CN102098082A

CN102098082A - Channel cluster tracking method and device

Info

Publication number: CN102098082A
Application number: CN2009102423207A
Authority: CN
Inventors: 张建华; 董第; 张平; 董伟辉; 刘光毅
Original assignee: China Mobile Communications Group Co Ltd
Current assignee: China Mobile Communications Group Co Ltd
Priority date: 2009-12-11
Filing date: 2009-12-11
Publication date: 2011-06-15
Anticipated expiration: 2029-12-11
Also published as: CN102098082B

Abstract

The invention discloses a channel cluster tracking method. The method comprises the following steps of: acquiring multi-channel impulse response data and relative geographic information of two transceiving ends, extracting multi-channel space time multi-dimensional parameter information from the acquired multi-channel impulse response data, clustering multiple channels by using a preset clustering algorithm, acquiring space time multi-dimensional parameters of a clustered center, calculating multi-channel component distance (MCD) value of each cluster of adjacent moments, estimating the change range of the MCD value of each cluster according to the acquired relative geographic information, acquiring the MCD threshold value of each corresponding cluster, and executing cluster tracking according to the acquired MCD value of each cluster and the calculated MCD value of each cluster. The invention also discloses a channel cluster tracking device at the same time. The method and the device can solve the source problem of living clusters when the clusters are fused, and meanwhile, improve the precision of a tracking algorithm.

Description

Channel cluster tracking method and device

Technical Field

The present invention relates to wireless communication technologies, and in particular, to a method and an apparatus for tracking a channel cluster.

Background

With the development of wireless communication technology, in order to effectively carry massive multimedia services, the requirements on the transmission rate and the spectrum efficiency of a wireless communication system are higher and higher. The multiple-Input multiple-Output (MIMO) technology is a key technology of a new generation mobile communication system for improving the transmission rate and the spectrum efficiency of a wireless communication system, and the technology suppresses channel fading by using multiple antennas, thereby improving the capacity and the spectrum utilization rate of the wireless communication system by multiples without increasing the bandwidth. However, the performance of the MIMO technology is also limited by the space-time fading characteristics of the channel, and the wideband channel exhibits new frequency selection characteristics, so that establishing a suitable wideband MIMO channel model and studying the space-time frequency three-dimensional fading characteristics of the wideband MIMO channel are the premise and key for taking the advantages of the MIMO technology into full play.

In the modeling process, measurement of a broadband wireless channel can be performed in an actual geographic propagation environment, or based on a typical propagation environment, a large amount of measurement is performed to obtain measured data, the measured data is analyzed and extracted, characteristic parameters of the propagation environment are calculated, then modeling or model correction is performed on the channel propagation characteristics of the environment according to the characteristic parameters obtained by calculation, and finally a relatively perfect broadband MIMO channel model is obtained, so that a reference basis is provided for the transmission technology, resource management and network planning in a wireless communication system.

In a space-time-frequency three-dimensional fading broadband channel, multipath propagates in a cluster form, and generation, death and survival of the multipath are one of important characteristics of the broadband channel, and reflect the dynamic characteristics of the multipath on adjacent time segments. Thus, the dynamic nature of the reconstruction clusters is an indispensable aspect of wideband MIMO channel modeling. The current approach for studying the dynamic behavior characteristics is to acquire measured data based on the geographical propagation environment, extract clusters, and determine the cluster generation, death, and survival behaviors by using a tracking algorithm.

Fig. 1 is a flowchart illustrating a conventional channel cluster tracking method. Referring to fig. 1, the process includes:

step 101, extracting channel multipath parameters sampled at a series of time points by using a multi-dimensional parameter estimation algorithm;

step 102, clustering the multipath by using a clustering algorithm, and respectively acquiring cluster center parameters of each cluster;

103, calculating Euclidean space distances between cluster centers of clusters at adjacent moments;

and step 104, tracking the channel cluster according to the cluster center parameters of each cluster at adjacent time and a preset Multipath Component Distance (MCD) threshold value.

In this step, if the minimum distance between a certain cluster (old cluster) at the next moment and any cluster (new cluster) at the previous moment exceeds the MCD threshold, it is considered that a new cluster is generated, and a cluster label is assigned to the new cluster;

and if the minimum distance between a certain cluster at the previous moment and any cluster at the next moment is smaller than the MCD threshold, the cluster is considered to be alive, the original cluster label is used, otherwise, the cluster label is considered to be dead, the cluster label is invalid, and the channel cluster is not tracked any more.

As can be seen from the above description, the conventional channel cluster tracking method mainly has the following two problems:

1) MCD threshold setting lacks rationality. The existing method for setting the MCD threshold value gives a uniform empirical value to all clusters, the empirical value lacks a theoretical basis, and the MCD threshold values of all clusters are the same.

