit to switch between transmission and reception; the data interaction and storage unit (1-4) is responsible for storing massive amounts of measured data and exchanging scheduling commands between multiple nodes. All units are connected via a high-speed bus and RF cables to jointly achieve multi-node collaborative measurement in complex low-altitude environments.

[0060] When the measurement node of this invention is configured as a control station, it serves as the command center of the entire collaborative measurement network. The core processing unit (1-2) first establishes a synchronization time basis based on the unified PPS second pulse signal provided by the spatiotemporal reference unit (1-1), and runs the collaborative topology planning module (1-2-2) to calculate and generate a global measurement scheduling table containing the transmission and reception timing and parallel quantity of each node. Subsequently, the control station transmits the scheduling instructions and synchronization information to the data interaction and storage unit (1-4), which drives the specially configured high-power amplifier of the data transmission link through the data transmission communication module (1-4-2). The dispatch instructions are broadcast to the remaining nodes via the data transmission antenna. At the same time, the control station strictly follows the dispatch table. The core processing unit (1-2) sends a transmit / receive switching control signal to the radio frequency front-end unit (1-3). In the transmit time slot, the high power amplifier link where the measurement transmit antenna is located is activated to radiate the optimized waveform generated by the measurement sequence generation and optimization module (1-2-3). Alternatively, in the receive time slot, the signal is switched to the low noise amplifier link where the measurement receive antenna is located. The small-scale fading extraction module (1-2-1) and the multi-domain real-time redundancy removal module (1-2-4) are used to process and store the echo signal in real time.

[0061] When the measurement node of this invention is configured as a non-control station, it acts as the execution unit of the cooperative network. Its data interaction and storage unit (1-4) receives broadcast commands from the control station through the data transmission antenna. The signal is amplified by a specially configured low-noise amplifier for the data transmission link and then sent to the data transmission communication module (1-4-2) for demodulation and command parsing. After obtaining the parsed global measurement schedule, the core processing unit (1-2) uses the positioning information and clock signal provided by the spatiotemporal reference unit (1-1) for calibration, and precisely controls the operation of the radio frequency front-end unit (1-3) according to the task time slot allocated to this node in the schedule. If the current task is a transmission task, the measurement transmission antenna is selected to send the measurement sequence. If the current task is a reception task, the measurement reception antenna is selected to capture the signal, and the effective channel feature data after multi-domain real-time deduplication and small-scale fading extraction is written into the high-speed storage module (1-4-1), thereby completing the cooperative measurement with the control station and other nodes.

[0062] Furthermore, to address the processing requirements of massive, highly dynamic channel data in low-altitude scenarios, the core processing unit (1-2) in this embodiment employs a parallel signal processing architecture to process the received data. Specifically, the analog baseband signal from the RF front-end unit (1-3) is divided into two paths after analog-to-digital conversion: the first path is input to the small-scale fading extraction module (1-2-1), which does not perform matched filtering but directly performs time-domain statistical analysis on the original signal to obtain the inter-node interference and background noise characteristics of the current measurement link; the second path is input to the multi-domain real-time deredundancy module (1-2-4), which first uses a locally generated pseudo-random sequence to perform sliding correlation with the received signal to extract the channel impulse response, then removes invalid noise components based on a dynamic threshold, and compresses the feature data of each dimension according to spatial correlation. The processed data (small-scale fading characteristics and effective multipath parameters) are written into the high-speed storage module (1-4-1) in the data interaction and storage unit (1-4), thereby significantly reducing the data storage volume and backhaul bandwidth pressure while ensuring the integrity of the measurement data, adapting to the limited payload resources of the UAV platform.

[0063] In this embodiment, the test scenario is set as a low-altitude communication scenario on a campus. The airborne antenna of the multi-UAV channel measurement unit is located 0.3m directly below the GPS positioning module. Due to the obstruction of buildings such as teaching buildings and dormitories in the campus scenario, the A2A, A2G, and G2G multi-link channels exhibit obvious non-stationary characteristics and multipath effects.

