A digital radar distributed weak signal target detection method and system

Abstract

The application discloses a kind of digital radar distributed weak signal target detection method and system, it is related to radar detection technical field, the method includes: the signal time delay of a plurality of distributed radar nodes is acquired, and the received signal-to-noise ratio is acquired;Based on the signal time delay and the received signal-to-noise ratio, the optimized coherence time is calculated and acquired, and the signal timing of each distributed radar node is corrected, and weak signal target is coherently accumulated;Based on the received signal-to-noise ratio and in combination with preset false alarm probability, weak signal detection threshold is acquired, and node weight is acquired based on signal time delay and the received signal-to-noise ratio, the detection result of each distributed radar node is fused, and weak signal target detection result is acquired.The application solves the technical problem that the effect of weak signal target in the prior art is poor in the cooperative detection of distributed radar.

Description

Technical Field

[0001] This invention relates to the field of radar detection technology, specifically to a digital radar distributed weak signal target detection method and system. Background Technology

[0002] In modern radar detection systems, distributed network detection has become a core technology for improving target perception capabilities in complex environments, especially for weak signal echo scenarios such as stealth targets and small, long-range targets. Multi-node collaborative processing can effectively utilize the advantages of spatial diversity and energy accumulation. However, existing distributed radar weak signal detection methods face multiple technical bottlenecks in practical applications. First, significant incoherent time delays exist when each radar node receives the echo of the same target, resulting in a severe decrease in accumulation gain or even cancellation when directly performing signal-level coherent accumulation. Second, the signal-to-noise ratio (SNR) of weak signal target echoes is extremely low. Traditional detection strategies based on fixed thresholds or single-node statistical characteristics struggle to balance low false alarm probabilities with high detection probabilities, and the differences in the dynamic range of SNR among different nodes further increase the difficulty of uniformly setting the global detection threshold. In addition, the fusion decision of multi-node detection results often adopts simple voting or equal-weighted merging, failing to fully consider the differentiated reliability factors such as signal delay accuracy and SNR quality of each node. This results in the fusion detection performance not effectively improving with the increase in the number of nodes, leading to low resource utilization. Summary of the Invention

[0003] This application provides a digital radar distributed weak signal target detection method and system to address the technical problem of poor performance in the cooperative detection of weak signal targets by distributed radar in the prior art.

[0004] In view of the above problems, this application provides a digital radar distributed weak signal target detection method and system.

[0005] In a first aspect, this application provides a digital radar distributed weak signal target detection method, the method comprising: Obtain the signal delay of multiple distributed radar nodes and the received signal-to-noise ratio; Based on the signal delay and the received signal-to-noise ratio, the optimized coherence time is calculated and obtained, and the signal timing of each distributed radar node is corrected to perform coherent accumulation on weak signal targets. Based on the received signal-to-noise ratio and a preset false alarm probability, a weak signal detection threshold is obtained. Node weights are then obtained based on signal delay and the received signal-to-noise ratio. The detection results of each distributed radar node are fused to obtain a weak signal target detection result.