2) When the clusters are fused, the source problem of the surviving clusters cannot be solved, namely when the MCD values of some clusters at the current moment and some clusters at the previous moment are both smaller than the set MCD threshold value, the cluster label of the cluster is not solved, and the label of which old cluster is used, so that the problem of the cluster occurring in the dynamic evolution process cannot be processed.

Disclosure of Invention

In view of the above, the main objective of the present invention is to provide a channel cluster tracking method, which solves the problem of the source of the surviving clusters when the clusters are merged and improves the accuracy of the tracking algorithm.

Another object of the present invention is to provide a channel cluster tracking apparatus, which solves the problem of the source of a surviving cluster when a cluster is merged, and improves the accuracy of a tracking algorithm.

In order to achieve the above object, the present invention provides a channel cluster tracking method, including:

acquiring multipath channel impulse response data and relative geographic information of a transmitting end and a receiving end;

extracting multi-path space-time multi-dimensional parameter information from the acquired multi-path channel impulse response data;

clustering the multipath by using a preset clustering algorithm, and acquiring a space-time multidimensional parameter of a clustering center;

calculating the maximum neighboring distance MCD value of each cluster at adjacent time;

estimating the change range of the MCD value of each cluster according to the obtained relative geographic information, and obtaining the corresponding MCD threshold value of each cluster;

and executing cluster tracking according to the acquired cluster MCD threshold value and the calculated cluster MCD value.

The step of acquiring the multipath channel impulse response data comprises:

on a preset measuring line, under the condition that the reference directions of antenna arrays at the transmitting end and the receiving end are kept unchanged in the measuring process, channel impulse response data on a preset time period sequence are obtained.

The relative geographic information includes: linear distance, direction of motion and line of sight direction included angle.

And extracting multi-path space-time multi-dimensional parameter information from the acquired multi-path channel impulse response data through a space alternation generalized expectation maximization SAGE algorithm.

The multi-path space-time multidimensional parameter information specifically includes: multipath power, multipath delay, horizontal departure angle, and horizontal arrival angle.

The method for clustering the multipath by using the preset clustering algorithm and acquiring the space-time multidimensional parameters of the clustering center comprises the following steps:

clustering the multipath by using a KPowermeans algorithm according to the multipath power contained in the multipath space-time multidimensional parameter information, and acquiring the space-time multidimensional parameter of the corresponding cluster core according to the acquired space-time multidimensional parameter information of the multipath.

The space-time multidimensional parameters of the cluster center comprise cluster center power, cluster center time delay, a cluster center horizontal departure angle and a cluster center horizontal arrival angle.

The calculation formula for calculating the maximum neighboring distance MCD value of each cluster at the adjacent time is as follows:

in the formula, the subscript i, j represents the cluster center number at adjacent time, MCD_AoA，ijMCD value representing AOA of cluster center arrival angles at adjacent time instants, MCD_AoD，ijMCD value representing the departure angle AOD of adjacent moments of the cluster center, MCD_τ，ijIs the MCD value of the cluster time delay of the adjacent time instants.

In the formula,denotes the horizontal departure angle, phi denotes the horizontal arrival angle, tau is the time delay, and zeta is the time delay adjustment factor.

The step of estimating the change range of the MCD value of each cluster according to the obtained relative geographic information and obtaining the corresponding MCD threshold value of each cluster comprises the following steps:

establishing a geometric structure of a transmitting end position, a receiving end position, a primary scattering body position and a secondary scattering body position according to the acquired relative geographic information and the space-time multidimensional parameter of the cluster center;

acquiring the change range of the secondary scatterers corresponding to each cluster at the current moment and the MCD value of the receiving end according to the established geometric structure;

and estimating the change range of the MCD value of each cluster at the current moment according to the position of the scatterer and the displacement of the receiving end, and acquiring the corresponding MCD threshold value of each cluster.