[0064] Step 1: Efficient Topology Planning and Intelligent Generation of Measurement Sequences for Multi-UAV Parallel and Serial Collaborative Measurements. This step aims to solve the interference problem in wide-area parallel measurements using optimized algorithms. First, the upper limit of parallel nodes is determined based on preset dynamic range requirements. Then, the transmission timing of each node is planned using a topology algorithm. Finally, a genetic algorithm is used to search for the combination of measurement waveforms with the lowest cross-correlation in a complex pseudo-random sequence space, providing a global scheduling scheme and optimal detection signal for subsequent experimental measurements. The specific implementation process is as follows:

[0065] 1.1) Determine the upper limit of the number of parallel UAVs transmitting signals simultaneously based on the dynamic range threshold. As the number of parallel measurement UAVs increases, the accumulation of interference will lead to a deterioration in the dynamic range of channel measurement results, necessitating the determination of the upper limit based on the dynamic range threshold. Determine the upper limit on the number of drones that can transmit signals simultaneously in parallel. In this embodiment, a dynamic range threshold is set. Combined with formulas (1) and (2), the following calculations are performed. The dynamic range of the channel after interference from a drone was measured to determine the optimal number of parallel transmission nodes.

[0066] (1)

[0067] In the formula, Indicates being received The dynamic range of the measured channel after interference from a drone can be expressed as follows:

[0068] (2)

[0069] In the formula, Indicates the first The measurement sequence and the first The sliding cross-correlation results of the measurement sequences , Indicates the first The results of the sliding autocorrelation of the measurement series.

[0070] 1.2) Based on the calculated parallel upper limit, calculate the lower bound of the number of switching operations, initialize the switching time slots, and generate a cyclic shift set. The system outputs the required number of switching operations and the specific launch timing of each UAV, achieving efficient traversal of the measurement link. The specific steps are as follows:

[0071] 1.2.1) Calculate the lower bound of the number of handovers. ,make ;

[0072] 1.2.2) Let Initialize in each switching time slot ;

[0073] 1.2.3) Using a seed set generate A set of cyclic shifters, from which to select... A set of non-mutually exclusive elements that satisfy... The set is placed greedily, prioritizing the scarcest time slots;

[0074] 1.2.4) If all measurement links are successfully traversed, output the required number of switching operations. And the launch sequence of each drone, otherwise... Repeat steps 1.2.2-1.2.4.

[0075] 1.3) Design different measurement signals for each UAV based on a weakly cross-correlated pseudo-random sequence. To suppress signal interference during parallel measurements by multiple UAVs and ensure the independence of channel measurements, this step follows the approach of "first defining the sequence generation relationship, then finding the optimal parameters," and is implemented through the following three sub-steps. For example... Figure 2 As shown, in order to suppress signal interference during parallel measurements, this embodiment uses a genetic algorithm to search for the optimal combination of measurement sequences that satisfies the low cross-correlation constraint. Figure 2 middle, This indicates the population size in the genetic algorithm. Indicates the first in the initial population Individual, Each individual This represents a set of parameters used to generate a measurement sequence; Indicates the first The fitness function value of each individual; Indicates the first The probability of an individual being selected in a roulette wheel selection method; Indicates the number of relevance evaluation items. Indicates the first One relevance evaluation item; Indicates the first The maximum fitness value in the population. The interval algebra representing the convergence criterion. The above represents the convergence threshold. Figure 2 The symbols in this document are used only to illustrate the genetic algorithm process and should be distinguished from the symbols used in other formulas in this paper to represent the number of measurement nodes, measurement sequence numbers, or measurement node numbers.

[0076] 1.3.1) Based on the measurement sequence length To determine chromosome size, the number of nodes measured is used. Determine the number of chromosomes, and randomly generate chromosomes containing The initial population of individuals.

[0077] 1.3.2) Determine that a weakly cross-correlated pseudo-random sequence is used as the basis for the measurement signal, and define the first... Measurement sequence of a drone The mapping relationship between it and its generated parameters is as follows:

[0078] (3)

[0079] In the formula, Indicates the sequence length. The generator function for the selected pseudo-random sequence. For the first Each UAV corresponds to a set of parameters for sequence generation. At this point, the measurement sequence characteristics of each UAV are entirely determined by its parameter set. Decision made. Subsequently, the generation parameters for each drone were... Chromosomes are encoded using a genetic algorithm. The chromosome size is determined based on the measurement sequence length, and the number of chromosomes is determined based on the number of measurement nodes, thus randomly generating chromosomes containing... The initial population of individuals.