[0006] Secondly, this application provides a digital radar distributed weak signal target detection system, comprising: The information acquisition module is used to acquire the signal delay of multiple distributed radar nodes and to acquire the received signal-to-noise ratio; The coherent accumulation module is used to calculate and obtain the optimized coherent time based on the signal delay and the received signal-to-noise ratio, and to correct the signal timing of each distributed radar node, and to coherently accumulate weak signal targets. The result acquisition module is used to obtain a weak signal detection threshold based on the received signal-to-noise ratio and a preset false alarm probability, and to obtain node weights based on signal delay and the received signal-to-noise ratio, and to fuse the detection results of each distributed radar node to obtain a weak signal target detection result.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application proposes a distributed weak signal target detection method and system for digital radar. By constructing a full-process collaborative processing mechanism from node timing correction and coherent accumulation optimization to adaptive configuration of detection threshold and weighted fusion of detection results, the detection capability of the distributed radar system for weak signal targets is significantly improved. First, the signal delay and received signal-to-noise ratio of each distributed radar node are obtained. Based on this, the node with the optimal signal-to-noise ratio is selected as the benchmark. By calculating the delay deviation coefficient and signal-to-noise ratio deviation coefficient between each node and the benchmark node, the coherent accumulation time window is adaptively optimized, and the signal timing of each node is accurately corrected, so that the echoes of multiple nodes are effectively aligned in the time domain. Then, linear superposition accumulation is performed, which significantly improves the coherent accumulation efficiency and signal energy enhancement of weak signal targets. Building upon this foundation, this method, for each distributed radar node, combines its received signal-to-noise ratio (SNR) with a preset false alarm probability. Through machine learning, it constructs and trains a convergent detection threshold configuration model, enabling dynamic and precise configuration of detection thresholds under different SNR conditions. This effectively overcomes the problems of insufficient detection sensitivity or excessively high false alarm rates in weak signal scenarios caused by traditional fixed thresholds, significantly improving the accuracy and stability of independent detection by each node. Furthermore, after node-level detection is completed, the delay deviation coefficient and SNR deviation coefficient of each node are comprehensively evaluated based on signal delay and received SNR. A weighted calculation is then performed to obtain the node credibility, generating node weights. Weighted fusion decisions are then made based on the detection results of each node, and the average credibility of multiple nodes is calculated and output as the final credibility. Compared to traditional fusion strategies using equal-weighted merging or simple voting mechanisms, the technical solution provided in this application significantly optimizes the rationality of fusion decisions, allowing nodes with high SNR and more accurate delay estimation to contribute more to the final decision, thereby effectively improving the overall accuracy and robustness of fusion detection. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a flowchart illustrating a distributed weak signal target detection method for digital radar provided in an embodiment of this application.

[0010] Figure 2 This is a schematic diagram of the structure of a digital radar distributed weak signal target detection system provided in an embodiment of this application.

[0011] The components represented by each number in the attached diagram are explained below: Information acquisition module 100, coherence accumulation module 200, result acquisition module 300. Detailed Implementation

[0012] This application provides a digital radar distributed weak signal target detection method and system to address the technical problem of poor performance in the cooperative detection of weak signal targets by distributed radar in the prior art.

[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0014] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0015] Example 1, as Figure 1 As shown, this application provides a digital radar distributed weak signal target detection method, wherein the method includes: S10: Obtain the signal delay of multiple distributed radar nodes and obtain the received signal-to-noise ratio.

[0016] When distributed radars are used to detect weak signal targets, the timing of receiving the echo from the same target can be significantly asynchronous due to differences in geographical location, clock reference, and signal propagation path among the nodes.

[0017] Step S10 in the method provided in this application embodiment includes: Multiple distributed radar nodes are acquired, and the time difference between the multiple distributed radar nodes receiving the echo of the same weak signal target is acquired as the signal delay. Calculate the ratio of target signal power to background noise power of the multiple distributed radar nodes to obtain the received signal-to-noise ratio.

[0018] In this embodiment of the application, the signal delay of multiple distributed radar nodes is obtained, and the received signal-to-noise ratio is obtained.

[0019] Specifically, firstly, multiple distributed radar nodes are acquired, and the time difference between their reception of the same weak-signal target echo is obtained as the signal delay. For example, each distributed radar node is deployed at a receiving site in a different geographical location, and each node is equipped with a broadband digital receiver and a high-precision timing module. When a weak-signal target enters the detection airspace, each node independently receives a single reflected echo signal from the target. The raw echo data received by each node are then collected and processed by a central processing unit. The central processing unit calls a time delay estimation processor based on generalized cross-correlation to perform pairwise comparisons of the echo signals from each node. Specifically, any node is selected as a reference node, and its echo signal and the echo signal from another node are input into the cross-correlation processor. A cross-correlation function is calculated through sliding correlation, and the peak position of this function corresponds to the time difference between the two nodes receiving the echo of the same target. This process is repeated for all nodes to obtain the signal delay value of each node based on the reference node. For example, the reference node is denoted as node 1, and the cross-correlation peak between node 1 and node 2 is located at the 152nd sampling point. Combined with the system sampling interval, the signal delay of node 2 relative to node 1 can be calculated to be several nanoseconds.

