Distributed communication signal detection method, device and system under limited communication range of single node
Patent Information
- Application Number
- CN202311009806.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-10
- Publication Date
- 2026-08-21
- Estimated Expiration
- 2043-08-10
AI Technical Summary
然而，传统的分布式信号检测方法主要是基于融合中心全局收集所有接收节点的原始数据进行的处理，这种方法在某些情况下可能不适用
[0034]本申请提供的单节点通信范围受限下的分布式通信信号检测方法、装置和系统，融合中心通过接收分布式布置的多个接收节点上传的样本协方差系数，该样本协方差系数是由各个接收节点根据本地收集到的观测数据集合在本地计算得到；融合中心进而根据接收的样本协方差系数，计算全部接收节点观测数据的样本协方差矩阵，在信号源不存在和信号源存在两种假设下，分别构建所述样本协方差矩阵与理论协方差矩阵的误差函数，并根据这两种假设下的误差函数的比值构建MFER检验统计量，基于所述样本协方差矩阵对MFER检验统计量中的未知变量采用矩阵分解算法进行估计，得到这两种假设下的全部接收节点观测数据的理论协方差矩阵的估计值，带入MFER检验统计量，并与预设的判决门限进行比较，来判决检测频段内是否存在信号源信号，从而完成分布式通信信号检测。
Smart Images
Figure CN117118565B_ABST
Abstract
Description
Technical Field
[0001] This application relates to distributed communication signal detection technology, specifically to a method, apparatus, and system for distributed communication signal detection under conditions where the communication range of a single node is limited. Background Technology
[0002] Two factors make signal detection challenging in practical applications. First, the signal-to-noise ratio (SNR) at each receiving node can be very low. With the increasing number of wireless communication devices, the growing complexity and variability of the electromagnetic environment, the deteriorating channel transmission conditions, and the increasing background noise and interference, the signal at the receiving end is often interfered with by strong background noise, sometimes even drowning out the useful signal, resulting in a very low received SNR. Second, noise can vary with time and location, leading to noise uncertainty. Under these conditions, false detections or missed detections may occur, and erroneous detection results will inevitably have a serious impact on modulation scheme identification and signal processing.
[0003] In modern communication systems, distributed detection using multiple receiver nodes improves signal detection performance under low signal-to-noise ratio (SNR) conditions compared to single-receiver-node methods. However, traditional distributed signal detection methods primarily rely on a fusion center globally collecting raw data from all receiver nodes. This approach may not be suitable in certain situations. For example, when the communication distance between the receiver node and the fusion center exceeds the minimum communication transmission requirement, resulting in incomplete raw data reception, or when data transmission errors degrade the quality of the raw data, the performance of traditional distributed signal detection methods will be significantly reduced. Summary of the Invention
[0004] The purpose of this application is to propose a distributed communication signal detection method, device and system under the condition of limited single-node communication range, which can reduce communication requirements, effectively utilize the observation data collected by each receiving node, and improve the detection probability of communication signals under low signal-to-noise ratio.
[0005] According to a first aspect of this application, a distributed communication signal detection method under conditions of limited single-node communication range is provided. The method is executed by a fusion center and includes the following steps:
[0006] S1, receive sample covariance coefficients uploaded by multiple distributed receiving nodes, wherein each receiving node collects observation data in the detection frequency band and broadcasts the collected observation data to the outside, and receives observation data transmitted by other receiving nodes. The sample covariance coefficients are calculated locally by each receiving node based on the set of observation data collected locally.
[0007] S2, Calculate the sample covariance matrix of all received node observation data based on the received sample covariance coefficients, wherein some coefficients are allowed to be missing in the sample covariance matrix;
[0008] S3. Under the two assumptions of the absence of a signal source and the presence of a signal source, respectively construct the error function of the sample covariance matrix and the theoretical covariance matrix, and construct the matrix factorization error ratio (MFER) test statistic based on the ratio of the error function under the two assumptions. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all the observation data of the receiving nodes under the two assumptions.
[0009] S4. Based on the sample covariance matrix, a matrix factorization algorithm is used to estimate the theoretical covariance matrix of all receiving node observation data under the two assumptions.
[0010] S5, substitute the estimated values of the theoretical covariance matrix under the two assumptions into the MFER test statistic, and compare it with the preset decision threshold to determine whether there is a signal source signal in the detection frequency band.
[0011] According to a second aspect of this application, another method for distributed communication signal detection under single-node communication range limitations is provided. The method is performed by a single receiving node arranged in a distributed manner and includes the following steps:
[0012] Collect observation data within the detection frequency band;
[0013] The collected observation data was broadcast outwards;
[0014] Receive observation data transmitted by other receiving nodes within a spatial distance that meets the minimum communication transmission requirements;
[0015] Calculate the sample covariance coefficient locally based on the locally collected set of observation data;
[0016] The calculated sample covariance coefficients are uploaded to the fusion center, which then calculates the sample covariance matrix of all received node observation data based on the sample covariance coefficients uploaded by multiple receiving nodes. Some coefficients in the sample covariance matrix are allowed to be missing. The fusion center then constructs error functions between the sample covariance matrix and the theoretical covariance matrix under two assumptions: the absence of a signal source and the presence of a signal source. Based on the ratio of these error functions under the two assumptions, a matrix factorization error ratio (MFER) test statistic is constructed. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all received node observation data under the two assumptions. Furthermore, the fusion center uses a matrix factorization algorithm based on the sample covariance matrix to estimate the theoretical covariance matrix of all received node observation data under the two assumptions. The estimated values of the theoretical covariance matrix under the two assumptions are substituted into the MFER test statistic and compared with a preset decision threshold to determine whether a signal source exists within the detection frequency band.