[0080] 1.3.3) Under the constraint of satisfying autocorrelation characteristics, find the optimal parameter set for each UAV. To minimize the peak value of the maximum cross-correlation between all drone pairs, the following min-max optimization problem is constructed:

[0081]

[0082] In the formula, Indicates generation The set of parameters required for a measurement sequence; Indicates the first The measurement sequence and the first The measurement sequence is in time delay The results of the sliding cross-correlation calculation are as follows; This indicates that the cross-correlation amplitude is taken, and the constraint condition ensures that the measured signal has ideal energy efficiency and autocorrelation characteristics.

[0083] 1.3.4) Intelligent search is performed in the candidate pseudo-random sequence space. The specific process is as follows: calculate the individual fitness using formula (2), which is the maximum cross-correlation peak between sequences. The smaller the cross-correlation peak, the higher the fitness. Then, select surviving individuals by using the roulette wheel selection method to screen the parent generation. Subsequently, perform single-point crossover and random mutation operations on the population, that is, exchange or fine-tune the sequence generation parameters to generate a new generation of population. Finally, perform convergence judgment. If the maximum number of iterations or fitness threshold requirements are met, the iteration ends and the optimal combination of measurement sequences that meets the low cross-correlation constraint and the corresponding parameter set are output. .

[0084] Step 2: Real-time Integrated Measurement of Channel and Interference. Utilizing the topology planning and optimization sequence obtained in Step 1, this step is responsible for task allocation for each node, signal transmission, and real-time extraction of channel characteristics. The specific implementation is as follows:

[0085] 2.1) The global scheduling table determined in the first step and the optimal measurement sequence parameter set optimized by the genetic algorithm are distributed to each UAV node through the data transmission link of the data interaction and storage unit (1-4), and a specific measurement sequence and task timing are assigned to each node.

[0086] 2.2) After receiving the instruction, each UAV node generates the corresponding measurement waveform in real time by the core processing unit (1-2) and drives the radio frequency front-end unit (1-3) to transmit the measurement signal into space synchronously or in a time-division manner within the preset transmission time slot.

[0087] 2.3) Receiver node captures superimposed signals in space :

[0088] (5)

[0089] In the formula, Indicates the number of drones. This indicates that the k-th node receives data from the k-th node. The measured signals of each node contain the mutual interference between the target signal and the node. Then, using formula (6), a time-domain sliding correlation operation is performed between the received signal and the locally corresponding known measurement sequence to separate and extract the channel response characteristics of the target link in real time.

[0090] (6)

[0091] In the formula, Indicates the first Measurement signals transmitted by a drone, Indicates the length of the measurement sequence.

[0092] Step 3: Low-Redundancy Feature Extraction of High-Dynamic Measurement Channels. This step aims to remove redundant information from the massive amounts of raw measurement data generated in low-altitude, high-dynamic scenarios through hardware-parallel real-time processing algorithms, considering the time, delay, and spatial domains. The system extracts statistical features of small-scale fading, identifies effective multipath components, and plans sparse sampling points in the spatial domain. This significantly reduces data storage size and backhaul bandwidth pressure while ensuring that key channel features are not lost, thus adapting to the limited payload resources of the UAV platform. For massive measurement data, this step employs a hardware-parallel real-time feature extraction method, including redundancy removal processing in the time, delay, and spatial domains, as detailed below:

[0093] 3.1) Real-time extraction of temporal equivalent fading features. Set the statistical window length. Within this window, the original fading waveform is decomposed into I-path and Q-path, assuming both are independent and identically distributed Gaussian variables. The mean of each path is calculated using an FPGA parallel computing architecture. and ) and variance ( and The system stores only these four statistical feature parameters to replace massive sampling points. The temporal fading envelope can then be reconstructed using these parameters. The equivalent fading feature hardware extraction architecture is as follows: Figure 3 As shown, the comparison between the actual small-scale fading amplitude and the reconstruction results is as follows: Figure 4 As shown.

[0094] 3.2) Real-time extraction of effective multipath features in the time delay domain. A dynamic threshold-based algorithm is used for multipath extraction, and redundancy in the time delay domain information is removed. A hardware parallel computing architecture is used to implement the extraction algorithm, achieving real-time redundancy removal of multipath in the time delay domain during the measurement process.