[0020] Further, the ratio of target signal power to background noise power of the multiple distributed radar nodes is calculated to obtain the received signal-to-noise ratio (SNR). For example, the digital receiver of each node simultaneously acquires environmental noise data during periods without a target while acquiring the target echo signal. For each node, the time-domain waveform of the range cell containing the target signal is extracted from the echo data, and its root mean square amplitude is calculated as the target signal power; a sampling sequence of the same length is extracted from the noise data during periods without a target, and its root mean square amplitude is calculated as the background noise power. The ratio of target signal power to background noise power is converted to decibels to obtain the received SNR of that node. The same operation is performed on all nodes to obtain the received SNR value for each node, which is stored together with the signal delay vector to form the complete input dataset for subsequent coherent accumulation and fusion detection.

[0021] By obtaining the signal delay of multiple distributed radar nodes receiving the echo of the same weak target, and calculating the ratio of target signal power to background noise power at each node, the received signal-to-noise ratio is obtained, providing an accurate and quantitative representation of the node state for the entire distributed collaborative detection system.

[0022] S20: Based on the signal delay and the received signal-to-noise ratio, calculate and obtain the optimized coherence time, correct the signal timing of each distributed radar node, and perform coherent accumulation on weak signal targets.

[0023] Distributed radar nodes performing signal-level coherent accumulation without precise synchronization will experience signal phase mismatch due to differences in echo delay between nodes. This not only fails to achieve effective energy superposition but may also cause destructive interference, resulting in a significant decrease in accumulation gain, which may even fall below the detection level of a single node.

[0024] Step S20 in the method provided in this application embodiment includes: Among the multiple distributed radar nodes, the node with the highest received signal-to-noise ratio is selected as the reference node; The absolute difference between the signal delay between each of the distributed radar nodes and the reference node is calculated as the ratio of the signal delay between the standard nodes, and this ratio is used as the synchronization deviation coefficient of that node. Calculate the ratio of the received signal-to-noise ratio of each of the distributed radar nodes to that of the reference node, and use it as the signal-to-noise ratio deviation coefficient; Based on the synchronization deviation coefficient and the signal-to-noise ratio deviation coefficient, a weighted calculation is performed to obtain the coherence time optimization coefficient, and the basic coherence time is optimized to obtain the optimized coherence time. Based on the signal delay, the signal reception timing of the remaining distributed radar nodes is adjusted so that the weak signal target echoes received by the remaining nodes are aligned with the weak signal target echo timing of the reference node. This includes correcting the timing of signals from each distributed radar node and performing coherent accumulation on weak signal targets, including: The signal reception timing of the reference node is used as the reference timing. Based on the signal delay, the signal reception timing of the remaining distributed radar nodes is adjusted so that the weak signal target echoes received by the remaining nodes are aligned with the weak signal target echo timing of the reference node. The optimized coherence time is used to linearly superimpose and accumulate the weak signal target echoes received by each distributed radar node after timing correction.

[0025] In this embodiment, firstly, the node with the highest received signal-to-noise ratio (SNR) among the distributed radar nodes is selected as the reference node. For example, the list of SNR values ​​of each distributed radar node obtained in step S10 is compiled. The SNR values ​​of all nodes are traversed, and the node with the largest value is selected and designated as the reference node for this collaborative processing. For example, if the SNR of node 2 is higher than that of other nodes such as node 1 and node 3, then node 2 is selected as the reference node.

[0026] Further, the ratio of the absolute difference in signal delay between each of the distributed radar nodes and the reference node to the signal delay between standard nodes is calculated as the synchronization deviation coefficient of that node. For example, the signal delay value of each node is obtained with a unified reference point as the reference, and the absolute value of the signal delay difference between each non-reference node and the reference node is calculated accordingly. The system-preset standard inter-node signal delay is obtained; this standard value represents the theoretical delay difference between two nodes receiving the echo from the same target under an ideal geometric layout. The delay difference of each non-reference node is divided by this standard delay value, and the resulting ratio is the synchronization deviation coefficient of that node. This coefficient reflects the degree of deviation of the node's actual synchronization error from the ideal state.

[0027] Further, the ratio of the received signal-to-noise ratio (SNR) of each of the distributed radar nodes to that of the reference node is calculated as the SNR deviation coefficient. For example, the SNR value of each non-reference node is divided by the SNR value of the reference node to obtain the SNR deviation coefficient for that node. This coefficient characterizes the relative level of the node's signal quality compared to the optimal node.

