High-speed moving complex scene wireless channel joint clustering and tracking method and system

Abstract

The application provides a high-speed moving complex scene wireless channel joint clustering and tracking method and system, and belongs to the wireless communication technical field, and comprises the following steps: for the multipath components of the previous several time points, an initial clustering model is constructed by using a clustering method based on a variational Bayesian Gaussian mixture model; according to the adjacent historical time points and the global clustering model, a distance-based decision method is used to perform joint clustering and tracking on the multipath components of the subsequent time points; and the joint clustering and tracking result is corrected. The application uses the multipath components of the previous several time points to construct an initial clustering model, realizes the identification of channel clusters of any shape, uses the connection between the multipath components of the adjacent several time points and the global, realizes the channel clustering while completing the multipath tracking, and the method only needs to be clustered once, effectively reduces the identification time and complexity, and the correction algorithm can effectively improve the identification precision and can complete the fast and accurate identification of the wireless channel clusters in the high-speed moving complex scene.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a method and system for joint clustering and tracking of wireless channels in high-speed mobile and complex scenarios. Background Art

[0002] With the continuous development of mobile communication technology, the widespread application of 6G communication networks in the future will expand their coverage to high-speed mobile scenarios such as rail transit and highways. These scenarios are characterized by high-speed vehicle movement and signal multipath effects. In such complex high-speed mobile environments, accurately understanding the propagation characteristics of wireless channels is crucial for optimizing communication systems and ensuring the stability of communication links.

[0003] For complex, time-varying, high-speed mobile scenarios, establishing an accurate channel model is a key step in describing the propagation characteristics of wireless channels. Cluster-based wireless channel modeling plays an important role in channel modeling. Its advantage in balancing complexity and accuracy has been adopted by many organizations as a standard channel model solution.

[0004] In a cluster-based wireless channel model, multipath components with similar channel characteristic parameters, such as power, delay, angle of departure, and angle of arrival, can be defined as a cluster. Therefore, accurately constructing a cluster-based wireless channel model requires properly clustering the characteristic parameters of the channel multipath components. Currently, several solutions have been proposed to address the wireless channel clustering problem. Among them, machine learning-based channel clustering methods have been widely used in this field. However, these methods are only suitable for clustering in time-invariant channels and have difficulty tracking multipath clusters in time-varying dynamic scenarios. In recent years, researchers have proposed tracking methods for multipath clusters in time-varying channels. However, these methods often require clustering multipath components at each moment, and the tracking effect is significantly affected by the clustering results at each moment. Alternatively, these methods only track the multipath components at the previous moment, ignoring the connections between multipath components at previous moments, resulting in suboptimal tracking performance. Summary of the Invention

[0005] The object of the present invention is to provide a method and system for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios, so as to solve at least one technical problem existing in the above-mentioned background technology.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios, comprising:

[0008] For the multipath components at the first few moments, a clustering method based on the variational Bayesian Gaussian mixture model is used to construct the initial channel clustering model;

[0009] The multipath components at subsequent moments track the global cluster model composed of the multipath components at adjacent historical moments and all past moments for joint clustering and tracking; including: calculating the distance between all multipath components at the current moment and all cluster centers in the adjacent historical moments; the multipath components at the current moment match a cluster with the closest distance in the cluster model of the adjacent historical moments; calculating the average eccentric distance of each cluster at the adjacent historical moments and using it as the distance judgment threshold for joint clustering and tracking; judging whether the multipath components at the current moment belong to the matching cluster based on the distance threshold for joint clustering and tracking, wherein all multipath components at the current moment are judged based on the determined joint clustering and tracking judgment threshold, if a certain If the distance between a multipath component and its nearest cluster center at an adjacent historical moment is less than the joint clustering and tracking distance decision threshold of the matching cluster, the multipath component is included in the matching cluster; otherwise, a new cluster number is assigned; new cluster numbers are assigned to multipath components that do not find their corresponding clusters in the clustering model at adjacent historical moments, with the principle of assigning the same new cluster number to multipath components with the same matching cluster; the distance between the cluster center of each new cluster and all cluster centers at adjacent historical moments is calculated; the average minimum cluster center distance of the global cluster is calculated and used as the distance decision threshold for the generation of new clusters; and whether a new cluster is generated is determined based on the distance threshold for the generation of the new cluster. If a new cluster is not generated, the new cluster is merged with the nearest global cluster;

[0010] The joint clustering and tracking result correction includes: calculating the ratio of the intra-cluster density to the intra-cluster variance of each global cluster; judging whether the intra-cluster density variance ratio of each global cluster is less than 1. If so, the cluster is split and re-clustered, and the new cluster is merged with the nearest global cluster.

[0011] Furthermore, the distances between all multipath components at the current moment and all cluster centers at adjacent historical moments are calculated as:

[0012]

[0013] in, represents the coordinate of the i-th multipath component in the time delay-arrival angle space at the n-th moment, represents the coordinates of the cluster center of the jth cluster at the n-1th time in the time delay-arrival angle space, is the distance between the point corresponding to the i-th multipath component at the n-th moment and the cluster center of the j-th cluster at the n-1-th moment;

[0014] Calculate the average centrifugal distance of each cluster at the current moment and use it as the distance threshold for joint clustering and tracking:

[0015]

[0016] in, represents the coordinate of the i-th multipath component in the j-th cluster at the n-1-th time in the time delay-arrival angle space, represents the cluster center of the jth cluster at the n-1th time, m represents the number of multipath components in the jth cluster at the n-1th time, represents the average centrifugal distance of the jth cluster at the n-1th moment;

[0017] The multipath component distance between the cluster center of each new cluster and all cluster centers at adjacent historical moments is calculated as:

[0018]

[0019] in represents the coordinates of the cluster center of the kth new cluster at the nth time in the time delay-arrival angle space, represents the cluster center of the jth cluster at the n-1th moment, It represents the distance between the center of the kth new cluster at the nth moment and the center of the jth cluster at the n-1th moment.

