Mobile target detection method, device, equipment and product based on cell-free network

Abstract

The application discloses a mobile target detection method and device based on a cell-free network, equipment and products, and relates to the technical field of wireless network communication. The method comprises the following steps: determining n target detection periods based on target variation degrees, and acquiring predicted target positions in the n target detection periods; evaluating channel states of each node to the predicted target positions and a fusion center according to the predicted target positions in the n target detection periods and a channel knowledge map; acquiring optimal node combinations and corresponding optimal periods by calculating normalized traces based on a fusion covariance and the channel knowledge map; performing target detection based on the optimal node combinations and the corresponding optimal periods, and estimating a target position by fusing node detection information obtained through target detection, wherein the node detection information comprises information detected by all nodes participating in detection. The method realizes continuous and accurate tracking of the target position, improves the detection accuracy, and guarantees the communication efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless network communication, and particularly relates to a mobile target detection method and device based on a cell-free network, equipment and products. BACKGROUND

[0002] The research of communication and sensing integrated system mainly focuses on centralized communication and sensing integrated system. Such a system deploys a macro base station in each cell, but the signal quality at the user end is affected due to the significant signal attenuation of the communication link between the base station and the user, resulting in a decline in communication performance. In the target tracking and positioning process, it is difficult to achieve reliable target detection and tracking because some targets may be far away from the base station. Therefore, a cell-free massive multiple-input multiple-output (MIMO) system is proposed. Unlike the centralized massive MIMO system, the cell-free MIMO system deploys a large number of spatially distributed single-antenna or multi-antenna access points, which are connected to a central processing unit through a high-speed link, and use spatial multiplexing technology to enable all access points to serve all users and targets in space on the same time-frequency resources.

[0003] At present, in the target detection and positioning process, the distance or the distance and the angle are generally measured by nodes, and each access point can be used as a node for target detection. However, in the face of the detection demand of the moving state of the target, the accuracy of the existing method often cannot meet the actual requirements.

[0004] Therefore, how to improve the detection accuracy of the moving target is a problem to be solved at present. SUMMARY

[0005] The main purpose of the present application is to provide a mobile target detection method, device, equipment and product based on a cell-free network, which aims to solve the technical problem of low detection accuracy of the moving target.

[0006] To achieve the above purpose, the present application provides a mobile target detection method based on a cell-free network, which comprises:

[0007] determining n target detection periods based on the target change degree, and obtaining the predicted target positions in the n target detection periods;

[0008] evaluating the channel states of each node to the predicted target positions and the fusion center according to the predicted target positions in the n target detection periods and the channel knowledge map;

[0009] based on the fusion covariance and the channel knowledge map, obtaining the optimal node combination and the corresponding optimal period by calculating the normalized trace;

[0010] Based on the optimal node combination and the corresponding optimal period, target detection is performed, and a target position is estimated by fusing node detection information obtained by target detection, the node detection information including information obtained by all nodes participating in detection.

[0011] In an embodiment, the step of determining n target detection periods based on the target variation degree and obtaining predicted target positions in the n target detection periods includes:

[0012] The target variation degree is calculated according to the target direction variation degree and the target speed variation degree.

[0013] Based on the target variation degree, the maximum value of the detection period is determined by the value of the pheromone, and the value of the pheromone is used to represent the detection demand variation degree.

[0014] The maximum value of the detection period is divided to obtain n target detection periods.

[0015] Based on the n target detection periods, predicted target positions in the n target detection periods are obtained.

[0016] In an embodiment, the step of determining the maximum value of the detection period based on the target variation degree by the value of the pheromone includes:

[0017] An initial value of the pheromone is obtained, and the initial value of the pheromone is used to reflect the detection demand variation state of the target.

[0018] The value of the pheromone is updated in a decreasing manner according to the evaporation coefficient, and the number of iterations is recorded and updated.

[0019] When the value of the pheromone decays to a set minimum threshold value, the maximum value of the detection period is determined according to the current number of iterations.

[0020] In an embodiment, the step of obtaining predicted target positions in the n target detection periods based on the n target detection periods includes:

[0021] In the target detection period, the current position and the current motion state of the target are obtained.

[0022] Based on the current position and the current motion state of the target, the corresponding predicted target positions of the target in each detection period are predicted by a state transition matrix to obtain predicted target positions in the n target detection periods.

[0023] In an embodiment, the step of obtaining the optimal node combination and the corresponding optimal period by calculating the normalized trace based on the fused covariance and the channel knowledge map includes:

[0024] Based on the channel knowledge map, an estimated error covariance matrix of each node combination is corrected by channel state data obtained from the channel knowledge map, to obtain a normalized trace of different node combinations;

[0025] By comparing the normalized traces of the different node combinations, a node combination corresponding to a minimum normalized trace is regarded as an optimal node combination, and the optimal period is a period corresponding to the optimal node combination.

[0026] In an embodiment, the step of evaluating the channel state of each node to the predicted target position and the fusion center according to the predicted target position in the n target detection periods and the channel knowledge map comprises:

[0027] Actual channel data is collected, and the actual channel data is input to an initial channel knowledge map for model training to obtain a target channel knowledge map;

[0028] The predicted target position information in the n target detection periods, the fusion center information and the node information of each node are input to the target channel knowledge map, to obtain the channel state of each node to the predicted target position and the fusion center.