[0095] 3.3) Real-time planning for sparse airspace sampling. To eliminate spatial redundancy, the system sampling rate is set to 200MHz, and the UAV flight speed is 1m / s. The specific implementation process includes the following four sub-steps:

[0096] 3.3.1) Based on the effective multipath component of the channel impulse response extracted in real time in step 3.2), the current power delay spectrum is calculated using formula (7). The channel response and effective multipath extraction results are as follows: Figure 5 As shown:

[0097] (7)

[0098] In the formula, express The spatial three-dimensional coordinates corresponding to the movement trajectory of the drone at any time. Indicates the effective multipath component.

[0099] 3.3.2) To quantify the redundancy of spatial location, formula (8) is used based on the current location. Position after displacement The spatial correlation coefficient of the power delay spectrum at a given location is calculated. Power delay spectrum correlation coefficient. Represented as:

[0100] (8)

[0101] This step quantifies the redundancy of spatial location by calculating the correlation between the power delay spectrum of the current location and the location after displacement.

[0102] 3.3.3) Set the correlation coefficient threshold Constraints are imposed on the sparse sampling points in spatial location. Considering the spatial correlation attenuation characteristics of low-altitude channels, this embodiment sets the threshold to [value missing]. The channel correlation at the current displacement is determined in real time using formula (9):

[0103] (9)

[0104] And update This process is repeated to determine the next sparse sampling point. The correlation coefficient is then calculated. When this occurs, it indicates that significant decorrelation has occurred in the channel spatial features. The system determines this location as a new valid sampling point and updates the reference location. This process utilizes a parallel computing architecture shared with the sliding correlation module on an FPGA, with the time required for a single correlation coefficient calculation controlled within 10μs, meeting the requirements for real-time planning.

[0105] 3.3.4) Use the updated reference position from step 3.3.3) as the input for the next round of calculation, and repeat steps 3.3.1) to 3.3.3) until the UAV completes the coverage of the current channel slice's preset trajectory, thereby completing the real-time planning of sparse airspace sampling for this area.

[0106] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0107] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A low-altitude channel measurement method based on multi-machine dynamic topology and weakly correlated signals, characterized in that, The method includes: S1. For low-altitude, high-dynamic scenarios involving parallel launches of multiple UAVs and mutual interference between multiple nodes, a dynamic range threshold is preset. Based on the dynamic range threshold, the upper limit of parallel nodes is determined and the launch sequence of each node is planned to avoid temporal interference. A candidate pseudo-random sequence space is constructed based on a weakly cross-correlated pseudo-random sequence. A genetic algorithm is used to search for the optimal measurement sequence with the lowest cross-correlation and the corresponding parameter set in the candidate pseudo-random sequence space to suppress waveform domain interference. Through the coordination of launch sequence and optimal measurement sequence, a global scheduling scheme for multi-UAV measurement is formed. S2, according to the global scheduling scheme planned in step S1, the actual measurement task is executed. Combined with the optimal measurement sequence searched in step S1, the received superimposed signal is processed in real time using time-domain sliding correlation operation to separate the target signal and suppress mutual interference between nodes in order to obtain multipath channel characteristics and obtain the original measurement data. S3. For the raw measurement data generated in step S2, a hardware parallel real-time processing algorithm is used to remove redundant information from three dimensions: time domain, time delay domain, and spatial domain. Small-scale fading statistical features are extracted, effective multipath components are identified, and spatial sparse sampling is completed. This reduces the data storage and backhaul bandwidth requirements while preserving key channel features.

2. The low-altitude channel measurement method based on multi-machine dynamic topology and weakly correlated signals according to claim 1, characterized in that, Step S1, the process of determining the upper limit of parallel nodes based on the dynamic range threshold, includes the following steps: To address the channel measurement accuracy requirements, hardware system performance, and scene interference characteristics in low-altitude, high-dynamic scenarios involving parallel launches of multiple UAVs and mutual interference between multiple nodes, a preset dynamic range threshold is established. According to the preset dynamic range threshold Determine the upper limit on the number of drones that can transmit signals simultaneously in parallel: In the formula, Indicates being received The dynamic range of the measured channel after interference from a single UAV is determined based on the background benchmark of multi-UAV mutual interference accumulation and the peak characteristics obtained by sequence correlation calculation, and is expressed as: In the formula, Indicates the first The measurement sequence and the first The sliding cross-correlation results of the measurement sequences , Indicates the first The results of the sliding autocorrelation of the measurement series.