[0020] Furthermore, the average minimum cluster center distance of the global cluster is calculated and used as the decision distance threshold for generating new clusters:

[0021]

[0022] Among them, c i and c j They represent the cluster centers of the i-th and j-th global clusters respectively, and M represents the number of global clusters.

[0023] Furthermore, whether a new cluster is generated is determined based on the distance threshold for the generation of the new cluster. If no new cluster is generated, the new cluster is merged with the nearest global cluster, including: determining whether the new cluster generated at the current moment is established based on the determined judgment distance threshold for the generation of the new cluster; if the distance between the cluster center of the new cluster and the cluster center of the nearest cluster at the adjacent historical moment is less than the distance judgment threshold for the generation of the new cluster, the new cluster is merged with the nearest global cluster; otherwise, it is considered to be a new cluster.

[0024] Furthermore, the intra-cluster density variance ratio of each global cluster is calculated as:

[0025]

[0026] where p i,k and p j,k denote the coordinates of the i-th and j-th multipath components in the k-th global cluster in the time delay-arrival angle space, m denotes the number of multipath components in the k-th global cluster, R k represents the intra-cluster density variance ratio of the kth global cluster.

[0027] Furthermore, it is determined whether the intra-cluster density variance ratio of each global cluster is less than 1. If so, the cluster is split and re-clustered, and the new cluster is merged with the nearest global cluster, including: according to the calculated intra-cluster variance ratio of each global cluster, it is determined whether the global cluster needs to be corrected. If the intra-cluster variance ratio of the global cluster is greater than 1, no correction is required, otherwise the global cluster is re-clustered. If the intra-cluster density variance ratio is less than 1, the global cluster is re-clustered. When between, the number of clusters is set to 2; if When , the number of clusters is set to 3.

[0028] In a second aspect, the present invention provides a high-speed mobile complex scene wireless channel joint clustering and tracking system, comprising:

[0029] A construction module is used to construct an initial channel clustering model for the multipath components at the previous several moments by using a clustering method based on a variational Bayesian Gaussian mixture model;

[0030] The cluster tracking module is used for the multipath components at subsequent moments to track the global cluster model composed of the multipath components at adjacent historical moments and all past moments for joint clustering and tracking; including: calculating the distance between all multipath components at the current moment and all cluster centers in the adjacent historical moments; matching the multipath components at the current moment with a cluster model with the closest distance in the adjacent historical moments; calculating the average centrifugal distance of each cluster at the adjacent historical moments and using it as the distance judgment threshold for joint clustering and tracking; judging whether the multipath components at the current moment belong to the matching cluster according to the distance threshold for joint clustering and tracking, wherein all multipath components at the current moment are judged according to the determined joint clustering and tracking judgment threshold. If the distance between a multipath component and its nearest cluster center at an adjacent historical moment is less than the joint clustering and tracking distance decision threshold of the matching cluster, the multipath component is included in the matching cluster; otherwise, a new cluster number is assigned; new cluster numbers are assigned to multipath components that do not find their corresponding clusters in the clustering model at adjacent historical moments, with the principle of assigning the same new cluster number to multipath components with the same matching cluster; the distance between the cluster center of each new cluster and all cluster centers at adjacent historical moments is calculated; the average minimum cluster center distance of the global cluster is calculated and used as the distance decision threshold for the generation of new clusters; whether a new cluster is generated is determined based on the distance threshold for the generation of the new cluster; if a new cluster is not generated, the new cluster is merged with the nearest global cluster;

[0031] The correction module is used to correct the clustering and tracking results. It includes: calculating the ratio of the intra-cluster density to the intra-cluster variance of each global cluster; judging whether the intra-cluster density variance ratio of each global cluster is less than 1. If so, the cluster is split and re-clustered, and the new cluster is merged with the nearest global cluster.

[0032] In a third aspect, the present invention provides a non-transitory computer-readable storage medium, which is used to store computer instructions. When the computer instructions are executed by a processor, the method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios as described in the first aspect is implemented.

[0033] In a fourth aspect, the present invention provides a computer device comprising a memory and a processor, wherein the processor and the memory communicate with each other, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the high-speed mobile complex scene wireless channel joint clustering and tracking method as described in the first aspect.

[0034] In a fifth aspect, the present invention provides an electronic device comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to execute instructions for implementing the high-speed mobile complex scene wireless channel joint clustering and tracking method as described in the first aspect.

[0035] The beneficial effects of the present invention are: it can quickly and accurately realize the identification of wireless channel clusters of arbitrary shapes in complex high-speed moving scenarios, and provide effective support for the construction of channel models in scenarios such as rail transit and highways.

[0036] Additional advantages of the present invention will be more clearly given in the following description or learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0038] Figure 1 This is a flow chart of the method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios according to an embodiment of the present invention.

[0039] Figure 2 This is a specific flow chart of performing joint clustering and tracking on a global cluster model composed of multipath components at adjacent historical moments and multipath components at all past moments according to an embodiment of the present invention.

[0040] Figure 3 This is a specific flow chart of the joint clustering and tracking result correction according to an embodiment of the present invention.

[0041] Figure 4Schematic diagram of a simulation scenario according to an embodiment of the present invention.

[0042] Figure 5 This is a schematic diagram of clustering multipath components at the first five time points t1 to t5 using a Gaussian mixture model according to an embodiment of the present invention.