The step of executing cluster tracking according to the obtained cluster MCD threshold value and the calculated cluster MCD value comprises:

and storing the new cluster with the MCD value smaller than the MCD threshold value of the old cluster into the candidate cluster set:

storing the old cluster with the MCD value smaller than the MCD threshold value of the old cluster into the aggregation cluster;

acquiring an MCD value of each cluster at adjacent time and an MCD threshold of an old cluster;

judging whether the MCD value of the new cluster at the adjacent moment is larger than the MCD threshold value of the corresponding old cluster, if so, allocating a new cluster number to the new cluster, otherwise, executing the following steps;

initializing aggregation cluster, setting variable initial value q of aggregation cluster number as 1, and aggregation cluster element number C_q＝M；

Traversing the old clusters of the aggregation cluster set and calculating the old clusters

Corresponding candidate cluster number L_kWherein Q is the number of aggregated clusters, if L_k0, old cluster k is dead cluster; if L is_kIf not, marking a new cluster with the minimum MCD value of the old cluster k in the aggregation cluster set in the candidate clusters, and reallocating the aggregation cluster set;

detecting whether an aggregation cluster set with the old cluster number larger than 1 exists, if so, traversing all candidate clusters corresponding to the aggregation cluster set with the old cluster number larger than 1, marking the candidate cluster element with the minimum MCD value as a survival cluster, deleting the candidate cluster element and the aggregation cluster element, updating the numbers of the aggregation cluster set and the aggregation cluster element, and returning to the step of traversing the old cluster of the aggregation cluster set; if not, marking the candidate cluster as a survival cluster, marking the rest clusters as new clusters, and allocating new cluster numbers to the new clusters.

An apparatus for channel cluster tracking, the apparatus comprising: a measurement module, a pre-processing module, an estimation module, and a tracking module, wherein,

the measuring module is used for acquiring multipath channel impulse response data and receiving and transmitting geographic information of two ends;

the preprocessing module is used for extracting multi-path space-time multidimensional parameter information from the acquired multi-path channel impulse response data, acquiring relative geographic information according to geographic information of the receiving end and the transmitting end, clustering the multi-paths by using a preset clustering algorithm and acquiring space-time multidimensional parameter information of a clustering center;

the estimation module is used for estimating the MCD threshold of each cluster according to the space-time multidimensional parameters of the cluster center and the relative geographic information;

and the tracking module calculates the MCD value of each cluster at the adjacent moment and tracks the clusters according to the obtained MCD threshold value of each cluster and the calculated MCD value of each cluster.

The preprocessing module comprises a parameter extraction sub-module, a clustering sub-module and a geographic information sub-module, wherein,

the parameter extraction submodule is used for receiving channel impulse response data, extracting space-time multidimensional parameter information containing multipath power, time delay and angle information according to a preset estimation algorithm and outputting the space-time multidimensional parameter information to the clustering submodule;

the clustering submodule is used for receiving the space-time multidimensional parameter information output by the parameter extraction submodule and clustering according to a preset clustering algorithm to obtain space-time multidimensional parameter information containing cluster core power, time delay and angle information;

and the geographic information submodule is used for carrying out format conversion on the received collected geographic data, carrying out interpolation and adjustment and obtaining relative geographic information.

The tracking sub-module includes a computation sub-module and a comparison and update sub-module, wherein,

the calculation submodule is used for calculating the MCD value of each cluster at the adjacent moment, generating an MCD matrix and outputting the MCD matrix to the comparison and update submodule;

and the comparison and update submodule is used for receiving the MCD matrix output by the calculation submodule and the MCD threshold value of each cluster output by the estimation module, and updating the state of the cluster according to the iteration target tracking algorithm set in advance.

The status of the clusters includes generation, death and survival of clusters, and the comparison and update submodule sets the numbers of the surviving clusters along with the old clusters, and the newly generated cluster numbers are arbitrarily designated from the numbers not yet assigned.

According to the technical scheme, the channel cluster tracking method and the channel cluster tracking device provided by the invention have the advantages that the multipath channel impulse response data and the relative geographic information of the receiving end and the transmitting end are obtained; extracting multi-path space-time multi-dimensional parameter information from the acquired multi-path channel impulse response data; clustering the multipath by using a preset clustering algorithm, and acquiring a space-time multidimensional parameter of a clustering center; calculating the MCD value of each cluster at adjacent time; estimating the change range of the MCD value of each cluster according to the obtained relative geographic information, and obtaining the corresponding MCD threshold value of each cluster; and executing cluster tracking according to the acquired cluster MCD threshold value and the calculated cluster MCD value. Therefore, the MCD threshold of each cluster is jointly estimated according to the position information, the motion state and the space-time multidimensional parameters of the cluster center at the receiving and sending ends, and each cluster has the corresponding MCD threshold, so that the method is closer to an actual propagation mechanism, the MCD threshold has reasonable physical significance and reliability, and the precision of a tracking algorithm is improved; and when cluster tracking is executed, a process of iteratively updating the cluster labels is introduced, so that the problem of cluster label distribution is solved, the phenomenon of cluster fusion can be more effectively dealt with, and the robustness of a tracking algorithm is improved.

Drawings

Fig. 1 is a flowchart illustrating a conventional channel cluster tracking method.