[0029] In an embodiment, the step of performing target detection based on the optimal node combination and the corresponding optimal period, and estimating the target position by fusing the node detection information obtained by target detection comprises:

[0030] A detection instruction issued by the fusion center is sent to a participating node, and the detection instruction is used to specify a detection time and the participating node;

[0031] Based on the optimal node combination and the corresponding optimal period, target detection is performed according to the detection instruction to obtain node detection information, and the node detection information includes distance data and angle data of each participating node and the target;

[0032] The node detection information is returned to the fusion center, and data fusion is performed on the node detection information to obtain fusion information;

[0033] The position information and the speed information of the target are calculated according to the fusion information, and the target position is estimated according to the position information and the speed information of the target.

[0034] In addition, to achieve the above object, the application further provides a mobile target detection device based on a cell-free network, which comprises:

[0035] A position prediction module is configured to determine n target detection periods based on a target change degree, and obtain predicted target positions in the n target detection periods;

[0036] a channel assessment module configured to assess channel states from each node to the predicted target position and the fusion center according to the predicted target position and the channel knowledge map in the n target detection periods;

[0037] a node optimization module configured to obtain an optimal node combination and a corresponding optimal period by calculating a normalized trace based on the fusion covariance and the channel knowledge map;

[0038] a target detection module configured to perform target detection based on the optimal node combination and the corresponding optimal period, and estimate a target position by fusing node detection information obtained by target detection, wherein the node detection information includes information obtained by all nodes participating in detection.

[0039] In addition, to achieve the above object, the present application further provides a storage medium, which is a computer readable storage medium, and has a computer program stored thereon, wherein the computer program is executed by a processor to implement the steps of the mobile target detection method based on a cell-free network.

[0040] In addition, to achieve the above object, the present application further provides a computer program product, which comprises a computer program, wherein the computer program is executed by a processor to implement the steps of the mobile target detection method based on a cell-free network.

[0041] The one or more technical solutions provided by the present application have at least the following technical effects:

[0042] n target detection periods are determined based on the target change degree, and predicted target positions in the n target detection periods are obtained; channel states of each node to the predicted target positions and a fusion center are evaluated according to the predicted target positions in the n target detection periods and a channel knowledge map; optimal node combinations and corresponding optimal periods are obtained by calculating a normalized trace based on the fusion covariance and the channel knowledge map; target detection is performed based on the optimal node combinations and the corresponding optimal periods, and a target position is estimated by fusing node detection information obtained by target detection, the node detection information including information obtained by detection of all participating nodes. By determining the detection period based on the target change degree, the system can adaptively adjust the detection frequency according to the motion state of the target. In each detection period, the predicted target position of the target is obtained, so that the system can more accurately locate the target. The prediction method combined with the state transition matrix makes the predicted target position more continuous and accurate, thereby improving the overall detection accuracy. The channel knowledge map is used to evaluate the channel state of each node to the target, to ensure that the node combination with the best channel state is selected in the detection process, and the optimal node combination and the optimal detection period are selected by calculating the normalized trace, to ensure the best configuration of resources while improving the detection and positioning accuracy. After each detection period ends, the target change degree is re-evaluated according to the updated target position, to ensure continuous and accurate tracking of the target position through the feedback mechanism, to ensure the detection accuracy, and to further significantly improve the communication efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0043] The accompanying drawings, which are incorporated herein and form a part of the specification, illustrate embodiments consistent with the present application and, together with the description, further serve to explain the principles of the application.

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, for those skilled in the art, other drawings can also be obtained based on these drawings without any creative effort.

[0045] Figure 1 The first schematic diagram of the centralized communication and perception integrated system scenario of the present application;

[0046] Figure 2 The second schematic diagram of the centralized communication and perception integrated system scenario of the present application;

[0047] Figure 3 The first embodiment of the mobile target detection method based on the cell-free network of the present application is shown in the flowchart.

[0048] Figure 4 The second embodiment of the mobile target detection method based on the cell-free network of the present application is shown in the flowchart.

[0049] Figure 5 This is a flowchart illustrating the third embodiment of the mobile target detection method based on a non-cellular network according to this application;

[0050] Figure 6 This is a schematic diagram of the signaling flow of a mobile target detection method without cellular network according to an embodiment of this application.

[0051] Figure 7 This is a schematic diagram of the module structure of a mobile target detection device based on a non-cellular network according to an embodiment of this application.

[0052] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0053] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0054] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0055] Currently, research on integrated communication and sensing systems mainly focuses on centralized integrated communication and sensing systems. For example... Figure 1 , Figure 2 As shown, this system deploys a macro base station in each cell. However, due to significant signal attenuation in the communication link between the base station and the user, the signal quality at the user end is affected, leading to a decline in communication performance. During target tracking and localization, reliable target detection and tracking are difficult because some targets may be far from the base station. To effectively address these issues, a cellular-free massive MIMO (Multiple-Input Multiple-Output) system has been proposed. Unlike centralized MIMO systems, cellular-free MIMO systems deploy a large number of spatially distributed single-antenna or multi-antenna access points. These access points are connected to the central processing unit via high-speed links. Through spatial multiplexing technology, all access points can provide services to all users and targets in space on the same time-frequency resources.