[0043] Figure 6 This is a schematic diagram of multipath component clustering after the joint clustering and tracking method is adopted at the sixth time t6 according to an embodiment of the present invention.

[0044] Figure 7 The 12th moment t described in the embodiment of the present invention 12 Schematic diagram of the emergence of new clusters.

[0045] Figure 8 The 15th to 17th time points t in the embodiment of the present invention 15 ~t 17 Schematic diagram of joint clustering and tracking results.

[0046] Figure 9 The 15th to 17th time points t in the embodiment of the present invention 15 ~t 17 Below is a schematic diagram of the correction of the joint clustering and tracking results.

[0047] Figure 10 This is a schematic diagram comparing the accuracy of the method described in the embodiment of the present invention and the k-means method. DETAILED DESCRIPTION

[0048] The embodiments of the present invention are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention and are not to be construed as limiting the present invention.

[0049] Those skilled in the art will understand that unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art to which this invention belongs.

[0050] It should also be understood that terms, such as those defined in commonly used dictionaries, should be understood to have a meaning consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless as defined herein.

[0051] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a," "an," "said," and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the stated features, integers, steps, operations, elements, and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, and / or groups thereof.

[0052] In the description of this specification, reference to the terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Those skilled in the art may combine and integrate different embodiments or examples described in this specification, as well as features of different embodiments or examples, unless otherwise contradictory.

[0053] To facilitate understanding of the present invention, the present invention is further explained below with reference to specific embodiments in conjunction with the accompanying drawings. However, the specific embodiments do not constitute a limitation on the embodiments of the present invention.

[0054] Those skilled in the art should understand that the drawings are merely schematic diagrams of embodiments, and the components in the drawings are not necessarily necessary for implementing the present invention.

[0055] The present invention provides a method for joint clustering and tracking of wireless channels in high-speed and complex scenarios. The method comprises the following steps: first, for the multipath components of the first several moments, an initial clustering model is constructed using a clustering method based on a variational Bayesian-Gaussian mixture model; second, based on adjacent historical moments and a global clustering model, a distance-based judgment method is used to jointly cluster and track the multipath components at subsequent moments; and finally, the results of the joint clustering and tracking are corrected. The present invention uses the multipath components of the first several moments to construct an initial clustering model, which can realize the identification of channel clusters of arbitrary shapes; utilizes the connection between the multipath components of several adjacent moments and the global to realize multipath tracking and channel clustering simultaneously; the method only requires clustering once, effectively reducing the identification time and complexity; and the correction algorithm can effectively improve the identification accuracy, and can complete the rapid and accurate identification of wireless channel clusters in high-speed and complex scenarios.

[0056] Example 1

[0057] In this embodiment 1, a high-speed mobile complex scene wireless channel joint clustering and tracking system is first provided, including:

[0058] A construction module is used to construct an initial channel clustering model for the multipath components at the previous several moments by using a clustering method based on a variational Bayesian Gaussian mixture model;

[0059] The cluster tracking module is used for the multipath components at subsequent moments to track the global cluster model composed of the multipath components at adjacent historical moments and all past moments for joint clustering and tracking; including: calculating the distance between all multipath components at the current moment and all cluster centers in the adjacent historical moments; matching the multipath components at the current moment with a cluster model with the closest distance in the adjacent historical moments; calculating the average centrifugal distance of each cluster at the adjacent historical moments and using it as the distance judgment threshold for joint clustering and tracking; judging whether the multipath components at the current moment belong to the matching cluster according to the distance threshold for joint clustering and tracking, wherein all multipath components at the current moment are judged according to the determined joint clustering and tracking judgment threshold. If the distance between a multipath component and its nearest cluster center at an adjacent historical moment is less than the joint clustering and tracking distance decision threshold of the matching cluster, the multipath component is included in the matching cluster; otherwise, a new cluster number is assigned; new cluster numbers are assigned to multipath components that do not find their corresponding clusters in the clustering model at adjacent historical moments, with the principle of assigning the same new cluster number to multipath components with the same matching cluster; the distance between the cluster center of each new cluster and all cluster centers at adjacent historical moments is calculated; the average minimum cluster center distance of the global cluster is calculated and used as the distance decision threshold for the generation of new clusters; whether a new cluster is generated is determined based on the distance threshold for the generation of the new cluster; if a new cluster is not generated, the new cluster is merged with the nearest global cluster;

[0060] The correction module is used to correct the clustering and tracking results. It includes: calculating the ratio of the intra-cluster density to the intra-cluster variance of each global cluster; judging whether the intra-cluster density variance ratio of each global cluster is less than 1. If so, the cluster is split and re-clustered, and the new cluster is merged with the nearest global cluster.

[0061] In this embodiment, the above-mentioned system is used to implement a method for joint clustering and tracking of wireless channels in high-speed mobile and complex scenarios, including:

[0062] S1: For the multipath components at the first few moments, an initial channel clustering model is constructed using a clustering method based on the Variational Bayesian Gaussian Mixture Model (VB-GMM).

[0063] S2: The multipath components at subsequent moments track the global cluster model composed of the multipath components of adjacent historical moments and all past moments for joint clustering and tracking;

[0064] S3: Joint clustering and tracking result correction;

[0065] Step S2 includes:

[0066] S201: Calculate the distances between all multipath components at the current moment and all cluster centers at adjacent historical moments;

[0067] S202: The multipath component at the current moment matches a cluster with the closest distance in the cluster model of the adjacent historical moment;

[0068] S203: Calculate the average centrifugal distance of each cluster at adjacent historical moments and use it as the distance decision threshold for joint clustering and tracking;

[0069] S204: Determine whether the multipath component at the current moment belongs to a matching cluster based on the joint clustering and tracking distance threshold;

[0070] S205: assigning new cluster numbers to multipath components that do not find their corresponding clusters in the clustering model at adjacent historical moments. The principle of assignment is to assign the same new cluster number to multipath components that match the same cluster.