Fig. 2 is a flowchart illustrating a channel cluster tracking method according to an embodiment of the present invention.

Fig. 3 is a schematic geometric structure diagram of a change range of the secondary scatterers and the MCD values at the receiving end corresponding to each cluster at the current time according to the embodiment of the present invention.

Fig. 4 is a schematic geometric structure diagram of a variation range of the values of the MCDs in each cluster at the current time according to the embodiment of the present invention.

Fig. 5 is a flowchart illustrating a method for performing cluster tracking according to an embodiment of the present invention.

Fig. 6 is a flowchart illustrating a method for performing cluster tracking according to an embodiment of the present invention.

Fig. 7 is a schematic structural diagram of a channel cluster tracking apparatus according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.

Fig. 2 is a flowchart illustrating a channel cluster tracking method according to an embodiment of the present invention. Referring to fig. 2, the process includes:

step 201, acquiring multipath channel impulse response data and relative geographic information of a transmitting end and a receiving end;

in this step, on a preset measurement route, under the condition that the reference directions of the antenna arrays at the transmitting and receiving ends are kept unchanged in the measurement process, channel impulse response data on a preset time period sequence and geographic information or relative geographic information of the corresponding transmitting and receiving ends are obtained.

Preferably, a measurement route with a fixed direction is planned and the transmitting end is ensured to be static, so that the reference direction of the antenna arrays at the transmitting end and the receiving end is kept unchanged in the measurement process. And driving the receiving end antenna to move on the measuring line at a fixed speed, and acquiring channel impulse response data according to a channel sampling rate matched with the moving rate of the receiving end, namely acquiring channel impulse response data samples at a sampling rate of sampling 3 channel snapshots at least every Doppler period.

In practical application, a precise instant positioning system can be equipped at the receiving end or the transmitting end and the receiving end so as to record the relative geographic positions of the transmitting end and the receiving end in real time. For example, when the receiving end is in a moving state, a differential Global Positioning System (GPS) or a sliding rail may be used, and the sending end may adopt a point attaching manner or the receiving end may obtain specific location information of the sending end in advance, and calculate and obtain the relative geographic information according to the recorded location information before and after the sending end.

The relative geographic information includes: linear distance, motion direction, and line-of-sight direction included angle, etc.

In practical application, the linear distance can be determined according to longitude and latitude in recorded GPS information, the movement direction can be obtained from GPS information before and after the receiving end, and the angle between the movement direction and the linear distance is the visual distance direction included angle.

Step 202, extracting multi-path space-time multi-dimensional parameter information from the acquired multi-path channel impulse response data;

in this step, the channel impulse response data describes the propagation characteristics of the channel multipath, and the channel impulse response is obtained, so that the multi-path space-time multidimensional parameters can be extracted from the obtained multi-path channel impulse response data by using the existing high-precision estimation calculation method. The high-precision estimation algorithm may be a Space-Alternating Generalized Expectation-maximization (SAGE) algorithm, but may be other estimation algorithms.

The multi-path space-time multidimensional parameter information specifically includes: multipath power, multipath delay, horizontal departure angle, and horizontal arrival angle. It should be noted that, in the embodiment of the present invention, since the angles of departure and arrival of the multipath in the vertical direction do not greatly affect the present invention, they are not considered.

Step 203, clustering the multipath by using a preset clustering algorithm, and acquiring a space-time multidimensional parameter of a clustering center;

in this step, the preset clustering algorithm may be a KPowerMeans algorithm, and the KPowerMeans algorithm is used to cluster the multipaths according to the multipath powers included in the multipath space-time multidimensional parameter information, and the space-time multidimensional parameters of the corresponding cluster core, that is, the cluster core power, the cluster core delay, the cluster core horizontal departure angle, and the cluster core horizontal arrival angle, are obtained according to the obtained multipath space-time multidimensional parameter information. The space-time multidimensional parameter of the corresponding cluster core is obtained according to the space-time multidimensional parameter information of the multipath, and the calculation method is the prior art and is not described herein again.

Step 204, calculating the Maximum neighbor distance (MCD) value of each cluster at the adjacent time;

in this step, the adjacent time may be two adjacent sampling periods, and the calculation formula of the MCD value is:

in which the indices i, j represent the time of dayn and cluster center number at time n +1, i.e. cluster center number at preceding and succeeding times, MCD_AoA，ijAn MCD value representing an Angle Of Arrival (AOA) at a cluster center Of neighboring time points as a function Of the angle Of Arrival at the cluster center Of neighboring time points as a normalized spatial distance; MCD_AoD，ijAn MCD value representing the cluster center departure Angle (AOD) at adjacent times is a function of the cluster center departure angle at the adjacent times, and is a normalized time delay domain distance; MCD_τ，ijIs the MCD value of the cluster center delay of adjacent time instants as a function of the cluster center delay of adjacent time instants. Wherein,

in the formula,

the specific weight of the time delay domain in calculating the MCD value, which is used to change the time delay domain, can be adjusted according to engineering experience, and generally, ζ is 1. Thus, there is 0 ≦ MCD_ij≤1。

In the above formula, MCD_AoA，ij、MCD_AoD，ijAnd MCD_τ，ijCan be obtained according to the space-time multidimensional parameters of the cluster center.