[0056] Since in target detection positioning, generally the distance or distance and angle is measured by the node, the node can be fixed or mobile. The system knows the position coordinates of the node, after detecting the distance or distance and angle with the target, the information is sent to the fusion center, and the position coordinates and velocity of the target are obtained by using Kalman filtering or particle filtering. Each access point can be used as a node to detect the target. Since multiple nodes are required to cooperate, the detection environment of different detection nodes in the target detection process, or the communication environment of different detection nodes and the fusion center in the communication process is very different. How to select the best node with good detection and communication environment and high positioning accuracy for detection, and realize the adaptive adjustment of tracking and detection parameters, faces great challenges.

[0057] The application provides a solution. First, a channel knowledge map of the detection area is constructed, n target detection periods are determined based on the target variation degree, and the predicted target position in the n target detection periods is obtained. The channel state of each node to the predicted target position and the fusion center is evaluated according to the predicted target position in the n target detection periods and the channel knowledge map. Based on the fusion covariance and the channel knowledge map, the optimal node combination and the corresponding optimal period are obtained by calculating the normalized trace. Target detection is performed based on the optimal node combination and the corresponding optimal period, and the target position is estimated by fusing the node detection information obtained by target detection. The node detection information includes the information detected by all nodes participating in the detection. By determining the detection period based on the target variation degree, the system can adaptively adjust the detection frequency according to the motion state of the target. In each detection period, by obtaining the predicted target position of the target, the system can more accurately locate the target. The prediction method combined with the state transition matrix makes the predicted target position more continuous and accurate, thereby improving the overall detection accuracy. The channel knowledge map is used to evaluate the channel state of each node to the target, to ensure that the node combination with the best channel state is selected in the detection process. The optimal node combination and the optimal detection period are selected by calculating the normalized trace, to ensure the positioning accuracy while realizing the best configuration of resources. After each detection period ends, the target variation degree is re-evaluated according to the updated target position, and the feedback mechanism is used to ensure the continuous and accurate tracking of the target position, to ensure the detection accuracy, and further to significantly improve the communication efficiency.

[0058] Based on this, the embodiment of the application provides a mobile target detection method based on a cell-free network. Figure 3 , Figure 3 FIG. 1 is a flowchart of a first embodiment of the mobile target detection method based on a cell-free network according to the application.

[0059] In this embodiment, the mobile target detection method based on a cell-free network includes steps S10-S40:

[0060] Step S10, determining n target detection periods based on the target variation degree, and obtaining predicted target positions in the n target detection periods.

[0061] It should be noted that the target variation degree can be used to measure the degree of change of the target in position, speed or motion mode, the target detection period can be understood as the time interval for one detection in the target detection process, and multiple periods can be used to continuously monitor the motion trajectory of the target. The predicted target position can be an estimation of the position of the target at a future time based on the current state of the target and the motion model.

[0062] Step S20, evaluating the channel state of each node to the predicted target position and the fusion center according to the predicted target position in the n target detection periods and the channel knowledge map.

[0063] It should be noted that the channel knowledge map can reflect the information atlas of the channel state between each node and the predicted target position and the fusion center, and be used to evaluate the signal transmission quality and related parameters such as signal attenuation. The fusion center can fuse the detection data of each node to finally obtain the position and movement vector of the target. The node can be understood as a device that acts as a signal receiver or transmitter in the system, usually referring to an access point or a sensor, which is used to detect the position and state of the target. The channel state can be understood as the transmission quality of each node to the current position of the target, which can include signal-to-noise ratio, attenuation, delay, channel capacity, etc.

[0064] Step S30, based on the fusion covariance and the channel knowledge map, obtaining the optimal node combination and the corresponding optimal period by calculating the normalized trace.

[0065] It should be noted that the normalized trace can be understood as a value obtained by fusing the channel data in the channel knowledge map and the fusion covariance, which is used to evaluate the effectiveness of different node combinations in target detection, and usually represents an error or signal quality index. The optimal node combination can be a set of nodes (or access points) selected in target detection, which can provide the best detection performance under the current channel conditions and target dynamic characteristics. The corresponding optimal period can be understood as the most suitable time period for detection based on the target variation degree and the channel conditions in a specific detection task.

[0066] Step S40, performing target detection based on the optimal node combination and the corresponding optimal period, and estimating the target position by fusing the node detection information obtained by target detection.

[0067] It should be noted that the node detection information includes the information detected by all nodes participating in the detection, and the detection information can be data about the state and position of the target detected by the nodes, which is used to update the motion state of the target.

[0068] In this embodiment, by determining the detection period based on the target change degree, the system can adaptively adjust the detection frequency according to the motion state of the target. In each detection period, by obtaining the predicted target position of the target, the system can more accurately locate the target. Combined with the prediction method of the state transition matrix, the predicted target position is more continuous and accurate, thereby improving the overall detection accuracy. Using the channel knowledge map to evaluate the channel state of each node to the target ensures that the node combination with the best channel state is selected in the detection process. By calculating the normalized trace to select the optimal node combination and the optimal detection period, the detection positioning accuracy is ensured while the resources are optimally configured. After each detection period ends, the target change degree is re-evaluated according to the updated target position, and the feedback mechanism ensures continuous and accurate tracking of the target position, ensuring the accuracy of the detection, and further significantly improving the communication efficiency.