[0071] S206: Calculate the distance between the cluster center of each new cluster and all cluster centers at adjacent historical moments;

[0072] S207: Calculate the average minimum cluster center distance of the global cluster and use it as the distance decision threshold for generating new clusters;

[0073] S208: Determine whether a new cluster is generated according to the distance threshold for generating the new cluster; if no new cluster is generated, merge the new cluster with the nearest global cluster.

[0074] Step S3 includes:

[0075] S301: Calculate the ratio of the intra-cluster density to the intra-cluster variance of each global cluster;

[0076] S302: Determine whether the intra-cluster density variance ratio of each global cluster is less than 1. If so, split the cluster and re-cluster it, and merge the new cluster with the nearest global cluster.

[0077] In step S201, the distances between all multipath components at the current moment and all cluster centers at adjacent historical moments are calculated according to the following formula (1):

[0078]

[0079] in represents the coordinate of the i-th multipath component in the time delay-arrival angle space at the n-th moment, represents the coordinates of the cluster center of the jth cluster at the n-1th time in the time delay-arrival angle space, is the distance between the point corresponding to the i-th multipath component at the n-th moment and the cluster center of the j-th cluster at the n-1-th moment.

[0080] In step S203, the average centrifugal distance of each cluster at the current moment is calculated according to the following formula (2) and used as the distance threshold for joint clustering and tracking:

[0081]

[0082] in represents the coordinate of the i-th multipath component in the j-th cluster at the n-1-th time in the time delay-arrival angle space, represents the cluster center of the jth cluster at the n-1th time, m represents the number of multipath components in the jth cluster at the n-1th time, represents the average centrifugal distance of the jth cluster at the n-1th moment.

[0083] Step S204 includes: judging all multipath components at the current moment according to the joint clustering and tracking decision threshold determined in step S203; if the distance between a multipath component and its nearest cluster center at an adjacent historical moment is less than the joint clustering and tracking distance decision threshold of the matching cluster, then including the multipath component in the matching cluster; otherwise, assigning a new cluster number.

[0084] In step S206, the multipath component distance between the cluster center of each new cluster and all cluster centers at adjacent historical moments is calculated according to the following formula (3):

[0085]

[0086] in represents the coordinates of the cluster center of the kth new cluster at the nth time in the time delay-arrival angle space, represents the cluster center of the jth cluster at the n-1th moment, It represents the distance between the center of the kth new cluster at the nth moment and the center of the jth cluster at the n-1th moment.

[0087] In step S207, the average minimum cluster center distance of the global cluster is calculated according to the following formula (4) and used as the decision distance threshold for generating new clusters:

[0088]

[0089] Among them, c i and c j They represent the cluster centers of the i-th and j-th global clusters respectively, and M represents the number of global clusters.

[0090] Step S208 includes: judging whether the new cluster generated at the current moment is valid according to the judgment distance threshold of the new cluster generated in step S207; if the distance between the cluster center of the new cluster and the cluster center of the nearest cluster at the adjacent historical moment is less than the distance judgment threshold of the new cluster generation, the new cluster is merged with the nearest global cluster; otherwise, it is considered to be a new cluster.

[0091] In step S301, the intra-cluster density variance ratio of each global cluster is calculated according to the following formula (5):

[0092]

[0093] where p i,k and p j,k denote the coordinates of the i-th and j-th multipath components in the k-th global cluster in the time delay-arrival angle space, m denotes the number of multipath components in the k-th global cluster, R k represents the intra-cluster density variance ratio of the kth global cluster.

[0094] Step S302 includes: according to the intra-cluster variance ratio of each global cluster calculated in S301, it is determined whether the global cluster needs to be corrected. If the intra-cluster variance ratio of the global cluster is greater than 1, no correction is required. Otherwise, the global cluster is re-clustered. When between, the number of clusters is set to 2; if When , the number of clusters is set to 3.

[0095] In this embodiment, the method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios also includes: creating a simulation scenario in simulation software based on ray tracing and obtaining channel simulation results, and then comparing them with the simulation results to evaluate the accuracy of the joint clustering and tracking results.

[0096] The number of multipath components used in constructing the initial channel clustering model in step S1 is determined based on the actual clustering effect and shape. Considering the clustering complexity and clustering effect, the ratio should not exceed 50% and should be able to better outline the shape of the channel cluster.

[0097] Example 2

[0098] like Figure 1 As shown, in this embodiment 2, a method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios is provided, which includes: S1: for the multipath components of the previous several moments, an initial channel clustering model is constructed using a clustering method based on machine learning; S2: the multipath components at subsequent moments track the global cluster model composed of the multipath components of adjacent historical moments and all past moments for joint clustering and tracking; S3: tracking and clustering result correction.

[0099] The number of multipath components used to construct the initial channel clustering model in step S1 is determined based on the actual clustering effect and shape. Considering the clustering complexity and clustering effect, the ratio should not exceed 50% and should be able to better outline the shape of the channel cluster.

[0100] As mentioned above, existing methods use clustering algorithms to cluster multipath components at each discrete moment, or only track multipath components at adjacent historical moments, resulting in high complexity and low accuracy. This application clusters multipath components only at the initial portion of moments, and while leveraging the connections between multipath components at adjacent moments, it also considers the connections with multipath components at historical moments. This fully exploits the time-varying characteristics of channels in complex high-speed mobile scenarios, and multiple times uses the clustering results from historical moments to jointly cluster and track the multipath components at the next moment, effectively improving clustering and tracking accuracy.