[0017] According to a third aspect of this application, a distributed communication signal detection device for single-node communication range limitation is provided. The device is located in a fusion center and includes the following modules:
[0018] The sample covariance coefficient receiving module is used to receive sample covariance coefficients uploaded by multiple distributed receiving nodes. Each receiving node collects observation data in the detection frequency band and broadcasts the collected observation data to the outside world, and receives observation data transmitted by other receiving nodes. The sample covariance coefficients are calculated locally by each receiving node based on the set of observation data collected locally.
[0019] The sample covariance matrix calculation module is used to calculate the sample covariance matrix of all received node observation data based on the received sample covariance coefficients. Some coefficients are allowed to be missing in the sample covariance matrix.
[0020] The MFER test statistic construction module is used to construct the error function of the sample covariance matrix and the theoretical covariance matrix under two assumptions: the absence of a signal source and the presence of a signal source. The module also constructs the matrix factorization error ratio (MFER) test statistic based on the ratio of the error functions under the two assumptions. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all receiving node observation data under the two assumptions.
[0021] The matrix factorization estimation module is used to estimate the theoretical covariance matrix of all receiving node observation data under the two assumptions based on the sample covariance matrix using a matrix factorization algorithm.
[0022] The comparison and decision module is used to substitute the estimated values of the theoretical covariance matrix under the two assumptions into the MFER test statistic and compare them with the preset decision threshold to determine whether there is a signal source signal within the detection frequency band.
[0023] According to a fourth aspect of this application, another distributed communication signal detection device for single-node communication range limitations is provided. The device is installed at a single receiving node in a distributed configuration and includes the following modules:
[0024] The acquisition module is used to acquire observation data within the detection frequency band;
[0025] The broadcast module is used to broadcast the collected observation data to the outside world;
[0026] The receiving module is used to receive observation data transmitted by other receiving nodes within a spatial distance that meets the minimum communication transmission requirements;
[0027] The sample covariance coefficient calculation module is used to calculate the sample covariance coefficient locally based on the locally collected set of observation data.
[0028] The sample covariance coefficient uploading module is used to upload the calculated sample covariance coefficients to the fusion center. The fusion center then calculates the sample covariance matrix of all received node observation data based on the sample covariance coefficients uploaded by multiple receiving nodes. Some coefficients are allowed to be missing in the sample covariance matrix. The fusion center then constructs error functions between the sample covariance matrix and the theoretical covariance matrix under two assumptions: the absence of a signal source and the presence of a signal source. Based on the ratio of these error functions under the two assumptions, a matrix factorization error ratio (MFER) test statistic is constructed. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all received node observation data under the two assumptions. Furthermore, the fusion center uses a matrix factorization algorithm based on the sample covariance matrix to estimate the theoretical covariance matrix of all received node observation data under the two assumptions. The estimated values of the theoretical covariance matrix under the two assumptions are substituted into the MFER test statistic and compared with a preset decision threshold to determine whether a signal source exists within the detection frequency band.
[0029] According to a fifth aspect of this application, a distributed communication signal detection system for single-node communication range limitations is provided, comprising a fusion center and multiple distributed receiving nodes.
[0030] Each receiving node collects observation data within the detection frequency band and broadcasts the collected observation data outwards. It also receives observation data transmitted by other receiving nodes within a spatial distance that meets the minimum communication transmission requirements, calculates the sample covariance coefficient locally based on the locally collected observation data set, and uploads the calculated sample covariance coefficient to the fusion center.
[0031] The fusion center includes a memory and a processor. The memory stores a computer program, which is loaded and executed by the processor. The processor processes the sample covariance coefficients uploaded by multiple receiving nodes to realize the distributed communication signal detection method under the single-node communication range limitation of the first aspect mentioned above.
[0032] According to a sixth aspect of this application, a computer-readable storage medium is provided that stores one or more computer programs, which, when executed by a processor, implement the distributed communication signal detection method under the single-node communication range limitation of the first aspect described above.