[0028] Further, a weighted calculation is performed based on the synchronization deviation coefficient and the signal-to-noise ratio deviation coefficient to obtain the coherence time optimization coefficient. This optimizes the basic coherence time, yielding the optimized coherence time. For example, a preset weight factor is assigned to both the synchronization deviation coefficient and the signal-to-noise ratio deviation coefficient, with the sum of the two weights being 1. For instance, the weights of both the synchronization deviation coefficient and the signal-to-noise ratio deviation coefficient can be set to 0.5. In practical applications, this can be adjusted based on the specific scenario. The synchronization deviation coefficient of each non-baseline node is multiplied by its corresponding weight, and the signal-to-noise ratio deviation coefficient is multiplied by its corresponding weight. These two products are then added together to obtain the coherence time optimization coefficient for that node. Multiplying this by the basic coherence time yields the optimized coherence time.

[0029] Furthermore, based on the signal delay, the signal reception timing of the remaining distributed radar nodes is adjusted so that the weak signal target echoes received by the remaining nodes are aligned with the weak signal target echo timing of the reference node.

[0030] First, the signal reception timing of the reference node is used as the reference timing.

[0031] Furthermore, based on the signal delay, the signal reception timing of the remaining distributed radar nodes is adjusted so that the weak signal target echoes received by the remaining nodes are aligned with the weak signal target echo timing of the reference node. For example, based on the signal delay difference between each non-reference node and the reference node, a time-shift operation is performed on the digital echo sequence of that node. The time-shift operation can be achieved through digital delay lines or frequency domain phase compensation, for example, shifting the entire echo sequence forward or backward by a corresponding number of sampling points in the digital domain, so that the arrival time of the target signal in the echo of that node is completely aligned with the arrival time of the same target signal in the echo of the reference node.

[0032] Furthermore, using the optimized coherence time, the weak signal target echoes received by each distributed radar node after time-series correction are linearly superimposed and accumulated. For example, the complex baseband echo data of all distributed radar nodes after time-series correction are linearly superimposed point-by-point within a time window of the optimized coherence time length. The superimposed complex signal sequence is the coherent accumulation output, whose signal-to-noise ratio is significantly improved compared to single-node echoes, and can be directly used for subsequent target detection and decision processing.

[0033] By selecting the node with the highest received signal-to-noise ratio (SNR) as the benchmark, the delay deviation coefficient and SNR deviation coefficient between each node and the benchmark node are calculated. These are then weighted and fused to generate coherence time optimization coefficients. The basic coherence time is adaptively optimized to obtain the optimal coherence time that best matches the current node state. Simultaneously, using the benchmark node's timing as a reference, the echoes from each node are precisely time-adjusted based on the signal delay, effectively aligning the weak signals from multiple nodes in the time domain. Then, linear superposition and accumulation using the optimized coherence time effectively suppresses energy loss caused by delay mismatch, enabling the weak signal target to achieve near-theoretical optimal accumulation gain under multi-node collaboration.

[0034] S30: Based on the received signal-to-noise ratio and combined with the preset false alarm probability, obtain the weak signal detection threshold, and based on the signal delay and the received signal-to-noise ratio, obtain the node weight, fuse the detection results of each distributed radar node, and obtain the weak signal target detection result.

[0035] Existing methods often use fixed detection thresholds, which make it difficult to balance low false alarm rates and high detection probabilities in scenarios with dynamically changing signal-to-noise ratios. Multi-node fusion strategies often use simple equal-weighted voting, which does not consider the significant differences in delay estimation accuracy and signal quality among different nodes, resulting in low-confidence nodes causing undue interference to the final results.

[0036] Step S30 in the method provided in this application embodiment includes: For each distributed radar node, obtain the false alarm probability; The false alarm probability is obtained based on the historical average false alarm probability of the distributed radar nodes. The false alarm probability and received signal-to-noise ratio of each distributed radar node are input into the detection threshold configuration model to obtain the weak signal detection threshold. The construction of the detection threshold configuration model includes: Acquire multiple sets of sample data, each set of sample data including received signal-to-noise ratio samples, preset false alarm probability samples, and weak signal detection thresholds for the corresponding samples; A detection threshold configuration model is constructed based on machine learning. The detection threshold configuration model is trained using the sample data until it converges, thus completing the construction of the detection threshold configuration model. Calculate the ratio of the signal delay of each of the distributed radar nodes to that of the reference node, and use it as the delay deviation coefficient of that node; By combining the time delay deviation coefficient and the signal-to-noise ratio deviation coefficient, a weighted calculation is performed to obtain the node credibility. Based on node credibility, node weights are calculated and obtained, and the detection results of each distributed radar node are fused to obtain the target detection result. The average confidence level of multiple distributed radar nodes is calculated as the result confidence level. This result is then combined with the target detection result to output the weak signal target detection result.