[0101] Specifically, if Figure 2 As shown, the step S2 includes: S201: calculating the multipath component distances between all multipath components at the current moment and all cluster centers at adjacent historical moments; S202: matching the multipath component at the current moment with a cluster with the nearest multipath component distance in the cluster model of the adjacent historical moment; S203: calculating the average centrifugal distance of each cluster at the adjacent historical moment and using it as the distance judgment threshold for joint clustering and tracking; S204: judging whether the multipath component at the current moment belongs to the cluster based on the distance threshold for joint clustering and tracking; S205: assigning a new cluster number to the multipath component that does not find its cluster in the cluster model of the adjacent historical moment, and assigning the same new cluster number to the multipath component that matches the same cluster; S206: calculating the multipath component distances between the cluster center of each new cluster and all cluster centers at adjacent historical moments; S207: calculating the average minimum cluster center distance of the global cluster and using it as the distance judgment threshold for the generation of a new cluster; S208: judging whether a new cluster is generated based on the distance threshold for the generation of the new cluster. If no new cluster is generated, the new cluster is merged with the nearest global cluster.

[0102] Specifically, if Figure 3 As shown, step S3 includes: S301: calculating the ratio of the intra-cluster density to the intra-cluster variance of each global cluster; S302: judging whether the intra-cluster density variance ratio of each global cluster is less than 1, if so, splitting and re-clustering the cluster, and merging the new cluster with the nearest global cluster.

[0103] In step S201, the multipath component distances between all multipath components at the current moment and all cluster centers at adjacent historical moments are calculated according to the following formula (1):

[0104]

[0105] in represents the coordinate of the i-th multipath component in the time delay-arrival angle space at the n-th moment, represents the coordinates of the cluster center of the jth cluster at the n-1th time in the time delay-arrival angle space, is the distance between the point corresponding to the i-th multipath component at the n-th moment and the cluster center of the j-th cluster at the n-1-th moment.

[0106] In step S203, the average centrifugal distance of each cluster at the current moment is calculated according to the following formula (2) and used as the distance threshold for joint clustering and tracking:

[0107]

[0108] in represents the coordinate of the i-th multipath component in the j-th cluster at the n-1-th time in the time delay-arrival angle space, represents the cluster center of the jth cluster at the n-1th time, m represents the number of multipath components in the jth cluster at the n-1th time, represents the average centrifugal distance of the jth cluster at the n-1th moment.

[0109] Step S204 includes: judging all multipath components at the current moment according to the joint clustering and tracking decision threshold determined in step S203; if the distance between a multipath component and its nearest cluster center at the adjacent historical moment is less than the joint clustering and tracking distance decision threshold of the cluster, then the multipath component is included in the matching cluster; otherwise, a new cluster number is assigned, and it is determined that a new cluster may be generated. The assignment of the new cluster number is based on which cluster these multipath components are closest to at the adjacent historical moment. For example, assuming that there are five clusters at the adjacent historical moment, with cluster numbers 1 to 5, if there are three multipath components a, b, and c at the current moment that need to be assigned new cluster numbers, and multipath components a and b are closer to cluster number 1, and c is closer to cluster number 2, then a and b form a new cluster and are assigned a new cluster number 6, and c is separately assigned a new cluster number 7.

[0110] In order to avoid tracking failure caused by the small correlation between the multipath components at the current moment and the adjacent historical moments, it is judged whether the multipath components in the new cluster can be tracked in the historical moments, that is, whether the new cluster is a new cluster compared with the global cluster at the historical moments.

[0111] In step S206, the multipath component distance between the cluster center of each new cluster and all cluster centers at adjacent historical moments is calculated according to the following formula (3):

[0112]

[0113] in represents the coordinates of the cluster center of the kth new cluster at the nth time in the time delay-arrival angle space, represents the cluster center of the jth cluster at the n-1th moment, It represents the distance between the center of the kth new cluster at the nth moment and the center of the jth cluster at the n-1th moment.

[0114] In step S207, the average minimum cluster center distance of the global cluster is calculated according to the following formula (4) and used as the distance decision threshold for generating new clusters:

[0115]

[0116] Among them, c i and c j They represent the cluster centers of the i-th and j-th global clusters respectively, and M represents the number of global clusters.

[0117] Step S208 includes: judging whether the new cluster generated at the current moment is valid according to the decision distance threshold of the new cluster generated in step S207; if the distance between the cluster center of the new cluster and the cluster center of the nearest cluster at the adjacent historical moment is less than the decision distance threshold of the new cluster generation, the new cluster is merged with the nearest global cluster; otherwise, it is considered to be a new cluster.

[0118] In order to further improve the clustering accuracy, in this embodiment, the results of joint clustering and tracking are corrected by calculating the intra-cluster density variance ratio.

[0119] In step S301, the intra-cluster density variance ratio of each global cluster is calculated according to the following formula (5):

[0120]

[0121] where p i,k and p j,k denote the i-th and j-th multipath components in the k-th global cluster, m denotes the number of multipath components in the k-th global cluster, R k represents the intra-cluster density variance ratio of the kth global cluster.

[0122] Step S302 includes: according to the intra-cluster variance ratio of each global cluster calculated in S301, it is determined whether the global cluster needs to be corrected. If the intra-cluster variance ratio of the global cluster is greater than 1, no correction is required. Otherwise, the global cluster is re-clustered. When between, the number of clusters is set to 2; if When , the number of clusters is set to 3.