[0033] The beneficial effects of the embodiments of this application are:
[0034] The distributed communication signal detection method, apparatus, and system provided in this application for single-node communication range limitation involves a fusion center receiving sample covariance coefficients uploaded by multiple distributed receiving nodes. These sample covariance coefficients are calculated locally by each receiving node based on its locally collected observation data set. The fusion center then calculates the sample covariance matrix of all receiving node observation data based on the received sample covariance coefficients. Under two assumptions—the absence of a signal source and the presence of a signal source—error functions are constructed between the sample covariance matrix and the theoretical covariance matrix. The ratio of these error functions under these two assumptions is used to construct an MFER test statistic. Based on the sample covariance matrix, the unknown variables in the MFER test statistic are estimated using a matrix factorization algorithm to obtain estimated values of the theoretical covariance matrix of all receiving node observation data under these two assumptions. These estimated values are then substituted into the MFER test statistic and compared with a preset decision threshold to determine whether a signal source exists within the detection frequency band, thus completing the distributed communication signal detection.
[0035] In this application, sample covariance coefficients are calculated locally by multiple receiving nodes and uploaded to the fusion center. The fusion center calculates the sample covariance matrix of all receiving node observation data based on the sample covariance coefficients uploaded by multiple receiving nodes. This effectively utilizes the observation data collected by each receiving node and improves the detection probability of communication signals under low signal-to-noise ratio conditions. Even when the communication range of a single node is limited, and the communication distance between some receiving nodes or between some receiving nodes and the fusion center exceeds the minimum communication transmission requirement, resulting in some missing coefficients in the calculated sample covariance matrix, the fusion center can still use the partially missing sample covariance matrix. By constructing the MFER test statistic, the unknown variables in the MFER test statistic are estimated using a matrix factorization algorithm, realizing distributed communication signal detection under conditions of limited single-node communication range, thereby greatly reducing communication requirements. At the same time, since the constructed MFER test statistic utilizes the correlation between the observation data of multiple receiving nodes, this application also has good noise immunity when the noise of the receiving nodes fluctuates. Attached Figure Description
[0036] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application, and those skilled in the art can obtain other drawings based on these drawings. In the drawings:
[0037] Figure 1 A flowchart illustrating a distributed communication signal detection method under limited single-node communication range provided in an embodiment of this application is shown.
[0038] Figure 2 This illustration shows a schematic diagram of the process of estimating the theoretical covariance matrix of all received node observation data using a matrix factorization algorithm, as provided in an embodiment of this application.
[0039] Figure 3 This illustration shows a schematic diagram of a distributed communication signal detection device with limited single-node communication range provided in an embodiment of this application.
[0040] Figure 4 A flowchart illustrating another distributed communication signal detection method under limited single-node communication range provided in an embodiment of this application is shown.
[0041] Figure 5 This invention provides a schematic diagram of another distributed communication signal detection device with limited single-node communication range, according to an embodiment of this application.
[0042] Figure 6This illustration shows a schematic diagram of the architecture of a distributed communication signal detection system with limited single-node communication range provided in an embodiment of this application.
[0043] Figure 7 A performance comparison chart of different distributed detection methods for QPSK modulated signals is shown. Detailed Implementation
[0044] Embodiments of this application will now be described in more detail with reference to the accompanying drawings. These embodiments are provided to enable a more thorough understanding of this application and to fully convey the scope of this application to those skilled in the art. Although exemplary embodiments of this application are shown in the drawings, it should be understood that this application can be implemented in various forms and should not be limited to the embodiments set forth herein.
[0045] Example 1
[0046] Figure 1 The illustration shows a flowchart of a distributed communication signal detection method under limited single-node communication range provided in an embodiment of this application. This method is executed by a fusion center. Figure 1 The methods shown include:
[0047] S1, receiving sample covariance coefficients uploaded by multiple distributed receiving nodes, wherein each receiving node collects observation data within the detection frequency band and broadcasts the collected observation data, and receives observation data transmitted by other receiving nodes, and the sample covariance coefficients are calculated locally by each receiving node based on the set of observation data collected locally.
[0048] In this application, each receiving node in a distributed configuration collects observation data within the detection frequency band, broadcasts the collected observation data, receives observation data transmitted by other receiving nodes within a spatial distance that meets the minimum communication transmission requirements, and calculates the sample covariance coefficient locally based on the locally collected observation data set.
[0049] Suppose the set of observation data collected by the j-th receiving node is in It is the raw data of length T collected by the j-th receiving node. If the set of node indices that have established communication with the i-th receiving node is a collection, then the sample covariance coefficient that needs to be calculated for the j-th receiving node is:
[0050]
[0051] Then each receiving node uploads the calculated sample covariance coefficients to the fusion center, which receives the sample covariance coefficients uploaded by multiple receiving nodes within a spatial distance that meets the minimum communication transmission requirements.
[0052] S2, calculate the sample covariance matrix of all received node observation data based on the received sample covariance coefficients, wherein some coefficients are allowed to be missing in the sample covariance matrix.
[0053] The fusion center calculates the sample covariance matrix of all the observation data from the receiving nodes based on the sample covariance coefficients uploaded by multiple receiving nodes, which can effectively utilize the observation data collected by each receiving node.
[0054] In actual testing, because the communication distance between some receiving nodes or between some receiving nodes and the fusion center exceeds the minimum communication transmission requirement, the fusion center cannot receive all the covariance coefficients, resulting in some coefficients being missing in the calculated sample covariance matrix, which is an incomplete sample covariance matrix or a partially missing sample covariance matrix.