[0069] Reference Figure 4 , Figure 4 is a flowchart of a second embodiment of the mobile target detection method based on a cell-free network of the present application, based on the first embodiment shown in the above Figure 3 The second embodiment of the mobile target detection method based on a cell-free network of the present application is proposed.

[0070] In the second embodiment, the step S10 comprises:

[0071] Step S101, calculating the target change degree according to the target direction change degree and the target speed change degree.

[0072] It should be noted that the target direction change degree can be the change amount of the direction of the target within a certain time period, usually expressed in degrees or radians, and the target speed change degree can represent the change of the speed of the target, usually calculated by the speed difference of the target at different time points. The target change degree can be used as a comprehensive index to consider the target direction change degree and the speed change degree.

[0073] For example, in each detection period, the detection nodes detect the target to obtain the position and movement vector of the target, and the data fusion center fuses the detection data of each node to finally obtain the position and movement vector of the target. Accordingly, the target direction change degree and the target speed change degree can be obtained, represented by formula (1) and formula (2) respectively.

[0074]

[0075]

[0076] In formula (1) and formula (2), and are the movement vectors of the target detected at time t and time t-T respectively, and Target moving rate detected at time t and time t-T, and is the speed detected by the system at time t and time t-T, the position and speed information of the next period are detected for the target. The initial speed is also detected by the system. o (t, t-T) is the target direction change degree, which represents the ratio of the angle between the speed direction at time t and time t-T to 2π. v o (t, t-T) is the target rate change degree, which represents the ratio of the difference between the speed at time t and time t-T to the speed at time t. The target change degree can be represented by formula (3):

[0077] S o (t, t-T) = θ o (t, t-T) + v o (t, t-T) (3)

[0078] S o (t, t-T) represents the target change degree, which is the sum of the target direction change degree and the target rate change degree. The greater the target change degree, the greater the target change, and the network detection needs to be more frequent.

[0079] Step S102, based on the target change degree, the value of pheromone is used to determine the maximum value of the detection period.

[0080] It should be noted that the value of pheromone is used to represent the detection requirement change degree, and the maximum value of the detection period is determined by the value of pheromone, wherein the maximum value of the detection period can be understood as the maximum detection time interval that can be sustained in a certain stage. With the movement of the target, the detection period needs to be adaptively changed. The detection requirement change degree becomes larger, and the detection period needs to be smaller, so that the target information can be detected in time; otherwise, the detection period needs to be larger, so as to save energy consumption. For example, the process of obtaining the detection requirement change degree is equivalent to the process of obtaining the pheromone in the ant colony algorithm, and the value of the detection requirement change degree is equivalent to the value of the pheromone.

[0081] Step S103, the maximum value of the detection period is divided to obtain n target detection periods.

[0082] For example, the maximum value of the detection period is represented as T max seconds, and 0-T max seconds is divided into n parts to obtain T j which can be represented by formula (4):

[0083]

[0084] In formula (4), T jThe jth target detection period is denoted as, and it can be understood that there are n target detection periods in total.

[0085] In step S104, the predicted target position in the n target detection periods is obtained based on the n target detection periods.

[0086] It should be noted that the predicted target position can be calculated according to the dynamic characteristics and motion model of the target in each detection period to obtain the expected position of the target in these periods.

[0087] In this embodiment, by calculating the target change degree, the system can evaluate the motion state of the target in real time, so as to dynamically adjust the detection period, ensure that the target can be effectively tracked in different situations, divide the maximum value of the detection period into n target detection periods, make the detection process more refined and accurate, and help to obtain multiple detection data in a short time. The maximum value of the detection period is determined by using the pheromone value, which can flexibly adjust the detection frequency according to the change of the detection demand, thereby improving the resource utilization rate and the detection accuracy.

[0088] In one embodiment, the step S102 comprises: obtaining an initial value of the pheromone, the initial value of the pheromone being used to reflect the change state of the target detection demand; updating the value of the pheromone in a decreasing manner according to a volatility coefficient, recording and updating the iteration number; and when the pheromone value decays to a set minimum threshold value, determining the maximum value of the detection period according to the current iteration number.

[0089] It should be noted that the initial value of the pheromone is used to represent the numerical value of the current detection demand of the target, which reflects the severity or urgency of the target change. The higher the pheromone value, the greater the demand for target detection. The volatility coefficient can be used as a parameter to control the decay rate of the pheromone, and the value range is usually between 0 and 1. The value of the pheromone can be the numerical value of the current recorded detection demand change state. With the passage of time, the value of the pheromone is updated according to the changes of the environment and the target. The iteration number can be the number of iteration operations in the process of updating the pheromone. Each time the pheromone is updated, the iteration number is increased by one. The set minimum threshold value can be a predefined value. When the pheromone value is below this value, it indicates that the detection demand has been reduced to the extent that the current detection period can be stopped.

[0090] For example, the initial value of the pheromone Q o (t) can be set as the change degree S o (t, t-T) perceived at the last time, denoted as formula (5):

[0091] Q o (t) = S o (t, t-T) (5)

[0092] From formula (5), we can see that the value of the pheromone in the previous test was Q. o (tT) equals the change in demand S from the previous detection. o (tT). Each time the pheromone is updated, its value decreases slightly, and this process iterates until it is completely evaporated. The rate of pheromone evaporation depends on the evaporation coefficient ρ, based on the pheromone Q obtained in this instance. o The value of (t) is the same as the previous pheromone value Q. o The result of the (tT) comparison is used to determine the pheromone update. The pheromone update is performed according to formula (6).