[0037] In this embodiment, a weak signal detection threshold is obtained based on the received signal-to-noise ratio and a preset false alarm probability. Node weights are obtained based on signal delay and the received signal-to-noise ratio. The detection results of each distributed radar node are then fused to obtain a weak signal target detection result.

[0038] Specifically, firstly, for each distributed radar node, the false alarm probability is obtained. This false alarm probability is based on the historical average false alarm probability of the distributed radar node. For example, the statistical frequency of false alarms in each target detection event recorded by the node over a past operating period is read, its historical average false alarm probability is calculated, and this average value is used as the false alarm probability of the node in the current detection period. For instance, the historical average false alarm probability of node 1 is 0.1%, and that of node 2 is 0.3%, and each node can be configured independently.

[0039] Furthermore, the false alarm probability and received signal-to-noise ratio of each distributed radar node are input into the detection threshold configuration model to obtain the weak signal detection threshold.

[0040] The construction of the detection threshold configuration model includes: First, multiple sets of sample data are acquired. Each set of sample data includes a received signal-to-noise ratio sample, a preset false alarm probability sample, and a weak signal detection threshold for the corresponding sample. For example, a large amount of sample data is collected from radar test ranges and historical mission data. Each set of samples contains three elements: the received signal-to-noise ratio value, the false alarm probability value, and the optimal detection threshold obtained through theoretical calculation or actual measurement calibration under that condition.

[0041] Furthermore, a detection threshold configuration model is constructed based on machine learning. For example, a three-layer fully connected neural network is built using the TensorFlow framework as the detection threshold configuration model. The first layer is the input layer with 2 nodes, corresponding to the two input features: received signal-to-noise ratio and false alarm probability. The second layer is the hidden layer, containing 64 neurons, using ReLU activation function, and a Dropout layer with a dropout rate of 0.2 is added to prevent overfitting. The third layer is the output layer with 1 node, using linear activation function, and outputting the predicted detection threshold.

[0042] Furthermore, the detection threshold configuration model is trained using the sample data until convergence, thus completing the construction of the detection threshold configuration model. For example, the sample dataset is divided into a training set and a validation set in an 8:2 ratio. The Adam optimizer is used for iterative training with mean squared error as the loss function. Training is considered converged when the validation set loss no longer decreases for several consecutive cycles, at which point training is stopped, and the model weights are saved.

[0043] Furthermore, the current received signal-to-noise ratio and false alarm probability of each node are concatenated into a two-dimensional feature vector, which is then input into the model for forward inference. The output of the model is the weak signal detection threshold of the current frame for that node.

[0044] Further, the ratio of the signal delay of each of the distributed radar nodes to that of the reference node is calculated as the delay deviation coefficient of that node. The delay deviation coefficient = current node signal delay / (reference node signal delay + 0.1). This coefficient reflects the degree of relative deviation between the node and the reference node in the time synchronization dimension. Here, 0.1 is a small real number set to prevent calculation errors when the reference node signal delay is 0; the specific value can be set based on the scenario.

[0045] Furthermore, a weighted calculation is performed by combining the delay deviation coefficient and the signal-to-noise ratio deviation coefficient to obtain the node credibility. For example, preset weight factors are assigned to both the delay deviation coefficient and the signal-to-noise ratio deviation coefficient, with the sum of the two weights being 1. In practical applications, the weight factors can be set based on the specific environment. The node credibility is obtained by multiplying the delay deviation coefficient of each node by its corresponding weight and the signal-to-noise ratio deviation coefficient by its corresponding weight, and then adding these two products. A higher node credibility indicates better signal quality and more accurate delay estimation during the current detection.