[0055] S3. Under the two assumptions of the absence of a signal source and the presence of a signal source, respectively construct the error function of the sample covariance matrix and the theoretical covariance matrix, and construct the matrix factorization error ratio (MFER) test statistic based on the ratio of the error function under the two assumptions. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all receiving node observation data under the two assumptions.
[0056] This step S3 specifically includes:
[0057] Signal detection can be formulated as a hypothesis testing problem, i.e., there are two hypotheses: Q0 indicates that the signal source does not exist, and Q1 indicates that the signal source exists.
[0058] Under these two assumptions, the error functions for constructing the sample covariance matrix and the theoretical covariance matrix are respectively:
[0059]
[0060]
[0061] In the formula, S is the sample covariance matrix of all receiver node observations, Σ0 is the theoretical covariance matrix of all receiver node observations under the Q0 assumption, and Σ1 is the theoretical covariance matrix of all receiver node observations under the Q1 assumption. Represents the square of the norm. Let p be a symmetric index set, where p represents the total number of receiving nodes, and S is the sample covariance coefficient corresponding to (i,j)∈Ω. ij It can be received by the fusion center, and
[0062]
[0063] The matrix factorization error ratio (MFER) test statistic ξ is constructed based on the ratio of the error functions under these two hypotheses. MFER for:
[0064]
[0065] MFER test statistic ξ MFER The unknown variables are Σ0 and Σ1.
[0066] S4. Based on the sample covariance matrix, a matrix factorization algorithm is used to estimate the theoretical covariance matrix of all receiving node observation data under the two assumptions.
[0067] Figure 2 This illustration shows a schematic diagram of the process for estimating the theoretical covariance matrix of all received node observation data using a matrix factorization algorithm, as provided in an embodiment of this application. Figure 2 As shown, step S4 specifically includes:
[0068] S40, Input the sample covariance matrix S;
[0069] S41, Set the initial feasible solution H 0 and Ψ 0 The initial iteration variable k = 0, and the step size η, where This represents the channel parameters from the signal source to each receiving node, where r represents the number of signal sources, and Ψ represents the number of receiving nodes. 0 It is a positive definite diagonal matrix representing the noise parameters at each receiving node;
[0070] S42, Update in And the calculation method for its i-th row is as follows: Where h i It is the i-th row of H.
[0071] S43, Update Ψ k+1 =Diag(SH) k+1 (H k+1 ) H );
[0072] S44, determine if it converges. The convergence condition is ||H||. k+1 -H k || F ≤10 -2 And ||Ψ k+1 -Ψ k || F ≤10 -2If the convergence condition is not met, then k = k + 1, and return to step S42 to continue execution;
[0073] S45, if the convergence condition is met, output
[0074] Under the Q1 assumption, the output is This is the estimated value of the theoretical covariance matrix Σ1 of all the observation data from the receiving nodes;
[0075] Under the Q0 assumption, let H k =0, thus obtaining the estimated value of the theoretical covariance matrix Σ0 of all received node observation data.
[0076] S5, substitute the estimated values of the theoretical covariance matrix under the two assumptions into the MFER test statistic, and compare it with the preset decision threshold to determine whether there is a signal source signal in the detection frequency band.
[0077] This step S5 specifically includes:
[0078] The MFER detector is constructed as follows:
[0079]
[0080] In the formula, γ is a preset decision threshold, which is set when the false alarm probability is specified; Let Σ0 be the estimate under the assumption Q0. Let Σ1 be the estimated value under the Q1 assumption.
[0081] The threshold value γ can be obtained through a Monte Carlo experiment: generate a large number of random noise samples, and then calculate the MFER test statistic ξ. MFER The empirical distribution is then used, followed by the user-specified false alarm rate P. FA Calculate the threshold γ. For example, when the user specifies P... FA When γ = 1%, ξ should be selected. MFER The minimum value in the top 1% of the empirical distribution.
[0082] To obtain the estimate of Σ0 under the Q0 assumption The estimated value of Σ1 under assumption Q1 Then, the estimated value and Substitute the MFER test statistic into the input and compare it with the pre-set decision threshold. When ξ... MFER If ξ > γ, then it is determined that a signal source signal exists within the detection frequency band. MFER If the value is less than or equal to γ, then it is determined that there is no signal source within the detection frequency band. This completes the distributed communication signal detection method described in this application.
[0083] Example 2
[0084] Figure 3 The illustration shows a flowchart of another distributed communication signal detection method under limited single-node communication range provided in an embodiment of this application, which is executed by a single receiving node arranged in a distributed manner. Figure 4 The method shown includes the following steps S31 to S35:
[0085] Step S31: Collect observation data within the detection frequency band.
[0086] Step S32: Broadcast the collected observation data.
[0087] Step S33: Receive observation data transmitted by other receiving nodes within a spatial distance that meets the minimum communication transmission requirements.