[0093] Q k (t)=(1-ρ) k Q o (t) (6)

[0094] In formula (6), k represents the number of pheromone renewals, and ρ represents the volatile coefficient, expressed as formula (7):

[0095]

[0096] In formula (7), when the detection demand changes little and is relatively stable, the volatilization time can be slightly longer, ρ μ The value of ρ2 can be slightly smaller; when the detection requirements vary moderately, the value of ρ2 can be set to be greater than that of ρ. μ Slightly larger; the detection requirements vary considerably, so the value of ρ3 can be set to be larger than ρ2. The range of α is (0,1).

[0097] When Q k ≈0.0001*Q o At this point, the pheromone can be considered to have completely evaporated. The number of iterations at this point is the new detection period T. max The new detection cycle T max This can be expressed as formula (8):

[0098]

[0099] It is understandable that the target detection update period T max Let's set it to the pheromone evaporation time. The faster the detection demand changes, the faster the pheromone evaporates, T... max A smaller value means a shorter detection interval; a slower change in detection demand means slower pheromone evaporation, T max The value is relatively larger, and the detection interval is longer. When the pheromone evaporates to a sufficiently low level, it is considered that the pheromone has evaporated completely, and the next target detection can be carried out.

[0100] In this embodiment, by obtaining the initial value of pheromone and dynamically updating the pheromone according to the detection demand changes, the system can adapt to the changes of the target state in real time, by setting the minimum threshold and the number of iterations to calculate the maximum value of the detection period, so that the detection period can be adjusted based on the actual demand, and by controlling the evaporation coefficient to control the decay rate of the pheromone, the system can effectively manage the detection resources. The value of the pheromone combined with the number of iterations provides data support for decision-making, improving the flexibility and response speed of detection.

[0101] In one embodiment, the step S104 comprises: obtaining the current position and the current motion state of the target within the target detection period; based on the current position and the current motion state of the target, predicting the corresponding predicted target position of the target in each detection period by a state transition matrix, to obtain the predicted target position in n target detection periods.

[0102] It should be noted that the node combination for target detection at k+1 time is usually determined at k time. For example, the position of the target at k+1 time is predicted at k time. Since the detection effective distance of the node is within a certain range, the nodes that can detect the target can be determined after the position at k+1 time is predicted. The maximum detection period T max is calculated, and the corresponding T j , the position at the next time can be predicted by formula (9). Assuming that the target is a slowly moving maneuvering point. For some complex motion models, a multi-model interaction method can be used. For simplicity, an approximate uniform speed model can be used, and the target state equation can be represented as formula (10). Formula (9) and formula (10) are as follows:

[0103]

[0104]

[0105] In formula (9) and formula (10), x t represents the target state at t time. x t = (x, v x , y, v y , z, v z ), wherein (x, y, z) is the position of the target; (v x , v y , v z ) is the speed in the x, y, and z directions. Wherein is the state transition matrix of the jth prediction. is the target state at t+T j time of the jth prediction, according to the predicted target position information of , the node detection range, it is determined which nodes can detect the target in the jth prediction.

[0106] In this embodiment, by obtaining the current position and motion state of the target and combining the state transition matrix for prediction, the position of the target in each detection period in the future can be more accurately estimated, and the effectiveness of detection can be improved. By updating the current position and motion state of the target in real time, the system can quickly adapt to the dynamic changes of the target, thereby realizing continuous tracking of the target and ensuring that the detection strategy is always consistent with the target behavior.

[0107] Referring to Figure 5 , Figure 5 The flowchart of the third embodiment of the mobile target detection method based on a cell-free network of the present application is based on the second embodiment shown in the above Figure 4 The third embodiment of the mobile target detection method based on a cell-free network of the present application is proposed.

[0108] In the third embodiment, the step S30 comprises:

[0109] Step S301, based on the fusion covariance and the channel knowledge map, the estimated error covariance matrix of each node combination is corrected by the channel state data obtained from the channel knowledge map, and the normalized trace of different node combinations is obtained.

[0110] It should be noted that the node selection can be based on the estimated error covariance matrix and the channel knowledge map. For example, in the system, the target state can be represented as formula (11) and formula (12):

[0111] x k+1 =Φ k x k +v k , k = 0, 1,..., (11)

[0112] y k =H k x k +W k , i = 0, 1,..., N (12)

[0113] wherein Φ k is the state transition matrix, x k is the target state at time k, v k is the process noise, y k is the observation vector, and H k is the state observation matrix.

[0114]

[0115]

[0116] In formula (13), The observations are represented as the values ​​corresponding to a single target. Formula (14) represents the state observation matrix, and formula (15) represents the measurement noise matrix. The observation noise W k The covariance can be expressed as formula (16):

[0117]

[0118] Formula (17) represents the observation noise covariance for a single target, where R k It is represented as the noise covariance matrix. Therefore, the standard result of Kalman filtering is expressed as Equation (18). The covariance representing the tracking error can be expressed as formula (19).