[0123] To verify the accuracy of this application, the method further includes the following steps: creating a simulation scenario in ray tracing-based simulation software, obtaining simulation results, and then comparing these results with the simulation results to evaluate the accuracy of the combined clustering and tracking results. The clustering accuracy of this embodiment is defined based on the simulation results, i.e., the proportion of correct clusters determined for each multipath component by the clustering method of this embodiment, compared to the simulation results.

[0124] Example 3

[0125] In this embodiment 3, a simulation scenario is established in a simulation platform based on ray tracing method, such as Figure 4The simulation scenario includes multiple buildings with an average height of 6 meters. The heights of the transmitting and receiving antennas are 6 meters and 2 meters respectively. The simulation area is approximately 120 meters by 65 meters. Sampling is performed every 2 meters. The wireless channel propagation parameters at 206 moments, i.e., t1, t2, ..., t 206 The specific parameters of the simulation scenario are shown in Table 1.

[0126] Table 1 Configuration parameters for urban street scene simulation

[0127]

[0128] In this embodiment 3, the specific process of the high-speed mobile complex scene wireless channel joint clustering and tracking method implemented based on the above simulation scenario is as follows:

[0129] In step 1, the VB-GMM clustering algorithm is used to construct an initial channel clustering model for the multipath components at the first few time points (in this example, the multipath components at times t1, t2, ..., t5). This algorithm considers both the mean and covariance of the multipath components, enabling clustering in both the delay and angle domains. Clustering based on the Gaussian mixture model uses a Gaussian distribution to describe the cluster structure, and the division of clusters is determined by the posterior probability of the prototypes. For multipath component data, the VB-GMM clustering algorithm assumes that the data is generated by a mixture model composed of multiple Gaussian distributions, with each Gaussian distribution corresponding to a cluster. VB-GMM uses a variational Bayesian approach for parameter estimation and inference. By introducing variational parameters to approximate the posterior distribution of the unknown parameters, VB-GMM effectively handles the complexity of parameter estimation and allows for adaptive determination of the number of clusters. Compared to the k-means clustering algorithm, the VB-GMM clustering algorithm considers both the mean and covariance of the sample points. This makes it applicable not only to circular clusters but also to any non-circular distribution. Figure 5 The figure shows the clustering results of the VB-GMM algorithm for the first several moments. It can be seen that it also has a good clustering effect on non-circular distributions.

[0130] Step 2: For the multipath components at each remaining moment, i.e., t6, t7, ..., t n-1 ,t n ,t n+1 ,...,t 206 , calculate the current time t according to formula (1) n Each multipath component and the adjacent historical moment t n-1 The distances between all cluster centers;

[0131]

[0132] in represents the coordinate of the i-th multipath component in the time delay-arrival angle space at the n-th moment, represents the coordinates of the cluster center of the jth cluster at the n-1th time in the time delay-arrival angle space, is the distance between the point corresponding to the i-th multipath component at the n-th moment and the cluster center of the j-th cluster at the n-1-th moment.

[0133] According to the calculated distance, each multipath component is matched with a cluster closest to the adjacent historical moment.

[0134] Step 3: Determine the adjacent historical time t according to formula (2) n-1 The average centrifugal distance of each cluster is used as the distance judgment threshold for joint clustering and tracking:

[0135]

[0136] in represents the coordinate of the i-th multipath component in the j-th cluster at the n-1-th time in the time delay-arrival angle space, represents the cluster center of the jth cluster at the n-1th time, m represents the number of multipath components in the jth cluster at the n-1th time, represents the average centrifugal distance of the jth cluster at the n-1th moment.

[0137] Step 4: Determine the current time t n Whether the multipath component can be included in its corresponding matching cluster, if the distance between the multipath component and the cluster center of the nearest cluster is less than the distance judgment threshold of the cluster, then the multipath component is divided into the matching cluster, otherwise proceed to the next step. Figure 6 The figure shows the cluster tracking effect at the sixth moment t6. It can be seen that all components in t6 are divided into the initial cluster model, and no new clusters are generated at this time.

[0138] Step 5: Assign a new cluster number to the multipath components that are not successfully divided into clusters in step 4. The principle of assignment is to assign the same new cluster number to multipath components that match the same cluster.

[0139] Step 6: Calculate the current time t according to formula (3) n The cluster center of each new cluster and the adjacent historical moment t n-1 The distance between the cluster centers of all clusters when :

[0140]

[0141] in represents the coordinates of the cluster center of the kth new cluster at the nth time in the time delay-arrival angle space, represents the cluster center of the jth cluster at the n-1th moment, It represents the distance between the center of the kth new cluster at the nth moment and the center of the jth cluster at the n-1th moment.

[0142] Step 7: Calculate the average minimum cluster center distance of the global cluster according to formula (4) as the distance decision threshold for generating new clusters:

[0143]

[0144] Among them, c i and c j They represent the cluster centers of the i-th and j-th global clusters respectively, and M represents the number of global clusters.

[0145] Step 8: Determine the current time t according to the distance decision threshold of the new cluster generated by the new cluster determined in step 7. n Whether the generated new cluster is established, if the cluster center of the new cluster is at the adjacent historical moment t n-1 If the distance between the cluster center of the nearest cluster is less than the decision distance threshold for generating a new cluster, the new cluster will be merged with the nearest global cluster, otherwise it will be considered a new cluster. Figure 7 As shown, at the 12th moment t 12 , the physical environment has changed relative to the historical moment, and new buildings have appeared in the environment, resulting in a large change in the channel characteristic parameters, making it impossible to track adjacent historical moments or global clusters. At this time, it is judged that a new cluster has appeared.