3. The low-altitude channel measurement method based on multi-machine dynamic topology and weakly correlated signals according to claim 1, characterized in that, In step S1, the planning process for the transmission timing of each node includes the following steps: A1, Calculate the lower bound of the number of switching operations. ,make In the formula, This represents the total number of UAV nodes participating in the cooperative channel measurement. This represents the number of UAV nodes simultaneously in the launch state within each switching time slot, i.e., the upper limit of the number of parallel launch nodes. ; A2, let Initialize in each switching time slot ; A3, using seed set generate A set of cyclic shifters, from which to select... A set of non-mutually exclusive elements that satisfy... The set is placed greedily, prioritizing the scarcest time slots; A4. If all measurement links are successfully traversed, output the required number of switching operations. And the launch sequence of each drone, end the process, otherwise... Return to step A1.

4. The low-altitude channel measurement method based on multi-machine dynamic topology and weakly correlated signals according to claim 1, characterized in that, In step S1, the process of obtaining the optimal measurement sequence includes the following steps: B1, using a weakly cross-correlated pseudo-random sequence as the basis of the measurement signal, defines the first... Measurement sequence of a drone The mapping relationship between it and its generated parameters: In the formula, Indicates the sequence length. The generator function for the selected pseudo-random sequence; For the first The set of sequence generation parameters for each UAV includes at least the measurement sequence length and the number of measurement nodes; Chromosome size is determined based on sequence length, and chromosome number is determined based on number of measurement nodes. The generation parameters for each UAV are then used. Chromosomes encoded as genetic algorithms are randomly generated and contain... The initial population of individuals; B2, under the constraint of satisfying autocorrelation characteristics, find the optimal parameter set for each UAV. To minimize the maximum cross-correlation peak among all drone pairs, a minimax optimization problem is constructed: In the formula, Indicates generation The set of parameters required for a measurement sequence; Indicates the first The measurement sequence and the first The measurement sequence is in time delay The results of the sliding cross-correlation calculation are as follows; This indicates the cross-correlation amplitude; B3 involves an intelligent search within the candidate pseudo-random sequence space, the specific process of which is as follows: A set of measurement sequences is treated as an individual, and its fitness is evaluated using the maximum cross-correlation peak among the sequences. The smaller the cross-correlation peak, the higher the fitness. Survival selection is performed using a roulette wheel selection method to choose parent individuals. Single-point crossover and random mutation operations are performed on the population to update the sequence generation parameters and produce a new generation. This process is repeated iteratively until the maximum number of iterations or the fitness threshold is met, outputting the optimal combination of measurement sequences that satisfies the low cross-correlation constraint and its corresponding parameter set. .

5. The low-altitude channel measurement method based on multi-machine dynamic topology and weakly correlated signals according to claim 1, characterized in that, Step S3 further includes the following steps: C1, Real-time extraction of time-domain equivalent fading features: Based on the original measurement data obtained in step S2, in a length of... Within the statistical window, the orthogonal fading waveform is denoted as and ,in, and represent the continuous sampled values ​​of the in-phase component of the quadrature fading waveform within the statistical window. This represents the continuous sampled values ​​of the orthogonal components of the orthogonal fading waveform within the statistical window. This indicates the number of sampling points contained within the statistical window; a hardware parallel computing architecture is used to calculate the mean and variance of the orthogonal fading waveform as unbiased estimates. , ,in and These are the mean and standard deviation, respectively. C2, Real-time extraction of effective multipath features in the time delay domain: For the original measurement data obtained in step S2, the effective multipath component extraction algorithm based on dynamic threshold is used to remove redundancy from the time delay domain information, and real-time redundancy removal of multipath in the time delay domain is completed to extract the effective multipath components of the channel impulse response. C3: Real-time planning for sparse spatial sampling, which includes the following three steps: C31: Calculate the power delay spectrum based on the effective multipath components of the channel impulse response extracted in real time in step C2: In the formula, express The spatial three-dimensional coordinates corresponding to the movement trajectory of the drone at any time. Indicates the effective multipath component. Indicates time delay; C32: Based on the correlation coefficient of the power delay spectrum, the spatial sparsity boundary of the channel slice is determined in real time to achieve redundancy removal of spatial sampling locations and power delay spectrum correlation coefficient. Represented as: in, This represents the spatial displacement relative to the reference spatial position. Indicates offset along the preset measurement trajectory The candidate sampling positions obtained afterwards Indicates the reference spatial position In time delay The power delay spectrum below; C33: Using correlation coefficient threshold Constraining the sparse sampling points in spatial location, the sampling points that satisfy the conditions are represented as: And update This is used to determine the next sparse sampling point; C34: Using the updated reference position as the input for the next round of calculation, repeat steps C31 to C33 until all preset spatial nodes of the current channel slice have been traversed, thus completing the real-time planning of spatial sparse sampling for the current region.