[0119]

[0120]

[0121] In formulas (18) and (19), x k / k The posterior estimate is represented by the fitness formula (20), x k / k-1 This is represented as the predicted state, specifically by formula (21), P k / k It is expressed as the posterior covariance, specifically as shown in formula (22). It is represented as the inverse matrix of the prediction error covariance matrix, and is expressed as in formula (23).

[0122]

[0123] P k / k =E[(x k / k -x k (x) k / k -x k )′|Y k ] (twenty two)

[0124] P k / k-1 =E[(x k / k-1 -x k (x) k / k-1 -x k )′|Y k-1 ] (twenty three)

[0125] Combining formulas (20) to (23) yields formula (24):

[0126]

[0127] Based on the above formulas, a prediction can be made to obtain formulas (25) to (28), which are expressed as follows:

[0128]

[0129]

[0130] The formula (29) is expressed by a time axis as follows:

[0131]

[0132] To further select the optimal combination from the alternative periods and nodes, the tracking effects of different period and node combinations are predicted. Since the estimation error covariance P of the extended Kalman filter can be calculated without obtaining the measurement information according to the above derivation conclusion, it means that the estimation error covariance matrix P of each node at the k+1 moment can be obtained at the k moment. k+1 / k+1 The selection of the node can be performed according to the estimation error covariance matrix. Considering the influence of the channel environment, the communication channel and the detection channel of each detection node can be obtained from the channel knowledge map, and then a modified fusion covariance is introduced, which can be expressed as formula (30) or formula (31).

[0133]

[0134] In the formula (30) and (31), represents the fusion covariance matrix of the optional N nodes at the predicted t+T j moment, wherein i is the i th node, C i represents the communication path loss from the i th node to the fusion center, M i represents the detection path loss from the i th node to the predicted target.

[0135] Considering that the fusion covariance in the embodiment has both position error and speed error, and the dimensions are not unified, a normalized trace S is introduced to define the advantages and disadvantages of the node, and the S of different period and node combinations is calculated, and the node combination and period with the smallest S are selected to participate in the detection and positioning in the next period. The normalized trace S can be expressed as formula (32):

[0136]

[0137] In the formula (32), p 11 represents the element in the first row and the first column of the fusion covariance matrix

[0138] In step S302, the node combination corresponding to the smallest normalized trace is regarded as the optimal node combination by comparing the normalized traces of the different node combinations, and the optimal period is the period corresponding to the optimal node combination.

[0139] ​For example, by calculating the next detection period and which nodes to perform detection in the next period, the calculation is performed in the fusion center, by comparing the normalized traces of different node combinations, the node combination corresponding to the minimum normalized trace is regarded as the optimal node combination, based on the optimal node combination, the fusion center sends instructions to the nodes participating in the detection in the next period, and the nodes perform detection in the next period. The results obtained by detection are sent to the fusion center, and the coordinate position of the target is located by calculation in the fusion center.

[0140] In this embodiment, by fusion calculation based on the channel knowledge map and the estimated error covariance matrix, the normalized traces of different node combinations are obtained, which can effectively evaluate the performance of each node combination in the target detection process. By comparing these normalized traces, the node combination corresponding to the minimum normalized trace is identified as the optimal node combination, and the corresponding period is identified as the optimal period, which ensures that the node combination with the best performance is selected in the target detection, thereby improving the accuracy and efficiency of detection, optimizing the configuration of resources, reducing redundant detection, and enhancing the overall performance of the system.

[0141] In one embodiment, based on the above embodiments and implementations, the step of evaluating the channel state of each node to the predicted target position and the fusion center based on the predicted target position and the channel knowledge map in n target detection periods includes: collecting actual channel data, inputting the actual channel data into the initial channel knowledge map for model training to obtain a target channel knowledge map; inputting the predicted target position information, fusion center information and node information in n target detection periods into the target channel knowledge map to obtain the channel state of each node to the predicted target position and the fusion center.

[0142] Exemplarily, the channel knowledge map can be used to derive the channel signal-to-noise ratio of the exploration channel, the channel knowledge map construction method based on the expectation maximization algorithm can be used to obtain the channel knowledge data of the local communication environment in any of the following ways: offline ray tracing simulation, offline field measurement, or online real-time measurement, the relevant channel knowledge can be statistically modeled according to the knowledge base, and the K-class parameters of the mixed statistical model can be estimated by using an iterative algorithm to construct the channel knowledge map reflecting the local signal propagation environment; when communication or positioning is needed, real-time position information can be obtained by using GPS, Beidou, cellular positioning, laser radar, and self-sensor positioning methods, and the channel knowledge of the target position can be obtained by using the previously constructed channel knowledge map and the inverse distance weighted method, so as to be used for environment perception and adaptive communication. Exemplarily, based on the geographic location information of the user terminal, the channel knowledge data reflecting the actual propagation environment of the local signal can be obtained in any of the following ways: offline ray tracing simulation, offline field measurement, or online real-time measurement, the statistical channel model parameters set can be increased from 1 to K according to the expert knowledge contained in the statistical channel model, and the K-class parameters of the mixed model can be estimated by using an iterative algorithm to construct the corresponding channel knowledge map.