[0088] Step S34: Calculate the sample covariance coefficient locally based on the locally collected observation data set.
[0089] As described above, suppose the set of observation data collected by the j-th receiving node is... in It is the raw data of length T collected by the j-th receiving node. If the set of node indices that have established communication with the i-th receiving node is a collection, then the sample covariance coefficient that needs to be calculated for the j-th receiving node is:
[0090] Step S35: The calculated sample covariance coefficients are uploaded to the fusion center so that the fusion center can perform communication signal detection based on the sample covariance coefficients uploaded by multiple receiving nodes.
[0091] Specifically, referring to the aforementioned method embodiment one, the fusion center calculates the sample covariance matrix of all received node observation data based on the sample covariance coefficients uploaded by multiple receiving nodes. Some coefficients in the sample covariance matrix are allowed to be missing. Then, under two assumptions—the absence of a signal source and the presence of a signal source—the fusion center constructs error functions between the sample covariance matrix and the theoretical covariance matrix, respectively. Based on the ratio of the error functions under the two assumptions, an MFER test statistic is constructed. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all received node observation data under the two assumptions. Furthermore, the fusion center uses a matrix factorization algorithm based on the sample covariance matrix to estimate the theoretical covariance matrix of all received node observation data under the two assumptions. The estimated values of the theoretical covariance matrix under the two assumptions are substituted into the MFER test statistic and compared with a preset decision threshold to determine whether a signal source signal exists within the detection frequency band, thereby achieving distributed communication signal detection.
[0092] Example 3
[0093] and Figure 1 The distributed communication signal detection method under the limited single-node communication range shown belongs to the same technical concept. The embodiments of this application also provide a distributed communication signal detection device. Figure 4 This illustration shows a schematic diagram of a distributed communication signal detection device with limited single-node communication range provided in an embodiment of this application. Figure 4 The device shown is located at the fusion center and includes the following modules:
[0094] The sample covariance coefficient receiving module 41 is used to receive the sample covariance coefficients uploaded by multiple receiving nodes arranged in a distributed manner. Each receiving node collects observation data in the detection frequency band and broadcasts the collected observation data to the outside world, and receives observation data transmitted by other receiving nodes. The sample covariance coefficients are calculated locally by each receiving node based on the set of observation data collected locally.
[0095] The sample covariance matrix calculation module 42 is used to calculate the sample covariance matrix of all received node observation data based on the received sample covariance coefficients. Some coefficients are allowed to be missing in the sample covariance matrix.
[0096] The MFER test statistic construction module 43 is used to construct the error function of the sample covariance matrix and the theoretical covariance matrix under two assumptions: the absence of a signal source and the presence of a signal source. The module also constructs the matrix factorization error ratio (MFER) test statistic based on the ratio of the error functions under the two assumptions. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all the observation data of the receiving nodes under the two assumptions.
[0097] The matrix factorization estimation module 44 is used to estimate the theoretical covariance matrix of all receiving node observation data under the two assumptions based on the sample covariance matrix using a matrix factorization algorithm.
[0098] The comparison and decision module 45 is used to substitute the estimated values of the theoretical covariance matrix under the two assumptions into the MFER test statistic and compare them with the preset decision threshold to determine whether there is a signal source signal in the detection frequency band.
[0099] Figure 4 The implementation process of each module in the device shown can be found in the aforementioned method embodiment one, and will not be repeated here.
[0100] Example 4
[0101] and Figure 3The distributed communication signal detection method under the limited single-node communication range shown belongs to the same technical concept, and the embodiments of this application also provide a distributed communication signal detection device. Figure 5 This illustration shows a schematic diagram of another distributed communication signal detection device with limited single-node communication range provided in an embodiment of this application. Figure 5 The device shown is set in a single receiving node in a distributed arrangement and includes the following modules:
[0102] Acquisition module 51 is used to acquire observation data within the detection frequency band;
[0103] Broadcast module 52 is used to broadcast the collected observation data to the outside world;
[0104] The receiving module 53 is used to receive observation data transmitted by other receiving nodes within a spatial distance that meets the minimum communication transmission requirements;
[0105] The sample covariance coefficient calculation module 54 is used to calculate the sample covariance coefficient locally based on the locally collected set of observation data.
[0106] The sample covariance coefficient uploading module 55 is used to upload the calculated sample covariance coefficients to the fusion center, so that the fusion center can calculate the sample covariance matrix of all received node observation data based on the sample covariance coefficients uploaded by multiple receiving nodes. Some coefficients are allowed to be missing in the sample covariance matrix. Then, the fusion center constructs error functions between the sample covariance matrix and the theoretical covariance matrix under two assumptions: the absence of a signal source and the presence of a signal source. Based on the ratio of the error functions under the two assumptions, a matrix MFER test statistic is constructed. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all received node observation data under the two assumptions. Furthermore, the fusion center uses a matrix factorization algorithm based on the sample covariance matrix to estimate the theoretical covariance matrix of all received node observation data under the two assumptions. The estimated values of the theoretical covariance matrix under the two assumptions are substituted into the MFER test statistic and compared with a preset decision threshold to determine whether a signal source signal exists within the detection frequency band.