[0146] Step 9: Calculate the intra-cluster density variance ratio according to formula (5):

[0147]

[0148] where p i,k and p j,k denote the i-th and j-th multipath components in the k-th global cluster, m denotes the number of multipath components in the k-th global cluster, R k represents the intra-cluster density variance ratio of the kth global cluster.

[0149] Step 10: Based on the intra-cluster variance ratio of each global cluster calculated in step 9, determine whether the global cluster needs to be corrected. Preferably, if the intra-cluster variance ratio of the global cluster is greater than 1, no correction is required. Otherwise, the global cluster is re-clustered using VB-GMM. Preferably, if the intra-cluster density variance ratio is less than 1, the global cluster is re-clustered. When between, the number of clusters is set to 2; if When , the number of clusters is set to 3. Figure 8 The results of joint clustering and tracking at the 15th to 17th moments are shown. Since the intra-cluster density variance ratio calculated for the 10th cluster is less than 1 and Therefore, correction is required, that is, the multipath components in the cluster are re-clustered with the number of clusters being 2. The corrected result is as follows Figure 9 As shown, it can be seen that after correction, the clusters are tighter and the clustering effect is better.

[0150] Finally, the accuracy of the combined clustering and tracking method was evaluated. The cluster to which each multipath component belongs was defined based on the simulation results obtained from the ray tracing simulation platform. The accuracy is the ratio of the number of correctly clustered multipath components to the total number of multipath components. Figure 10 As shown in Figure 3, the clustering effects of the joint clustering and tracking method and the k-means method at each moment are compared. It can be seen that the joint clustering and tracking method has better clustering accuracy.

[0151] The above experiments verified the joint clustering and tracking method proposed in this embodiment. The initial clustering model was constructed through the VB-GMM clustering algorithm, which enabled the identification of channel clusters of arbitrary shapes. By utilizing the connection between multipath components at several adjacent moments and globally, channel clustering was achieved while multipath tracking was performed. The joint clustering and tracking results were corrected through the cluster correction algorithm, effectively reducing complexity and improving accuracy. This provides a reference for modeling wireless channels based on scattering clusters in complex, time-varying, high-speed mobile scenarios in future 6G communication networks.

[0152] Example 4

[0153] This embodiment 4 provides a non-transitory computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios described above is implemented. The method includes:

[0154] S1: For the multipath components of the first few moments, a machine learning-based clustering method is used to construct an initial channel clustering model; S2: The multipath components at subsequent moments track the global cluster model composed of the multipath components of adjacent historical moments and all past moments for joint clustering and tracking; S3: The tracking clustering results are corrected.

[0155] Example 5

[0156] This embodiment 5 provides a computer device, including a memory and a processor, wherein the processor and the memory communicate with each other, the memory stores program instructions executable by the processor, and the processor invokes the program instructions to execute the above-described method for joint clustering and tracking of wireless channels in complex high-speed mobile scenarios. The method includes:

[0157] S1: For the multipath components of the first few moments, a machine learning-based clustering method is used to construct an initial channel clustering model; S2: The multipath components at subsequent moments track the global cluster model composed of the multipath components of adjacent historical moments and all past moments for joint clustering and tracking; S3: The tracking clustering results are corrected.

[0158] Example 6

[0159] This embodiment 6 provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to cause the electronic device to execute instructions for implementing the above-described high-speed mobile complex scenario wireless channel joint clustering and tracking method, the method including:

[0160] S1: For the multipath components of the first few moments, a machine learning-based clustering method is used to construct an initial channel clustering model; S2: The multipath components at subsequent moments track the global cluster model composed of the multipath components of adjacent historical moments and all past moments for joint clustering and tracking; S3: The tracking clustering results are corrected.

[0161] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0162] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0163] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0164] These computer program instructions can also be loaded onto a computer or other programmable data processing device, and a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide the functions for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0165] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solutions disclosed in the present invention without the need for creative work should be included in the scope of protection of the present invention.

Claims

1. A method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios, characterized by: include: For the multipath components at the first few moments, a clustering method based on the variational Bayesian Gaussian mixture model is used to construct the initial channel clustering model; The multipath components at subsequent moments track the global cluster model composed of the multipath components at adjacent historical moments and all past moments for joint clustering and tracking; including: calculating the distance between all multipath components at the current moment and all cluster centers in the adjacent historical moments; the multipath components at the current moment match a cluster with the closest distance in the cluster model of the adjacent historical moments; calculating the average eccentric distance of each cluster at the adjacent historical moments and using it as the distance judgment threshold for joint clustering and tracking; judging whether the multipath components at the current moment belong to the matching cluster based on the distance threshold for joint clustering and tracking, wherein all multipath components at the current moment are judged based on the determined joint clustering and tracking judgment threshold, if a certain If the distance between a multipath component and its nearest cluster center at an adjacent historical moment is less than the joint clustering and tracking distance decision threshold of the matching cluster, the multipath component is included in the matching cluster; otherwise, a new cluster number is assigned; new cluster numbers are assigned to multipath components that do not find their corresponding clusters in the clustering model at adjacent historical moments, with the principle of assigning the same new cluster number to multipath components with the same matching cluster; the distance between the cluster center of each new cluster and all cluster centers at adjacent historical moments is calculated; the average minimum cluster center distance of the global cluster is calculated and used as the distance decision threshold for the generation of new clusters; and whether a new cluster is generated is determined based on the distance threshold for the generation of the new cluster. If a new cluster is not generated, the new cluster is merged with the nearest global cluster; The joint clustering and tracking result correction includes: calculating the ratio of the intra-cluster density to the intra-cluster variance of each global cluster; judging whether the intra-cluster density variance ratio of each global cluster is less than 1. If so, the cluster is split and re-clustered, and the new cluster is merged with the nearest global cluster.

