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

Abstract

The invention provides a high-speed moving complex scene wireless channel joint clustering and tracking method and system, and belongs to the technical field of wireless communication, and the method comprises the steps: employing a clustering method based on a variational Bayesian Gaussian mixture model to construct an initial clustering model for multipath components at a plurality of previous moments; according to the adjacent historical moments and the global clustering model, performing combined clustering and tracking on the multipath components at the subsequent moments by adopting a distance-based judgment method; and correcting the result of the combined clustering and tracking. According to the method, an initial clustering model is constructed by utilizing multipath components at a plurality of previous moments, and identification of channel clusters in any shape is realized; realizing multipath tracking and finishing channel clustering at the same time by utilizing the relation of multipath components among several adjacent moments and a global situation; according to the method, only one-time clustering is needed, so that the identification time and complexity are effectively reduced; by adopting a correction algorithm, the identification precision can be effectively improved, and the rapid and accurate identification of the wireless channel cluster in a high-speed moving complex scene can be completed.

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 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 its coverage to high-speed mobile scenarios including rail transit and highways, which are characterized by high-speed vehicle movement and signal multipath effects. In such a complex high-speed mobile environment, accurately understanding the propagation characteristics of wireless channels is crucial to 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 and has been adopted by many organizations as a standard channel model solution for its advantage of balancing complexity and accuracy.

[0004] In a wireless channel model based on a cluster structure, multipath components with similar channel characteristic parameters can be defined as a cluster, such as power, delay, departure angle, and arrival angle. Therefore, accurately constructing a wireless channel model based on a cluster structure requires reasonably clustering the characteristic parameters of the channel multipath components. At present, some solutions have been proposed for the problem of wireless channel clustering. Among them, channel clustering methods based on machine learning have been widely used in this field. However, these methods are suitable for clustering non-time-varying channels, and it is difficult to track multipath clusters in time-varying dynamic scenarios. In recent years, some scholars have proposed some tracking methods for the tracking problem of multipath clusters in time-varying channels, but these methods often require clustering of multipath components at each moment, and the tracking effect is greatly affected by the clustering results of multipath components at each moment; or only track the multipath components at the previous moment, ignoring the connection between multipath components in historical moments, so that the tracking method effect cannot reach the best. 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 several 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 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 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; a new cluster number is assigned to the multipath component that has not found its cluster in the clustering model at adjacent historical moments, and the principle of assignment is that the multipath components with consistent matching clusters are assigned the same new cluster number; 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 to generate a new cluster is determined according to the distance threshold for the generation of the new cluster, and if not, 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, splitting and re-clustering the cluster, and merging the new cluster 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 delay-arrival angle space at the n-th time, 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 eccentric 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 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 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 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] Further, 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 between 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 multipath components at a plurality of previous 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 at adjacent historical moments; the multipath components at the current moment match a cluster with the closest distance in the cluster model at adjacent historical moments; calculating the average eccentric distance of each cluster at 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; a new cluster number is assigned to the multipath component that does not find its cluster in the clustering model at adjacent historical moments, and the principle of assignment is that the multipath components with consistent matching clusters are assigned the same new cluster number; 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 to generate a new cluster is determined according to the distance threshold for the generation of the new cluster, and if no new cluster is generated, the new cluster is merged with the nearest global cluster;

[0031] The correction module is used for joint clustering and tracking result correction, including: 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, splitting and re-clustering the cluster, and merging the new cluster 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 so that the electronic device executes 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 as follows: the identification of wireless channel clusters of any shape in high-speed moving and complex scenarios can be realized quickly and accurately, and effective support can be provided 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 the 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 accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying 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 4A schematic 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 is the 12th time 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 described in the embodiment of the present invention 15 ~t 17 Schematic diagram of the joint clustering and tracking results.

[0046] Fig. 9 The 15th to 17th time points t described 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] Fig.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, and 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 with the same or similar functions. The embodiments described below by the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0049] It should be understood by those skilled in the art 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 that 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 features, integers, steps, operations, elements and / or components, but does not exclude 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, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", 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 one or more embodiments or examples in a suitable manner. Different embodiments or examples described in this specification and features of different embodiments or examples may be combined and combined by those skilled in the art without contradiction.

[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, and 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 mobile complex scenarios, the method comprising the following steps: first, for the multipath components of the first several moments, a clustering method based on a variational Bayesian Gaussian mixture model is used to construct an initial clustering model; second, according to the adjacent historical moments and the global clustering model, a distance-based judgment method is used to perform joint clustering and tracking on the multipath components of the subsequent moments; 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 recognition of channel clusters of arbitrary shapes; uses the connection between the multipath components of the adjacent moments and the global to realize multipath tracking and complete channel clustering; the method only needs to perform clustering once, which effectively reduces the recognition time and complexity; the correction algorithm can effectively improve the recognition accuracy, and can complete the rapid and accurate recognition of wireless channel clusters in high-speed mobile 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 multipath components at a plurality of previous 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 at adjacent historical moments; the multipath components at the current moment match a cluster with the closest distance in the cluster model at adjacent historical moments; calculating the average eccentric distance of each cluster at 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; a new cluster number is assigned to the multipath component that does not find its cluster in the clustering model at adjacent historical moments, and the principle of assignment is that the multipath components with consistent matching clusters are assigned the same new cluster number; 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 to generate a new cluster is determined according to the distance threshold for the generation of the new cluster, and if no new cluster is generated, the new cluster is merged with the nearest global cluster;

[0060] The correction module is used for joint clustering and tracking result correction, including: 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, splitting and re-clustering the cluster, and merging the new cluster with the nearest global cluster.