[0107] Figure 5 The implementation process of each module in the device shown can also be found in the aforementioned method embodiment one, and will not be repeated here.
[0108] Example 5
[0109] The distributed communication signal detection method and apparatus under the single-node communication range limitation of the foregoing embodiments belong to the same technical concept. The embodiments of this application also provide a distributed communication signal detection system. Figure 6This illustration shows a schematic diagram of the architecture of a distributed communication signal detection system with limited single-node communication range, as provided in an embodiment of this application. Figure 6 As shown, it includes a fusion center 60 and multiple receiving nodes (61, 62, 63, ...) arranged in a distributed manner.
[0110] Each receiving node (61, 62, 63, ...) collects observation data within the detection frequency band and broadcasts the collected observation data outwards. It also receives observation data transmitted by other receiving nodes within a spatial distance that meets the minimum communication transmission requirements, calculates the sample covariance coefficient locally based on the locally collected observation data set, and uploads the calculated sample covariance coefficient to the fusion center 60.
[0111] The fusion center 60 includes a memory and a processor. The memory stores a computer program, which is loaded and executed by the processor. The processor processes the sample covariance coefficients uploaded by multiple receiving nodes to realize the distributed communication signal detection method under single-node communication range limitation in the aforementioned method embodiment 1.
[0112] Figure 6 For the implementation process of the fusion center 60 in the system shown, please refer to the aforementioned method embodiment one, which will not be repeated here.
[0113] Example 6
[0114] This application also provides a computer-readable storage medium that stores one or more computer programs. When executed by a processor, the one or more computer programs implement the distributed communication signal detection method under single-node communication range limitation described in the first embodiment of the aforementioned method, which will not be elaborated further here.
[0115] Example 7
[0116] This embodiment further verifies and illustrates the technical effects of the above embodiments of this application through simulation experiments.
[0117] First, a simulated signal source generates a communication signal. The communication signal modulation method is QPSK modulation, and the carrier frequency is 128×10⁻⁶. 3 Hz, symbol rate 20×10 3 Baud, with a detection time of 0.5s per node and a sampling rate of 120×10⁻⁶. 3 The low-pass filter has a passband attenuation factor of 4 and a stopband attenuation factor of 20 Hz. Monte Carlo simulations were performed 500 times. The signal transmission and reception process includes baseband signal generation, shaping filtering, modulation, the modulated signal passing through different spatial diversity channels to reach the receiving node, demodulation, and filtering to obtain the received data from each node.
[0118] In distributed detection, assuming there are p distributed receiving nodes receiving signals from a signal source from different directions, we can obtain p spatial diversity channels. Assuming each receiving node collects T data points within the observation time, the signal received by the i-th receiving node during the observation time can be represented as...
[0119]
[0120] In the formula, i = 1,...,p, t = 1,...,T, x i,t η represents the received signal from the receiving node in each spatial diversity channel. i,t Let be the received noise of the i-th receiving node during the observation time, with a mean of 0 and a variance of . The β follows a complex Gaussian distribution, and the received noise at different receiving nodes is statistically independent of each other; i Let f be the channel parameters from the signal source to the i-th receiving node, which are unknown complex parameters used to describe the channel propagation effects; i,t This indicates the signal source signal.
[0121] Next, the distributed receiving nodes will calculate the correlation coefficient locally and upload it to the fusion center. Assume the set of observation data collected by the j-th receiving node is... in It is the raw data of length T collected by the j-th receiving node. If the set of node indices that have established communication with the i-th receiving node is a collection, then the sample covariance coefficient that needs to be calculated for the j-th receiving node is:
[0122] Then each receiving node uploads the calculated sample covariance coefficients to the fusion center.
[0123] Finally, the fusion center executes the MFER detector of this application to determine whether there is a signal source signal within the detection frequency band.
[0124] Figure 7 A performance comparison chart of different distributed detection methods for QPSK modulated signals is shown. Figure 7In the figure, the horizontal axis SNR (dB) represents the signal-to-noise ratio, and the vertical axis Probability of Detection represents the detection probability. The comparison objects include: ED (Energy Detector), AGM (Arithmetic to Geometric Mean), EMR (Eigenvalue-Moment-Ratio), MME (Maximum-Minimum Eigenvalue), SLE (Scaled Largest Eigenvalue), and MFER (Matrix Factorization Error Ratio), which represents the matrix factorization error ratio detection method proposed in this application. It includes two cases: one is that all data samples are observed (100% observed), and the other is that only a portion of the data samples are observed (70% observed).
[0125] from Figure 7 As can be seen from the performance comparison chart, if the complete sample covariance matrix S is used, the MFER detection method proposed in this application achieves the best performance among all detection methods. However, all benchmark methods require the receipt of complete data samples. It is very important that the MFER detection method proposed in this application can still achieve better performance than all benchmark methods (which require the complete sample covariance matrix) even when using only the sample covariance matrix of partial observation data.