2. The method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios according to claim 1 is characterized in that: The distance between all multipath components at the current moment and all cluster centers at adjacent historical moments is calculated as: in, represents the coordinate of the i-th multipath component in the time delay-arrival angle space at the n-th moment, represents the coordinates of the cluster center of the jth cluster at the n-1th time in the time delay-arrival angle space, is the distance between the point corresponding to the i-th multipath component at the n-th moment and the cluster center of the j-th cluster at the n-1-th moment; Calculate the average centrifugal distance of each cluster at the current moment and use it as the distance threshold for joint clustering and tracking: in, represents the coordinate of the i-th multipath component in the j-th cluster at the n-1-th time in the time delay-arrival angle space, represents the cluster center of the jth cluster at the n-1th time, m represents the number of multipath components in the jth cluster at the n-1th time, represents the average centrifugal distance of the jth cluster at the n-1th moment; The multipath component distance between the cluster center of each new cluster and all cluster centers at adjacent historical moments is calculated as: in represents the coordinates of the cluster center of the kth new cluster at the nth time in the time delay-arrival angle space, represents the cluster center of the jth cluster at the n-1th moment, It represents the distance between the center of the kth new cluster at the nth moment and the center of the jth cluster at the n-1th moment.

3. The method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios according to claim 1 is characterized in that: Calculate the average minimum cluster center distance of the global cluster and use it as the decision distance threshold for generating new clusters: Among them, c i and c j They represent the cluster centers of the i-th and j-th global clusters respectively, and M represents the number of global clusters.

4. The method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios according to claim 1 is characterized in that: Whether a new cluster is generated is determined according to the distance threshold for the generation of the new cluster. If no new cluster is generated, the new cluster is merged with the nearest global cluster, including: determining whether the new cluster generated at the current moment is established according to the determined judgment distance threshold for the generation of the new cluster; if the distance between the cluster center of the new cluster and the cluster center of the nearest cluster at the adjacent historical moment is less than the distance judgment threshold for the generation of the new cluster, the new cluster is merged with the nearest global cluster; otherwise, it is considered to be a new cluster.

5. The method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios according to claim 1 is characterized in that: The intra-cluster density variance ratio of each global cluster is calculated as: where p i,k and p j,k denote the coordinates of the i-th and j-th multipath components in the k-th global cluster in the time delay-arrival angle space, m denotes the number of multipath components in the k-th global cluster, R k represents the intra-cluster density variance ratio of the kth global cluster.

6. The method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios according to claim 1 is characterized in that: Determine whether the intra-cluster density variance ratio of each global cluster is less than 1. If so, split the cluster and re-cluster it, and merge the new cluster with the nearest global cluster, including: according to the calculated intra-cluster variance ratio of each global cluster, determine whether the global cluster needs to be corrected. If the intra-cluster variance ratio of the global cluster is greater than 1, no correction is required, otherwise the global cluster is re-clustered; if the intra-cluster density variance ratio is less than 1, ... When between, the number of clusters is set to 2; if When , the number of clusters is set to 3.

7. A high-speed mobile complex scene wireless channel joint clustering and tracking system, characterized by: include: A construction module is used to construct an initial channel clustering model for the multipath components at the previous several moments by using a clustering method based on a variational Bayesian Gaussian mixture model; The cluster tracking module is used for the multipath components at subsequent moments to track the global cluster model composed of the multipath components at adjacent historical moments and all past moments for joint clustering and tracking; including: calculating the distance between all multipath components at the current moment and all cluster centers in the adjacent historical moments; matching the multipath components at the current moment with a cluster model with the closest distance in the adjacent historical moments; calculating the average centrifugal distance of each cluster at the adjacent historical moments and using it as the distance judgment threshold for joint clustering and tracking; judging whether the multipath components at the current moment belong to the matching cluster according to the distance threshold for joint clustering and tracking, wherein all multipath components at the current moment are judged according to the determined joint clustering and tracking judgment threshold. If the distance between a multipath component and its nearest cluster center at an adjacent historical moment is less than the joint clustering and tracking distance decision threshold of the matching cluster, the multipath component is included in the matching cluster; otherwise, a new cluster number is assigned; new cluster numbers are assigned to multipath components that do not find their corresponding clusters in the clustering model at adjacent historical moments, with the principle of assigning the same new cluster number to multipath components with the same matching cluster; the distance between the cluster center of each new cluster and all cluster centers at adjacent historical moments is calculated; the average minimum cluster center distance of the global cluster is calculated and used as the distance decision threshold for the generation of new clusters; whether a new cluster is generated is determined based on the distance threshold for the generation of the new cluster; if a new cluster is not generated, the new cluster is merged with the nearest global cluster; The correction module is used to correct the clustering and tracking results. It includes: calculating the ratio of the intra-cluster density to the intra-cluster variance of each global cluster; judging whether the intra-cluster density variance ratio of each global cluster is less than 1. If so, the cluster is split and re-clustered, and the new cluster is merged with the nearest global cluster.

8. A non-transitory computer-readable storage medium, characterized in that The non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by the processor, the method for joint clustering and tracking of wireless channels in high-speed mobile complex scenarios according to any one of claims 1 to 6 is implemented.

9. A computer device, characterized in that: It includes a memory and a processor, the processor and the memory communicate with each other, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the high-speed mobile complex scene wireless channel joint clustering and tracking method according to any one of claims 1 to 6.

10. An electronic device, characterized in that: include: A processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to execute instructions for implementing the high-speed mobile complex scene wireless channel joint clustering and tracking method as described in any one of claims 1-6.

Images