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

[0062] S1: For the multipath components at the first several moments, a clustering method based on the Variational Bayesian Gaussian Mixture Model (VB-GMM) is used to construct the initial clustering model of the channel;

[0063] S2: 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;

[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 eccentric distance of each cluster at adjacent historical moments and use it as the distance decision threshold for joint clustering and tracking;

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

[0070] S205: assigning new cluster numbers to multipath components that have not been found to belong to a cluster in the clustering model at adjacent historical moments, with the principle of assigning 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: judging whether to generate a new cluster according to the distance threshold of the new cluster, and if not, merging 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 delay-arrival angle space at the n-th time, 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 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, It 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 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 established 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, 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 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, determining whether the global cluster needs correction. 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 between 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 selection of the number of multipath components for constructing the initial clustering model of the channel in step S1 is determined according to the actual clustering effect and shape. Considering the clustering complexity and clustering effect, the ratio should not exceed 50% and can 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, and the method includes: S1: for the multipath components of the first several moments, an initial clustering model of the channel is constructed by 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 selection of the number of multipath components for constructing the initial clustering model of the channel in step S1 is determined according to the actual clustering effect and shape. Considering the clustering complexity and clustering effect, the ratio should not exceed 50% and can 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, which is highly complex and has low accuracy. The present application clusters multipath components only at the initial part of the moment, and while utilizing the connection between multipath components at adjacent moments, it also considers the connection with multipath components at historical moments, making full use of the time-varying characteristics of channels in complex high-speed mobile scenarios, and using the clustering results at historical moments multiple times to jointly cluster and track multipath components at the next moment, effectively improving clustering and tracking accuracy.

[0101] Specifically, 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 a cluster with the nearest multipath component distance in the cluster model of the adjacent historical moment with the multipath component at the current moment; S203: calculating the average eccentric distance of each cluster at the adjacent historical moment and using it as the distance decision threshold for joint clustering and tracking; S204: judging whether the multipath component at the current moment belongs to the cluster according to 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 decision threshold for the generation of new clusters; S208: judging whether a new cluster is generated 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.

[0102] Specifically, 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 delay-arrival angle space at the n-th time, 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 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, It 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 the cluster center closest to it 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, and 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 5 clusters at the adjacent historical moment, and the cluster numbers are 1 to 5 respectively, if there are 3 multipath components a, b, and c at the current moment that need to be assigned new cluster numbers, and the a and b multipath components are closer to the cluster with cluster number 1, and c is closer to the cluster with cluster number 2, then a and b form a new cluster and are assigned a new cluster number 6, and c is assigned a new cluster number 7 alone.

[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 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 established 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, 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, and 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, determining whether the global cluster needs correction. 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 between When , the number of clusters is set to 3.

[0123] In order to verify the accuracy of the present application, the method further comprises the following steps: creating a simulation scenario in simulation software based on the ray tracing method and obtaining simulation results, and then comparing them with the simulation results to evaluate the accuracy of the joint clustering and tracking results. The clustering accuracy rate in the present embodiment is defined according to the simulation results, that is, the proportion of the correct cluster determined by the clustering method of the present embodiment for each multipath component compared with 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 scene contains multiple buildings with an average height of 6m. The heights of the transmitting antenna and the receiving antenna are 6m and 2m respectively. The simulation area is about 120m×65m. Sampling is performed every 2 meters. The wireless channel propagation parameters at 206 moments are stored through ray tracing simulation, namely 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] Step 1: For the multipath components at the first several moments, in this case, the multipath components at moments t1, t2, ..., t5, the VB-GMM clustering algorithm is used to construct the initial channel clustering model. The algorithm considers the mean and covariance information of the multipath components at the same time, and realizes the clustering of the multipath components in the delay domain and the angle domain. Clustering based on the Gaussian mixture model uses Gaussian distribution to describe the clustering structure, and the clustering division is determined by the posterior probability of the prototype. For multipath component data, the VB-GMM clustering algorithm assumes that these data are generated by a mixture model composed of multiple Gaussian distributions, and each Gaussian distribution corresponds to a cluster cluster. VB-GMM uses the variational Bayesian method for parameter estimation and inference. By introducing variational parameters to approximate the posterior distribution of unknown parameters, the complexity of parameter estimation can be effectively handled, and the number of clusters can be adaptively determined. Compared with the k-means clustering algorithm, the VB-GMM clustering algorithm considers the mean and covariance of the sample points at the same time, and is not only applicable to the case where the cluster is a circular distribution, but also to any other 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 for non-circular distributions.