[0126] In summary, the proposed solution calculates sample covariance coefficients locally at multiple receiving nodes and uploads them to the fusion center. The fusion center then calculates the sample covariance matrix of all receiving node observations based on these coefficients. This effectively utilizes the observation data collected by each receiving node, improving the detection probability of communication signals under low signal-to-noise ratio conditions. Even when the communication distance between some receiving nodes or between some receiving nodes and the fusion center exceeds the minimum communication transmission requirement, resulting in missing coefficients in the calculated sample covariance matrix, the fusion center can still use the partially missing sample covariance matrix. By constructing the MFER test statistic and estimating the unknown variables in the MFER test statistic using matrix factorization, distributed communication signal detection under limited single-node communication range conditions is achieved, significantly reducing communication requirements. Furthermore, since the constructed MFER test statistic utilizes the correlation between observation data from multiple receiving nodes, the proposed solution also exhibits good noise immunity when receiving node noise fluctuates.
[0127] Finally, it should be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0128] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.
Claims
1. A method for detecting distributed communication signals under limited single-node communication range, characterized in that, The method is executed by the fusion center and includes the following steps: S1, receive sample covariance coefficients uploaded by multiple distributed receiving nodes, wherein each receiving node collects observation data in the detection frequency band and broadcasts the collected observation data to the outside, and receives observation data transmitted by other receiving nodes. The sample covariance coefficients are calculated locally by each receiving node based on the set of observation data collected locally. S2, Calculate the sample covariance matrix of all received node observation data based on the received sample covariance coefficients, wherein some coefficients are allowed to be missing in the sample covariance matrix; S3. Under the two assumptions of the absence of a signal source and the presence of a signal source, respectively construct the error function of the sample covariance matrix and the theoretical covariance matrix, and construct the matrix factorization error ratio (MFER) test statistic based on the ratio of the error function under the two assumptions. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all the observation data of the receiving nodes under the two assumptions. S4. Based on the sample covariance matrix, a matrix factorization algorithm is used to estimate the theoretical covariance matrix of all receiving node observation data under the two assumptions. S5, substitute the estimated values of the theoretical covariance matrix under the two assumptions into the MFER test statistic, and compare it with the preset decision threshold to determine whether there is a signal source signal in the detection frequency band.
2. The method according to claim 1, characterized in that, Step S1 specifically includes: Each receiving node collects observation data within the detection frequency band and broadcasts the collected observation data. It also receives observation data transmitted by other receiving nodes within a spatial distance that meets the minimum communication transmission requirements. Assume the set of observation data collected by the i-th receiving node is... ,in It is the first The length collected by each receiving node is The original data, Is with the first If there is a set of node indices for communication established by several receiving nodes, then the sample covariance coefficient to be calculated for the i-th receiving node is: ； Then each receiving node uploads the calculated sample covariance coefficients to the fusion center, which receives the sample covariance coefficients uploaded by multiple receiving nodes within a spatial distance that meets the minimum communication transmission requirements.
3. The method according to claim 2, characterized in that, Step S3 specifically includes: Signal detection can be formulated as a hypothesis testing problem, i.e., there are two hypotheses: This indicates that the signal source does not exist. This indicates that the signal source exists; Under these two assumptions, the error functions for constructing the sample covariance matrix and the theoretical covariance matrix are respectively: ； ； In the formula, It is the sample covariance matrix of all the observation data from the receiving nodes. for Assume the theoretical covariance matrix of all received node observation data. for Assume the theoretical covariance matrix of all received node observation data. Represents the square of the norm. It is a set of symmetric indices. This represents the total number of receiving nodes. When its corresponding sample covariance coefficient It can be received by the fusion center, and ； Construct the MFER test statistic based on the ratio of the error functions under the two hypotheses. for: ； MFER test statistic The unknown variables in are and .
4. The method according to claim 3, characterized in that, Step S4 specifically includes: S40, Input the sample covariance matrix ; S41, Set the initial feasible solution and Initial iteration variables Step length ,in This represents the channel parameters from the signal source to each receiving node. Indicates the number of signal sources. It is a positive definite diagonal matrix representing the noise parameters at each receiving node; S42, Update ,in And its first The method for calculating rows is as follows ,in yes The OK, ; S43, Update ; S44, determine if it converges; the convergence condition is... ,and If the convergence condition is not met, then Return to step S42 and continue execution; S45, if the convergence condition is met, output ; exist Assuming the output is That is, the theoretical covariance matrix of all the observation data from the receiving nodes. The estimated value; exist Assuming that, let =0, thus obtaining the theoretical covariance matrix of all received node observation data. The estimated value .
5. The method according to claim 4, characterized in that, Step S5 specifically includes: The MFER test statistic is constructed as follows: ； In the formula It is a preset decision threshold, the Set the false alarm probability when specified. for Assuming The estimated value, for Assuming The estimated value; when If so, it is determined that there is a signal source signal within the detection frequency band. If the signal source is not detected within the frequency band, it is determined that there is no signal source signal.