[0046] Furthermore, based on node credibility, node weights are calculated and obtained, and the detection results of each distributed radar node are fused to obtain the target detection result. For example, the credibility of all nodes is normalized: the credibility of each node is divided by the sum of the credibility of all nodes to obtain the node weight. The node weight reflects the proportion of contribution of the node's detection result to the final fusion decision. After obtaining its own detection threshold, each node performs a threshold comparison on its coherently accumulated signal and outputs the detection result. The central processing unit collects the detection results of all nodes and performs weighted voting according to the above node weights. If the sum of the detection results of all nodes multiplied by the corresponding weight is greater than the preset fusion threshold, the target is determined to exist; otherwise, the target is determined not to exist. After weighted fusion, the target detection result of this detection is output, including the presence / absence and the target position parameters obtained by weighted average.

[0047] Furthermore, the average confidence level of multiple distributed radar nodes is calculated as the result confidence level. This result is then combined with the target detection result for output to obtain the weak signal target detection result. For example, the node confidence levels of all nodes are summed and divided by the total number of nodes to obtain the result confidence level. This confidence level value serves as the confidence level label for this detection conclusion and is encapsulated and output along with the target detection result to form a complete weak signal target detection result. For example, the output format is "Target exists, distance 25 km, bearing 60 degrees, result confidence level 0.87".

[0048] For each distributed radar node, by combining its received signal-to-noise ratio (SNR) and false alarm probability, a convergent detection threshold configuration model is trained to dynamically obtain a weak signal detection threshold adapted to the current signal state, achieving precise control of detection sensitivity. Simultaneously, based on signal delay and received SNR, the delay deviation coefficient and SNR deviation coefficient of each node are comprehensively evaluated. Node credibility is weighted and calculated to generate fusion weights. Weighted decisions are made on the detection results of each node, and the average credibility of multiple nodes is calculated and output as the final credibility. This step significantly improves the adaptability and rationality of the detection threshold setting, while ensuring that the fusion decision fully reflects the quality differences of each node, effectively suppressing the negative contribution of low-quality nodes, and significantly enhancing the overall reliability, accuracy, and interpretability of the distributed radar system for detecting weak signal targets.

[0049] Example 2, as Figure 2 As shown, based on the same inventive concept as the distributed weak signal target detection method for digital radar provided in Embodiment 1, this embodiment of the invention also provides a distributed weak signal target detection system for digital radar, comprising: The information acquisition module 100 is used to acquire the signal delay of multiple distributed radar nodes and to acquire the received signal-to-noise ratio; The coherent accumulation module 200 is used to calculate and obtain the optimized coherent time based on the signal delay and the received signal-to-noise ratio, and to correct the signal timing of each distributed radar node, and to perform coherent accumulation on weak signal targets. The result acquisition module 300 is used to obtain a weak signal detection threshold based on the received signal-to-noise ratio and a preset false alarm probability, and to obtain node weights based on signal delay and the received signal-to-noise ratio, and to fuse the detection results of each distributed radar node to obtain a weak signal target detection result.

[0050] In one embodiment, the information acquisition module 100 is further configured to: Multiple distributed radar nodes are acquired, and the time difference between the multiple distributed radar nodes receiving the echo of the same weak signal target is acquired as the signal delay. Calculate the ratio of target signal power to background noise power of the multiple distributed radar nodes to obtain the received signal-to-noise ratio.

[0051] In one embodiment, the coherence accumulation module 200 is further configured to: Among the multiple distributed radar nodes, the node with the highest received signal-to-noise ratio is selected as the reference node; The absolute difference between the signal delay between each of the distributed radar nodes and the reference node is calculated as the ratio of the signal delay between the standard nodes, and this ratio is used as the synchronization deviation coefficient of that node. Calculate the ratio of the received signal-to-noise ratio of each of the distributed radar nodes to that of the reference node, and use it as the signal-to-noise ratio deviation coefficient; Based on the synchronization deviation coefficient and the signal-to-noise ratio deviation coefficient, a weighted calculation is performed to obtain the coherence time optimization coefficient, and the basic coherence time is optimized to obtain the optimized coherence time. Based on the signal delay, the signal reception timing of the remaining distributed radar nodes is adjusted so that the weak signal target echoes received by the remaining nodes are aligned with the weak signal target echo timing of the reference node. This includes correcting the timing of signals from each distributed radar node and performing coherent accumulation on weak signal targets, including: The signal reception timing of the reference node is used as the reference timing. Based on the signal delay, the signal reception timing of the remaining distributed radar nodes is adjusted so that the weak signal target echoes received by the remaining nodes are aligned with the weak signal target echo timing of the reference node. The optimized coherence time is used to linearly superimpose and accumulate the weak signal target echoes received by each distributed radar node after timing correction.