Step 205, estimating the variation range of the MCD value of each cluster according to the obtained relative geographic information, and obtaining the corresponding MCD threshold value of each cluster;

in this step, the relative geographical position is obtained based on the geographical information at the transmitting and receiving ends, and the variation range of the MCD value of each cluster is estimated.

Specifically, first, a change range of the MCD value of the secondary scatterer and the receiver corresponding to each cluster at the current time is acquired.

Fig. 3 is a schematic geometric structure diagram of a change range of the secondary scatterers and the MCD values at the receiving end corresponding to each cluster at the current time according to the embodiment of the present invention. Referring to fig. 3, T and R denote a transmitting end and a receiving end, respectively; a1 and a2 represent positions of primary scatterers, and a represents a farthest position of the primary scatterers; b1 and B2 are positions of secondary scatterers, and B is the farthest position of the secondary scatterer; gamma ray₁，γ₂Represents the horizontal angle between the cluster and the line-of-sight propagation path when arriving with only one scatter; phi denotes the horizontal angle between the cluster and the line-of-sight propagation path at the time of arrival of the secondary scatter.

The geometry diagram in fig. 3 is based on two assumptions:

(1) the propagation of the signal experiences a maximum of two scatterers. In practical applications, since the reflection loss introduced by the scatterer is large, the probability of the occurrence of a third-order reflection component in the received signal is very small, and thus, the assumption is reasonable.

(2) The movement speed of the scatterer is far less than that of the mobile station and can be ignored. This is generally true in practical propagation environments.

In step 203, the cluster-center delay τ has been acquired, and therefore, the path length l ═ C · τ traveled by the cluster propagation can be calculated according to the cluster-center delay τ, where C is the speed of light and l can be represented as the total length of TA1B1R in fig. 3, or as the total length of TA2B 2R. I.e. l ═ l_TA1+l_A1B1+l_B1ROr l ═ l_TA2+l_A2B2+l_B2R。

For each cluster, referring to fig. 3, the farthest distance a of the primary scatterer can be preliminarily determined_maxLet TR ═ d, apply cosine law to calculate, arrange, can get at last:

similarly, the maximum distance b of the secondary scatterers can be calculated_max：

According to the actual measurement environment, the radius a of the non-scattering body area with the transmitting end and the receiving end as the origin can be respectively given_minAnd b_minCorresponding to TC and RD in fig. 3, respectively. Typically, in an indoor scenario, a may be set_min＝b_min1 m; in an outdoor scene, a_min＝10m，b_min5 m. Thus, the positions of the primary scatterer and the secondary scatterer are on DB and CA, respectively. Therefore, the unique distance value b from the secondary scatterer to the receiving end can be calculated according to the distance value a from the primary scatterer to the transmitting end.

Suppose that the primary scatterer is located at A₁Where the secondary scatterer is located at B₁And (3) arranging according to the cosine theorem, wherein the following steps are carried out:

wherein,

a obtained according to the above calculation_minAnd a_maxThe value range [ b ] of b can be calculated_a，min，b_a，max]Thus, it can be determined that the distance from the secondary scatterer to the sink should be in the interval β ═ max { b_a，min，b_min},b_a，max]And (4) the following steps.

And secondly, estimating the change range of the MCD value of each cluster at the current moment, and acquiring the corresponding MCD threshold value of each cluster.

Fig. 4 is a schematic geometric structure diagram of a variation range of the values of the MCDs in each cluster at the current time according to the embodiment of the present invention. Referring to fig. 4, T is the originating position; a1 and A2 are primary scatteringsA body position; b1 and B2 are secondary scatterer positions; r is the receiving end position;

and

respectively the adjacent time positions of the receiving ends.