[0143] In the embodiment, by obtaining the channel data of different nodes to the predicted target and the fusion center, the system can deeply understand the characteristics of the communication environment, the channel knowledge map provides a comprehensive view of the channel state between the nodes and the predicted target and the fusion center, helps the system to more effectively allocate communication and detection resources when detecting the target, and by constructing the channel knowledge map, the system can select the optimal node when detecting the target, improve the accuracy of target positioning and tracking, and reduce the influence of signal attenuation. The construction of the channel knowledge map enables the system to perform adaptive communication, optimize the working parameters of the nodes, and thus improve the communication quality and the reliability of target detection.

[0144] In one embodiment, the step of performing target detection based on the optimal node combination and the corresponding optimal period, and estimating the target position by fusing the node detection information obtained by target detection, includes: sending a detection instruction issued by the fusion center to the participating nodes, the detection instruction being used to specify the detection time and the participating nodes; performing target detection according to the detection instruction based on the optimal node combination and the corresponding optimal period to obtain node detection information, the node detection information including distance data and angle data of each participating node to the target; returning the node detection information to the fusion center, and performing data fusion on the node detection information to obtain fusion information; calculating the position information and the speed information of the target according to the fusion information, and estimating the target position according to the position information and the speed information of the target.

[0145] It should be noted that after the fusion center calculates the position information and the speed information according to the node detection information, the target position in different periods can be predicted, the nodes in the effective detection distance in each period are determined, the next network detection period and the nodes of the detection target are determined based on the estimated error covariance matrix of the signal-to-noise ratio correction in the channel knowledge map, and the fusion center reissues the detection time to the nodes that need to participate in the detection based on the updated detection period and the nodes of the detection target, and a new round of detection is performed. For example, the signaling flow of the mobile target detection method of the cell-free network is as shown in Figure 6 .

[0146] In the embodiment,

[0147] It should be noted that the above examples are only used for understanding the present application and do not constitute a limitation on the mobile target detection method of the cell-free network based on the present application. More forms of simple transformation based on this technical concept are within the protection scope of the present application.

[0148] The present application also provides a mobile target detection device based on a cell-free network. Please refer to Figure 7 , the mobile target detection device based on the cell-free network comprises:

[0149] The position prediction module 10 is configured to determine n target detection periods based on the target change degree and obtain the predicted target positions in the n target detection periods.

[0150] The channel evaluation module 20 is configured to evaluate the channel state of each node to the predicted target position and the fusion center according to the predicted target positions in the n target detection periods and the channel knowledge map.

[0151] The node optimization module 30 is configured to obtain the optimal node combination and the corresponding optimal period by calculating the normalized trace based on the fusion covariance and the channel knowledge map.

[0152] The target detection module 40 is configured to perform target detection based on the optimal node combination and the corresponding optimal period, and estimate the target position by fusing the node detection information obtained by target detection. The node detection information includes the information detected by all nodes participating in the detection.

[0153] The mobile target detection device based on the cell-free network provided in the application adopts the mobile target detection method based on the cell-free network in the above embodiment, and can solve the technical problem of low detection accuracy of the mobile target. Compared with the prior art, the mobile target detection device based on the cell-free network provided in the application has the same beneficial effects as the mobile target detection method based on the cell-free network provided in the above embodiment, and other technical features in the mobile target detection device based on the cell-free network are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0154] The application provides a computer readable storage medium having computer readable program instructions (i.e. computer programs) stored thereon for executing the mobile target detection method based on the cell-free network in the above embodiment.

[0155] The computer readable storage medium provided in the application may, for example, be a U disk, but is not limited to an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, system or device, or any combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection with one or more conductive wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present embodiment, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system or device. The program code contained on the computer readable storage medium can be transmitted by any suitable medium, including but not limited to electrical wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0156] The above computer readable storage medium can be included in the mobile target detection device based on the cell-free network; or can exist separately without being assembled into the mobile target detection device based on the cell-free network.

[0157] The computer readable storage medium described above carries one or more programs, when the one or more programs are executed by the mobile target detection device based on a non-cellular network, the mobile target detection device based on a non-cellular network is caused to: determine n target detection periods based on a target change degree, and obtain predicted target positions in the n target detection periods; evaluate channel states of each node to the predicted target positions and a fusion center according to the predicted target positions in the n target detection periods and a channel knowledge map; obtain an optimal node combination and a corresponding optimal period by calculating a normalized trace based on a fusion covariance and the channel knowledge map; perform target detection based on the optimal node combination and the corresponding optimal period, and estimate a target position by fusing node detection information obtained by target detection, the node detection information including information detected by all nodes participating in detection.

[0158] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0159] The flow diagrams and the block diagrams in the drawings are illustrations of architectures, functionalities, and operations of possible implementations of systems, methods, and computer program products according to various embodiments of present application. In this regard, each block in the flow diagrams or block diagrams can represent a module, a procedure, or a part of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may

[0160] The modules described in the embodiments of the present application can be implemented in the form of software or in the form of hardware. In some cases, the name of the module does not constitute a limitation on the module itself.

[0161] The readable storage medium provided by the present application is a computer readable storage medium, which stores computer readable program instructions (i.e., a computer program) for executing the above-mentioned mobile target detection method based on a cell-free network, and can solve the technical problem of low detection accuracy of a mobile target. Compared with the prior art, the computer readable storage medium provided by the present application has the same beneficial effects as the mobile target detection method based on a cell-free network provided by the above-mentioned embodiments, and will not be described here.