6. A low-altitude channel measurement device based on multi-machine dynamic topology and weakly correlated signals, characterized in that, The device includes a spatiotemporal reference unit (1-1), a core processing unit (1-2), a radio frequency front-end unit (1-3), and a data interaction and storage unit (1-4). The spatiotemporal reference unit (1-1) is used to provide the device with a unified high-precision time synchronization signal and spatial location information, and transmit them to the core processing unit (1-2); the core processing unit (1-2) is used to generate measurement sequences according to cooperative topology planning, control the radio frequency front-end unit (1-3) to generate and transmit measurement signals, and at the same time, the core processing unit (1-2) is also used to perform multi-domain real-time redundancy removal processing and small-scale fading characteristic analysis on the measurement signals received by the radio frequency front-end unit (1-3) after transmission through the wireless channel, and extract effective channel feature data; The radio frequency front-end unit (1-3) is used to amplify and transmit the measurement signal with high power, and to amplify the received signal with low noise before transmitting it to the core processing unit (1-2); the data interaction and storage unit (1-4) is used to store the channel characteristic data output by the core processing unit (1-2), and to perform scheduling command interaction between multiple nodes through the data transmission link.

7. The low-altitude channel measurement device based on multi-machine dynamic topology and weakly correlated signals according to claim 6, characterized in that, The core processing unit (1-2) adopts a parallel processing architecture and integrates a collaborative topology planning module (1-2-2), a measurement sequence generation module (1-2-3), a small-scale fading extraction module (1-2-1), and a multi-domain real-time deduplication module (1-2-4). The collaborative topology planning module (1-2-2) is connected to the measurement sequence generation module (1-2-3), and the output of the measurement sequence generation module (1-2-3) is connected to the transmit signal input of the RF front-end unit (1-3). The receive signal output of the RF front-end unit (1-3) is connected to a parallel splitting node, which is connected to the input of the small-scale fading extraction module (1-2-1) and the input of the multi-domain real-time deduplication module (1-2-4) to realize parallel splitting processing of the received signal.

8. The low-altitude channel measurement device based on multi-machine dynamic topology and weak correlation signals according to claim 6, characterized in that, The radio frequency front-end unit (1-3) includes a measurement high-power amplifier, a measurement low-noise amplifier, a measurement transmitting antenna, and a measurement receiving antenna; the input terminal of the measurement high-power amplifier is connected to the analog signal output port of the core processing unit (1-2), and its output terminal is connected to the measurement transmitting antenna; the measurement receiving antenna is connected to the input terminal of the measurement low-noise amplifier, and the output terminal of the measurement low-noise amplifier is connected to the analog signal input port of the core processing unit (1-2); the control interface of the core processing unit (1-2) is connected to the enable terminal of the amplifier or the radio frequency switch to control the physical on / off state of the transceiver link.

9. The low-altitude channel measurement device based on multi-machine dynamic topology and weak correlation signals according to claim 6, characterized in that, The data processing unit (1-4) includes a high-speed storage module (1-4-1), a data transmission communication module (1-4-2), a dedicated data transmission RF front-end, and a data transmission omnidirectional antenna. The high-speed storage module (1-4-1) is connected to the data output end of the core processing unit (1-2) via a high-speed data bus. The baseband interface of the data transmission communication module (1-4-2) is connected to the core processing unit (1-2), and its RF interface is connected to the dedicated data transmission RF front-end. The dedicated data transmission RF front-end includes a data transmission high-power amplifier and a data transmission low-noise amplifier. Its output and input ends are both connected to the data transmission omnidirectional antenna to construct a wireless command and data interaction channel independent of the measurement link.

Images