Due to the fact that

It has been found that for the current time, when the secondary scatterer is located at max { b }_a，min，b_minAnd b_a，maxAt time, corresponding to maximum and minimum MCD respectively_AoAAnd (4) changing. For MCD_τIn clusters

For example, MCD_τIs dependent on

And

the path difference Δ d between

B_{1} B_{t_{0}} = b_{0},

B_{1} R_{t_{1}} = b_{1},

R_{t_{0}} R_{t_{1}} = s,

It is possible to obtain:

order to

<math><mrow><mfrac><mrow><mo>&PartialD;</mo><mi>Δd</mi></mrow><mrow><mo>&PartialD;</mo><msub><mi>b</mi><mn>0</mn></msub></mrow></mfrac><mo>≤</mo><mn>0</mn><mo>,</mo></mrow></math>

Obtaining:

s²(cos²α-1)≤0

the above formula is clearly true, and thus Δ d is b₀Monotonically increasing function of, i.e. MCD_τDecreases with increasing secondary scatterer distance b. Therefore, the range of variation epsilon of the value of each old cluster MCD decreases as b increases.

According to the formula:

in the formula, MCD_AoD＝0；

MCD_AoAThe range of values may be defined by b e β ═ max { b [_a，min，b_min}，b_a，max]Obtaining;

MCD_τthe range of values can be determined by Δ d, so that the range of variation ε of the old cluster MCD can be obtained

<math><mrow><mi>ϵ</mi><mo>&Element;</mo><mi>ϵ</mi><mo>=</mo><mo>[</mo><msub><mi>ϵ</mi><msub><mi>b</mi><mi>max</mi></msub></msub><mo>,</mo><msub><mi>ϵ</mi><msub><mi>b</mi><mi>min</mi></msub></msub><mo>]</mo><mo>.</mo></mrow></math>

In practical application, if s and b are considered₀Is relatively small, and is therefore calculated

And

often relatively close. For convenience, one may choose

Thus, the range of variation of the old cluster MCD value is a certain value, hereinafter referred to as the old cluster MCD threshold or MCD threshold.

And step 206, executing cluster tracking according to the acquired cluster MCD threshold value and the calculated cluster MCD value.

In this step, cluster tracking is performed, that is, the generation, death and survival of clusters, that is, the generation, death and survival of new clusters, old clusters are determined according to the obtained cluster MCD threshold values, the calculated cluster MCD values and a preset iterative tracking algorithm.

The following description will be made about the concept of the cluster according to the present invention.

And (3) new cluster generation: a new cluster (i.e., the cluster at the current time) whose MCD value is greater than the MCD threshold value of any old cluster (i.e., the cluster at the previous time);

candidate clustering: a new cluster set with an old cluster MCD value smaller than the old cluster MCD threshold value;

death cluster: when a certain old cluster does not have a corresponding candidate cluster, the old cluster is a dead cluster;

aggregation and clustering: an old cluster set with an MCD value of a new cluster being smaller than an MCD threshold of the old cluster;

survival clusters: when a candidate cluster of an old cluster has only one element and the old cluster is not in the aggregated cluster set, the old cluster and the new cluster corresponding to the old cluster are both called a surviving cluster.

Cluster numbering: an identifier of the cluster. The new cluster is assigned a new cluster number and the surviving clusters follow the old cluster number.

The candidate clusters and the aggregation clusters only appear in the process of the algorithm, and each old cluster and each new cluster only have one or more of new clusters, dead clusters and alive clusters at the end of the algorithm.

As can be seen from the above, in the channel cluster tracking method according to the embodiments of the present invention, by obtaining multipath channel impulse response data and relative geographic information at the transmitting end and the receiving end, space-time multidimensional parameter information of multipaths is extracted from the obtained multipath channel impulse response data, clustering the multipaths by using a preset clustering algorithm, obtaining space-time multidimensional parameters of cluster centers, calculating MCD values of clusters at adjacent times, estimating a variation range of MCD values of the clusters according to the obtained relative geographic information, obtaining MCD threshold values of corresponding clusters, and performing cluster tracking according to the obtained MCD threshold values of the clusters and the calculated MCD values of the clusters. Therefore, the position information, the motion state and the space-time multidimensional parameters of the cluster center at the receiving and sending ends are considered, the MCD threshold of each cluster is jointly estimated, and each cluster has the corresponding MCD threshold, so that the method is closer to an actual propagation mechanism, the MCD threshold has reasonable physical significance and reliability, and the precision of a tracking algorithm is improved; further, when cluster tracking is executed, a process of iteratively updating the cluster label is introduced into the tracking algorithm, so that the problem of cluster label distribution is solved. When multi-cluster fusion occurs, the new cluster label continues to use the old cluster label closest to the new cluster label in the candidate cluster, and other clusters continue to search the latest new cluster in the next iteration process until a new cluster is found or the old cluster is judged to die, so that the source problem of the surviving cluster is solved, the phenomenon of cluster fusion can be more effectively dealt with, and the robustness of the tracking algorithm is improved.