[0162] The present application also provides a computer program product comprising a computer program, which, when executed by a processor, implements the steps of the above-mentioned mobile target detection method based on a cell-free network.

[0163] The computer program product provided by the present application can solve the technical problem of low detection accuracy of a mobile target. Compared with the prior art, the computer program product provided by the present application has the same beneficial effects as the mobile target detection method based on a cell-free network provided by the above-mentioned embodiments, and will not be described here.

[0164] The above only describes some embodiments of the present application, and does not limit the patent scope of the present application. Any equivalent structural transformation, direct / indirect application in other related technical fields based on the technical concept of the present application, and the content of the specification and drawings are included in the patent protection scope of the present application.

Claims

1. A method for mobile target detection based on a cell-free network, characterized in that The method comprises: determining n target detection periods based on target variation degree, and obtaining predicted target positions in the n target detection periods; evaluating channel states of each node to the predicted target positions and a fusion center according to the predicted target positions in the n target detection periods and a channel knowledge map; obtaining optimal node combination and corresponding optimal period through calculating normalized trace based on fusion covariance and the channel knowledge map; performing target detection based on the optimal node combination and the corresponding optimal period, and estimating target positions through fusing node detection information obtained in the target detection, wherein the node detection information comprises information detected by all nodes participating in the detection.

2. The method of claim 1, wherein, The step of determining n target detection periods based on target variation degree and obtaining predicted target positions in the n target detection periods comprises: calculating target variation degree according to target direction variation degree and target speed variation degree; determining a maximum value of a detection period through a value of pheromone based on the target variation degree, wherein the value of pheromone is used to represent detection demand variation degree; dividing the maximum value of the detection period to obtain n target detection periods; obtaining predicted target positions in the n target detection periods based on the n target detection periods.

3. The method of claim 2, wherein, The step of determining a maximum value of a detection period through a value of pheromone based on the target variation degree comprises: obtaining an initial value of pheromone, wherein the initial value of pheromone is used to reflect a detection demand variation state of a target; updating the value of pheromone in a decreasing manner according to a volatilization coefficient, recording and updating iteration times; when the value of pheromone decays to a set minimum threshold value, determining the maximum value of the detection period according to the current iteration times.

4. The method of claim 2, wherein, The step of obtaining predicted target positions in the n target detection periods based on the n target detection periods comprises: obtaining a current position and a current motion state of a target in a target detection period; predicting corresponding predicted target positions of the target in each detection period through a state transition matrix based on the current position and the current motion state of the target, and obtaining the predicted target positions in the n target detection periods.

5. The method of claim 1, wherein, The step of obtaining optimal node combination and corresponding optimal period through calculating normalized trace based on fusion covariance and the channel knowledge map comprises: obtaining normalized traces of different node combinations by correcting an estimated error covariance matrix of each node combination through channel state data obtained from the channel knowledge map based on the channel knowledge map; regarding a node combination corresponding to a minimum normalized trace as the optimal node combination through comparing the normalized traces of the different node combinations, and regarding a period corresponding to the optimal node combination as the optimal period.

6. The method of any one of claims 1 to 5, wherein, The step of evaluating channel states of each node to predicted target positions and a fusion center according to predicted target positions in n target detection periods and a channel knowledge map comprises: collecting actual channel data, inputting the actual channel data into an initial channel knowledge map for model training, and obtaining a target channel knowledge map; The predicted target position information in n target detection periods, the fusion center information and the node information are input into the target channel knowledge map to obtain channel states of each node to the predicted target position and the fusion center.

7. The method of claim 1, wherein, The step of performing target detection based on the optimal node combination and the corresponding optimal period and estimating the target position by fusing the node detection information obtained by target detection comprises: sending a detection instruction issued by the fusion center to the participating nodes, the detection instruction being used to specify a detection time and the participating nodes; performing target detection according to the detection instruction based on the optimal node combination and the corresponding optimal period to obtain node detection information, the node detection information comprising distance data and angle data of each participating node to the target; returning the node detection information to the fusion center and performing data fusion on the node detection information to obtain fusion information; calculating target position information and speed information according to the fusion information and estimating the target position according to the target position information and speed information.

8. A mobile target detection apparatus based on a cell-free network, characterized by The device comprises: a position prediction module configured to determine n target detection periods based on a target variation degree and to obtain predicted target positions in the n target detection periods; a channel evaluation module configured to evaluate channel states of each node to the predicted target positions and a fusion center according to the predicted target positions in the n target detection periods and a channel knowledge map; a node optimization module configured to obtain an optimal node combination and a corresponding optimal period by calculating a normalized trace based on a fusion covariance and the channel knowledge map; a target detection module configured to perform target detection based on the optimal node combination and the corresponding optimal period and to estimate a target position by fusing node detection information obtained by target detection, the node detection information comprising information detected by all participating nodes.

9. A storage medium, characterized by The storage medium is a computer-readable storage medium, and the storage medium stores a computer program. When the computer program is executed by a processor, the steps of the mobile target detection method based on a cell-free network according to any one of claims 1 to 7 are implemented.

10. A computer program product, characterised in that, The computer program product comprises a computer program. When the computer program is executed by a processor, the steps of the mobile target detection method based on a cell-free network according to any one of claims 1 to 7 are implemented.

Images