[0130] Step 2: for each remaining multipath component at each time, 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 time t n-1 The distances of all cluster centers;

[0131]

[0132] in represents the coordinate of the i-th multipath component in the delay-arrival angle space at the n-th time, 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, a cluster with the closest adjacent historical moment is matched for each multipath component.

[0134] Step 3: Determine the adjacent historical time t according to formula (2) n-1 The average eccentric distance of each cluster is used as the distance decision 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 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, It 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 decision threshold of the cluster, then the multipath component is divided into the matching cluster, otherwise, continue to the next step. Figure 6 The cluster tracking effect at the sixth moment t6 is shown. 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 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 time t n-1 If the distance between the cluster center of the nearest cluster is less than the decision distance threshold for the generation of a new cluster, the new cluster is merged with the nearest global cluster, otherwise it is 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, and R k Represents the intra-cluster density variance ratio of the kth global cluster.

[0149] Step 10: Determine whether the global cluster needs correction based on the intra-cluster variance ratio of each global cluster calculated in step 9. Preferably, if the intra-cluster variance ratio of the global cluster is greater than 1, no correction is required. Otherwise, VB-GMM is used to re-cluster the global cluster. Preferably, if the intra-cluster density variance ratio is less than 1, the global cluster needs correction. When between, the number of clusters is set to 2; if between When , the number of clusters is set to 3. Figure 8 The figure shows the effect of joint clustering and tracking at the 15th to 17th moments. Since the intra-cluster density variance ratio calculated for the 10th cluster is less than 1 and Therefore, correction is needed, that is, the multipath components in the cluster are re-clustered with the condition that the number of clusters is 2. The corrected result is as follows Fig. 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 joint 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 number of all multipath components. Fig.10 As shown in FIG. 1 , 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 verify the joint clustering and tracking method proposed in this embodiment. The initial clustering model is constructed through the VB-GMM clustering algorithm, and the recognition of channel clusters of arbitrary shapes is realized. By utilizing the connection between multipath components at adjacent moments and globally, channel clustering is completed while multipath tracking is achieved, and the joint clustering and tracking results are corrected through the clustering correction algorithm, which effectively reduces the complexity and improves the accuracy. It provides a reference for the modeling of 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, wherein the non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the above-mentioned high-speed mobile complex scene wireless channel joint clustering and tracking method is implemented. The method includes:

[0154] S1: For the multipath components of the first few moments, a clustering method based on machine learning is used to build 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: Tracking clustering result correction.

[0155] Example 5

[0156] This embodiment 5 provides a computer device, including a memory and a processor, the processor and the memory communicate with each other, the memory stores program instructions executable by the processor, the processor calls the program instructions to execute the high-speed mobile complex scene wireless channel joint clustering and tracking method as described above, the method comprising:

[0157] S1: For the multipath components of the first few moments, a clustering method based on machine learning is used to build 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: Tracking clustering result correction.

[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, 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, so that the electronic device executes instructions for implementing the high-speed mobile complex scene wireless channel joint clustering and tracking method as described above, the method comprising:

[0160] S1: For the multipath components of the first few moments, a clustering method based on machine learning is used to build 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: Tracking clustering result correction.

[0161] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, 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 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 flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, 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 flowchart and / or block diagram. 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 capable of directing a computer or other programmable data processing device to operate 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 A 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 for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0165] Although the above describes the specific implementation mode 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 without creative work on the basis of the technical solution disclosed in the present invention 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 in that: include: For the multipath components at the first several 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 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 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; a new cluster number is assigned to the multipath component that has not found its cluster in the clustering model at adjacent historical moments, and the principle of assignment is that the multipath components with consistent matching clusters are assigned the same new cluster number; 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 to generate a new cluster is determined according to the distance threshold for the generation of the new cluster, and if not, 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, splitting and re-clustering the cluster, and merging the new cluster with the nearest global cluster.

2. The high-speed mobile complex scene wireless channel joint clustering and tracking method 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 delay-arrival angle space at the n-th time, 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 eccentric 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 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 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: The average minimum cluster center distance of the global cluster is calculated and used 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 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. When between, the number of clusters is set to 2; if between When , the number of clusters is set to 3.

7. A high-speed mobile complex scene wireless channel joint clustering and tracking system, characterized in that: include: A construction module is used to construct an initial channel clustering model for multipath components at a plurality of previous 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 at adjacent historical moments; the multipath components at the current moment match a cluster with the closest distance in the cluster model at adjacent historical moments; calculating the average eccentric distance of each cluster at 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; a new cluster number is assigned to the multipath component that does not find its cluster in the clustering model at adjacent historical moments, and the principle of assignment is that the multipath components with consistent matching clusters are assigned the same new cluster number; 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 to generate a new cluster is determined according to the distance threshold for the generation of the new cluster, and if no new cluster is generated, the new cluster is merged with the nearest global cluster; The correction module is used for joint clustering and tracking result correction, including: 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, splitting and re-clustering the cluster, and merging the new cluster 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 high-speed mobile complex scene wireless channel joint clustering and tracking method as described in any one of claims 1-6 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 as described in any one of claims 1-6.

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 so that the electronic device executes 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