[0052] In one embodiment, the result acquisition module 300 is further configured to: For each distributed radar node, obtain the false alarm probability; The false alarm probability is obtained based on the historical average false alarm probability of the distributed radar nodes. The false alarm probability and received signal-to-noise ratio of each distributed radar node are input into the detection threshold configuration model to obtain the weak signal detection threshold. The construction of the detection threshold configuration model includes: Acquire multiple sets of sample data, each set of sample data including received signal-to-noise ratio samples, preset false alarm probability samples, and weak signal detection thresholds for the corresponding samples; A detection threshold configuration model is constructed based on machine learning. The detection threshold configuration model is trained using the sample data until it converges, thus completing the construction of the detection threshold configuration model. Calculate the ratio of the signal delay of each of the distributed radar nodes to that of the reference node, and use it as the delay deviation coefficient of that node; By combining the time delay deviation coefficient and the signal-to-noise ratio deviation coefficient, a weighted calculation is performed to obtain the node credibility. Based on node credibility, node weights are calculated and obtained, and the detection results of each distributed radar node are fused to obtain the target detection result. The average confidence level of multiple distributed radar nodes is calculated as the result confidence level. This result is then combined with the target detection result to output the weak signal target detection result.

[0053] In summary, the embodiments of this application have at least the following technical effects: This application proposes a distributed weak signal target detection method and system for digital radar. By constructing a full-process collaborative processing mechanism from node timing correction and coherent accumulation optimization to adaptive configuration of detection threshold and weighted fusion of detection results, the detection capability of the distributed radar system for weak signal targets is significantly improved. First, the signal delay and received signal-to-noise ratio of each distributed radar node are obtained. Based on this, the node with the optimal signal-to-noise ratio is selected as the benchmark. By calculating the delay deviation coefficient and signal-to-noise ratio deviation coefficient between each node and the benchmark node, the coherent accumulation time window is adaptively optimized, and the signal timing of each node is accurately corrected, so that the echoes of multiple nodes are effectively aligned in the time domain. Then, linear superposition accumulation is performed, which significantly improves the coherent accumulation efficiency and signal energy enhancement of weak signal targets. Building upon this foundation, this method, for each distributed radar node, combines its received signal-to-noise ratio (SNR) with a preset false alarm probability. Through machine learning, it constructs and trains a convergent detection threshold configuration model, enabling dynamic and precise configuration of detection thresholds under different SNR conditions. This effectively overcomes the problems of insufficient detection sensitivity or excessively high false alarm rates in weak signal scenarios caused by traditional fixed thresholds, significantly improving the accuracy and stability of independent detection by each node. Furthermore, after node-level detection is completed, the delay deviation coefficient and SNR deviation coefficient of each node are comprehensively evaluated based on signal delay and received SNR. A weighted calculation is then performed to obtain the node credibility, generating node weights. Weighted fusion decisions are then made based on the detection results of each node, and the average credibility of multiple nodes is calculated and output as the final credibility. Compared to traditional fusion strategies using equal-weighted merging or simple voting mechanisms, the technical solution provided in this application significantly optimizes the rationality of fusion decisions, allowing nodes with high SNR and more accurate delay estimation to contribute more to the final decision, thereby effectively improving the overall accuracy and robustness of fusion detection.

[0054] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0055] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0056] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A distributed weak signal target detection method for digital radar, characterized in that, include: Obtain the signal delay of multiple distributed radar nodes and the received signal-to-noise ratio; Based on the signal delay and the received signal-to-noise ratio, the optimized coherence time is calculated and obtained, and the signal timing of each distributed radar node is corrected to perform coherent accumulation on weak signal targets. Based on the received signal-to-noise ratio and a preset false alarm probability, a weak signal detection threshold is obtained. Node weights are obtained based on signal delay and the received signal-to-noise ratio. The detection results of each distributed radar node are then fused to obtain a weak signal target detection result.