Fig. 5 is a flowchart illustrating a method for performing cluster tracking according to an embodiment of the present invention. Referring to fig. 5, the process includes:

step 501, acquiring MCD values of clusters at adjacent moments;

in this step, the MCD values of the neighboring time clusters can be obtained in step 204. For example, for a channel cluster containing M old clusters and N new clusters, each old cluster contains N MCD values and each new cluster contains M MCD values.

Preferably, the resulting MCD values are combined into an M × N MCD matrix D, where M and N represent the number of old and new clusters, respectively. Note D_i，jThe MCD values for the ith old cluster and the jth new cluster; d_i，·(i ═ 1, 2, …, M) row i of D, the MCD vector for old cluster i and all new clusters; d_·，j(j ═ 1, 2, …, N) is the jth column of D, which is the MCD vector for the new cluster j and all old clusters.

Step 502, obtaining an old cluster MCD threshold value;

in this step, the MCD threshold of each old cluster may be calculated and obtained from step 205.

Step 503, judging whether the MCD value of the new cluster at the adjacent time is larger than the MCD threshold value of the corresponding old cluster, if so, executing step 504, otherwise, executing step 505;

in this step, D is determined by traversing j to 1, 2, …, N_·，jIf each element value in the column is greater than the corresponding old cluster MCD threshold value, compared to the old cluster MCD threshold value, step 504 is performed.

For example, as described above, let the MCD threshold of the old cluster including M old clusters be ε₁，ε₂，...，ε_mIf D is_m，j＞ε_m(M1, 2.. times, M; j 1, 2.. times, N), then step 504 is performed.

Step 504, determining the new cluster as a new cluster, and allocating a new cluster number to the new cluster;

in this step, new clusters other than the new cluster are candidate clusters.

Step 505, initializing all old clusters belonging to a virtual aggregation cluster, setting an initial value q of an aggregation cluster number variable to 1, and setting the number C of aggregation cluster elements_q＝M；

In this step, if the number of aggregation clusters is Q, there are Q aggregation clusters in total of 1, 2. C_qThe number of old clusters representing the Q (Q ═ 1, 2.. Q) th candidate cluster set.

Step 506, traverse the old cluster of the aggregated cluster set, calculate the old cluster

Corresponding candidate cluster number L_kIf L is_k0, old cluster k is dead cluster; if L is_kIf not, marking a new cluster with the minimum MCD value of the old cluster k in the aggregation cluster set in the candidate clusters, and reallocating the aggregation cluster set;

step 507, detecting whether an aggregation cluster set with the number of old clusters larger than 1 exists, if not, marking the candidate cluster as a survival cluster, and the rest clusters as new clusters, allocating new cluster numbers, and ending the cluster tracking process; if yes, go to step 508;

step 508, traversing all candidate clusters corresponding to the aggregation cluster with the old cluster number greater than 1, marking the candidate cluster element with the minimum MCD value as a surviving cluster, deleting the candidate cluster element and the aggregation cluster element, updating the numbers of the aggregation cluster and the aggregation cluster element, and returning to step 506.

Fig. 6 is a flowchart illustrating a method for performing cluster tracking according to an embodiment of the present invention. Referring to fig. 6, the process includes:

step 601, calculating MCD values of N new clusters and M old clusters;

step 602, obtaining an MCD threshold value of each old cluster;

step 603, setting the initial value i to 1, and recording the MCD values of the new cluster i and all the old clusters respectively;

step 604, recording the number n of the MCD values of the new cluster i and the old cluster larger than the MCD threshold value of the corresponding old cluster;

step 605, determining whether N is equal to N, if yes, executing step 606, otherwise, executing step 607;

step 606, marking the new cluster i as a new cluster, allocating a new cluster number, and executing step 607;

step 607, determining whether i is equal to N, if so, executing step 608, otherwise, if i is i +1, executing step 603;

step 608, setting the initial value j to 1 and C to M, and recording the number L of candidate clusters corresponding to the aggregation cluster j;

step 609, judging whether L is equal to zero, if so, executing step 611, otherwise, executing step 610;

step 610, marking a new cluster with the minimum MCD value of the old cluster j in the aggregation cluster set in the candidate clusters, reallocating the aggregation cluster set, and executing step 611;

step 611, determine whether j is equal to C, if yes, go to step 612, otherwise, j equals j +1, go to step 608;

step 612, determining whether an aggregation cluster exists, if so, executing step 613, otherwise, executing step 615;

step 613, marking the candidate cluster corresponding to the latest aggregation cluster as a survival cluster, and deleting the survival cluster from the candidate cluster;

step 614, deleting the latest aggregation cluster, updating the aggregation cluster number C, and returning to execute step 608;

and step 615, marking the candidate cluster as a survival cluster and marking the non-candidate cluster as a new cluster.