6. A method for detecting distributed communication signals under limited single-node communication range, characterized in that, The method is executed by a single receiving node in a distributed configuration and includes the following steps: Collect observation data within the detection frequency band; The collected observation data was broadcast outwards; Receive observation data transmitted by other receiving nodes within a spatial distance that meets the minimum communication transmission requirements; Calculate the sample covariance coefficient locally based on the locally collected set of observation data; The calculated sample covariance coefficients are uploaded to the fusion center, which then calculates the sample covariance matrix of all received node observation data based on the sample covariance coefficients uploaded by multiple receiving nodes. Some coefficients in the sample covariance matrix are allowed to be missing. The fusion center then constructs error functions between the sample covariance matrix and the theoretical covariance matrix under two assumptions: the absence of a signal source and the presence of a signal source. Based on the ratio of these error functions under the two assumptions, a matrix factorization error ratio (MFER) test statistic is constructed. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all received node observation data under the two assumptions. Furthermore, the fusion center uses a matrix factorization algorithm based on the sample covariance matrix to estimate the theoretical covariance matrix of all received node observation data under the two assumptions. The estimated values of the theoretical covariance matrix under the two assumptions are substituted into the MFER test statistic and compared with a preset decision threshold to determine whether a signal source exists within the detection frequency band.
7. A distributed communication signal detection device under single-node communication range limitation, characterized in that, The device is located in the fusion center and includes the following modules: The sample covariance coefficient receiving module is used to receive sample covariance coefficients uploaded by multiple distributed receiving nodes. Each receiving node collects observation data in the detection frequency band and broadcasts the collected observation data to the outside world, and receives observation data transmitted by other receiving nodes. The sample covariance coefficients are calculated locally by each receiving node based on the set of observation data collected locally. The sample covariance matrix calculation module is used to calculate the sample covariance matrix of all received node observation data based on the received sample covariance coefficients. Some coefficients are allowed to be missing in the sample covariance matrix. The MFER test statistic construction module is used to construct the error function of the sample covariance matrix and the theoretical covariance matrix under two assumptions: the absence of a signal source and the presence of a signal source. The module also constructs the matrix factorization error ratio (MFER) test statistic based on the ratio of the error functions under the two assumptions. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all the observation data of the receiving nodes under the two assumptions. The matrix factorization estimation module is used to estimate the theoretical covariance matrix of all receiving node observation data under the two assumptions based on the sample covariance matrix using a matrix factorization algorithm. The comparison and decision module is used to substitute the estimated values of the theoretical covariance matrix under the two assumptions into the MFER test statistic and compare them with the preset decision threshold to determine whether there is a signal source signal within the detection frequency band.
8. A distributed communication signal detection device under single-node communication range limitation, characterized in that, The device is installed at a single receiving node in a distributed configuration and includes the following modules: The acquisition module is used to acquire observation data within the detection frequency band; The broadcast module is used to broadcast the collected observation data to the outside world; The receiving module is used to receive observation data transmitted by other receiving nodes within a spatial distance that meets the minimum communication transmission requirements; The sample covariance coefficient calculation module is used to calculate the sample covariance coefficient locally based on the locally collected set of observation data. The sample covariance coefficient uploading module is used to upload the calculated sample covariance coefficients to the fusion center. The fusion center then calculates the sample covariance matrix of all received node observation data based on the sample covariance coefficients uploaded by multiple receiving nodes. Some coefficients in the sample covariance matrix are allowed to be missing. The fusion center then constructs error functions between the sample covariance matrix and the theoretical covariance matrix under two assumptions: the absence of a signal source and the presence of a signal source. Based on the ratio of these error functions under the two assumptions, a matrix factorization error ratio (MFER) test statistic is constructed. The unknown variable in the MFER test statistic is the theoretical covariance matrix of all received node observation data under the two assumptions. Furthermore, the fusion center uses a matrix factorization algorithm based on the sample covariance matrix to estimate the theoretical covariance matrix of all received node observation data under the two assumptions. The estimated values of the theoretical covariance matrix under the two assumptions are substituted into the MFER test statistic and compared with a preset decision threshold to determine whether a signal source exists within the detection frequency band.
9. A distributed communication signal detection system under single-node communication range limitation, characterized in that, Including a fusion center and multiple receiving nodes arranged in a distributed manner, Each receiving node collects observation data within the detection frequency band and broadcasts the collected observation data outwards. It also receives observation data transmitted by other receiving nodes within a spatial distance that meets the minimum communication transmission requirements, calculates the sample covariance coefficient locally based on the locally collected observation data set, and uploads the calculated sample covariance coefficient to the fusion center. The fusion center includes a memory and a processor. The memory stores a computer program, which is loaded and executed by the processor to process the sample covariance coefficients uploaded by multiple receiving nodes, thereby realizing the distributed communication signal detection method under the single-node communication range limitation as described in any one of claims 1 to 5.
10. A computer-readable storage medium storing one or more computer programs, which, when executed by a processor, implement the distributed communication signal detection method under single-node communication range limitation as described in any one of claims 1 to 5.