2. The digital radar distributed weak signal target detection method according to claim 1, characterized in that, Obtain the signal delay of multiple distributed radar nodes and the received signal-to-noise ratio, including: Multiple distributed radar nodes are acquired, and the time difference between the multiple distributed radar nodes receiving the echo of the same weak signal target is acquired as the signal delay. Calculate the ratio of target signal power to background noise power of the multiple distributed radar nodes to obtain the received signal-to-noise ratio.

3. The digital radar distributed weak signal target detection method according to claim 2, characterized in that, Based on the signal delay and the received signal-to-noise ratio, the optimized coherence time is calculated, including: Among the multiple distributed radar nodes, the node with the highest received signal-to-noise ratio is selected as the reference node; The absolute difference between the signal delay between each of the distributed radar nodes and the reference node is calculated as the ratio of the signal delay between the standard nodes, and this ratio is used as the synchronization deviation coefficient of that node. Calculate the ratio of the received signal-to-noise ratio of each of the distributed radar nodes to that of the reference node, and use it as the signal-to-noise ratio deviation coefficient; Based on the synchronization deviation coefficient and the signal-to-noise ratio deviation coefficient, a weighted calculation is performed to obtain the coherence time optimization coefficient, and the basic coherence time is optimized to obtain the optimized coherence time.

4. The digital radar distributed weak signal target detection method according to claim 3, characterized in that, Correcting the timing of signals from each distributed radar node and performing coherent accumulation on weak signal targets, including: The signal reception timing of the reference node is used as the reference timing. Based on the signal delay, the signal reception timing of the remaining distributed radar nodes is adjusted so that the weak signal target echoes received by the remaining nodes are aligned with the weak signal target echo timing of the reference node. The optimized coherence time is used to linearly superimpose and accumulate the weak signal target echoes received by each distributed radar node after timing correction.

5. The digital radar distributed weak signal target detection method according to claim 4, characterized in that, Based on the received signal-to-noise ratio and combined with a preset false alarm probability, a weak signal detection threshold is obtained, including: For each distributed radar node, obtain the false alarm probability; The false alarm probability and received signal-to-noise ratio of each distributed radar node are input into the detection threshold configuration model to obtain the weak signal detection threshold.

6. The digital radar distributed weak signal target detection method according to claim 5, characterized in that, The false alarm probability is obtained based on the historical average false alarm probability of the distributed radar nodes.

7. The digital radar distributed weak signal target detection method according to claim 6, characterized in that, The construction of the detection threshold configuration model includes: Acquire multiple sets of sample data, each set of sample data including received signal-to-noise ratio samples, preset false alarm probability samples, and weak signal detection thresholds for the corresponding samples; A detection threshold configuration model is constructed based on machine learning. The detection threshold configuration model is trained using the sample data until it converges, thus completing the construction of the detection threshold configuration model.

8. The digital radar distributed weak signal target detection method according to claim 7, characterized in that, Based on the signal delay and the received signal-to-noise ratio, node weights are obtained, and the detection results of each distributed radar node are fused to obtain weak signal target detection results, including: Calculate the ratio of the signal delay of each of the distributed radar nodes to that of the reference node, and use it as the delay deviation coefficient of that node; By combining the time delay deviation coefficient and the signal-to-noise ratio deviation coefficient, a weighted calculation is performed to obtain the node credibility. Based on node credibility, node weights are calculated and obtained, and the detection results of each distributed radar node are fused to obtain the target detection result.

9. The digital radar distributed weak signal target detection method according to claim 8, characterized in that, Obtaining target detection results also includes: The average confidence level of multiple distributed radar nodes is calculated as the result confidence level. This result is then combined with the target detection result to output the weak signal target detection result.

10. A digital radar distributed weak signal target detection system, characterized in that, For implementing the digital radar distributed weak signal target detection method according to any one of claims 1-9, the system comprises: The information acquisition module is used to acquire the signal delay of multiple distributed radar nodes and to acquire the received signal-to-noise ratio; The coherent accumulation module is used to calculate and obtain the optimized coherent time based on the signal delay and the received signal-to-noise ratio, and to correct the signal timing of each distributed radar node, and to coherently accumulate weak signal targets. The result acquisition module is used to obtain a weak signal detection threshold based on the received signal-to-noise ratio and a preset false alarm probability, and to obtain node weights based on signal delay and the received signal-to-noise ratio, and to fuse the detection results of each distributed radar node to obtain a weak signal target detection result.

Images