The following describes a channel cluster tracking apparatus according to an embodiment of the present invention.

Fig. 7 is a schematic structural diagram of a channel cluster tracking apparatus according to an embodiment of the present invention. Referring to fig. 7, the apparatus includes: a measurement module, a pre-processing module, an estimation module, and a tracking module, wherein,

in this embodiment, the measurement module obtains multipath channel impulse response data and geographic information of both the receiving end and the transmitting end according to a channel sampling rate matched with the receiving end mobile rate. The method can be applied to uplink or downlink MIMO systems.

In practical application, a precise instant positioning system can be configured in the measurement module to record the geographical position information of the transmitting end and the receiving end. In addition, the measurement module can calibrate the channel cluster tracking device.

in this embodiment, the estimation module determines the length of a propagation path by using the time delay of a cluster, determines the scatterer direction by using the angle information of the transceiving end, further jointly constrains the scatterer distribution region, then further reduces the scatterer distribution region according to the radius of the scatterer-free region in the actual measurement environment, and finally jointly determines the MCD threshold of each cluster according to the position of the scatterer and the displacement of the mobile station, where different clusters correspond to different MCD thresholds.

In the above example, the estimation module receives the outputs of the clustering submodule and the geographic information submodule, and estimates the MCD threshold of each cluster at each time.

A tracking sub-module comprising a calculation sub-module and a comparison and update sub-module, wherein,

The comparison and update sub-module determines the cluster state based on the respective MCD threshold for each cluster, which is determined jointly by the position of the scatterer and the displacement of the mobile station. Judging the state of the cluster in an iterative mode, and checking whether a plurality of old clusters are converged into a new cluster (cluster aggregation) after obtaining a judgment result once. If so, the new cluster continues to use the old cluster number closest to the new cluster, and iteration is repeated until no aggregate cluster exists.

The status of the clusters includes the production, death and survival status of the clusters.

In this embodiment, the comparison and update submodule uses the numbers of the old clusters for the surviving clusters, and the newly generated cluster numbers are arbitrarily specified from the numbers that have not been assigned so as not to affect the state of the clusters.

The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for tracking a channel cluster, the method comprising:

2. The method of claim 1, wherein the step of acquiring multipath channel impulse response data comprises:

3. The method of claim 2, wherein the relative geographic information comprises: linear distance, direction of motion and line of sight direction included angle.

4. The method according to any one of claims 1 to 3, wherein the space-time multidimensional parameter information of the multipath is extracted from the acquired multipath channel impulse response data through a space-alternating generalized expectation-maximization (SAGE) algorithm.

5. The method according to claim 4, wherein the multi-path space-time multidimensional parameter information specifically includes: multipath power, multipath delay, horizontal departure angle, and horizontal arrival angle.

6. The method of claim 5, wherein the step of clustering the multipaths using a predetermined clustering algorithm and obtaining the space-time multidimensional parameters of the cluster center comprises:

7. The method of claim 6, wherein the space-time multidimensional parameters for the cluster center comprise cluster center power, cluster center latency, cluster center horizontal departure angle, and cluster center horizontal arrival angle.

8. A method according to any one of claims 1 to 3, wherein the calculation formula for calculating the maximum neighbour distance MCD value of each cluster at adjacent time instants is:

9. The method of claim 8,

in the formula,

denotes the horizontal departure angle, phi denotes the horizontal arrival angle, tau is the time delay, and zeta is the time delay adjustment factor.

10. The method of claim 1, wherein the step of estimating a variation range of the MCD values of each cluster according to the obtained relative geographic information, and obtaining the MCD threshold value of each corresponding cluster comprises:

11. The method of claim 10, wherein performing cluster tracking based on the obtained cluster MCD thresholds and the calculated cluster MCD values comprises:

12. An apparatus for tracking a cluster of channels, the apparatus comprising: a measurement module, a pre-processing module, an estimation module, and a tracking module, wherein,

13. The apparatus of claim 12, wherein the pre-processing module comprises a parameter extraction sub-module, a clustering sub-module, and a geographic information sub-module, wherein,

14. The apparatus of claim 12 or 13, wherein the tracking sub-module comprises a computation sub-module and a comparison and update sub-module, wherein,

15. The apparatus of claim 14, wherein the status of the clusters includes generation, death and survival of clusters, the compare and update submodule setting numbers of surviving clusters along worn clusters, newly generated cluster numbers arbitrarily designated from numbers not yet assigned.