CN103117823A - Short wave channel model building method - Google Patents

Abstract

The invention discloses a modeling method of a shortwave channel model, comprising: extracting M samples of a certain link from a database storing shortwave channel parameter samples, and obtaining multipath broadening parameters and Doppler parameters in each sample Stretch parameters; use the multipath spread parameters and Doppler spread parameters of each sample as column vectors to construct a channel parameter matrix, and normalize each vector in the matrix; define each column of the channel parameter matrix as an array Points, use the two-dimensional clustering combination algorithm to calculate the cluster centers in each array point; detect whether the number of array points covered within the neighborhood radius of each cluster center meets the clustering requirements, and if so, the cluster center Denormalization processing is performed, and the multipath broadening and Doppler broadening parameters corresponding to the cluster centers after denormalization processing are used as the shortwave channel model. The invention solves the problem that the model error obtained by the existing channel modeling technology is relatively large, and the characteristics of the link channel cannot be accurately described.

Description

A Modeling Method for HF Channel Modeling

technical field

The invention relates to the fields of shortwave channel modeling and mathematical statistics, in particular to a shortwave channel model modeling method.

Background technique

The shortwave channel is a typical time-varying parameter channel. The ionosphere changes randomly in the time domain, space domain and frequency domain, which makes the shortwave channel parameters change randomly, so that it is impossible to use an accurate channel model to describe the specific time domain, Channel characteristics in frequency and space domains. At present, the common method in the world is to use the channel model formed by ITU-RF.1487 according to the typical channel parameters of high, middle and low latitudes. However, due to the time-varying characteristics of short-wave channel parameters, the error between the channel model suggested by ITU and the actual channel model is too large, and the real-time channel parameters obtained by channel measurement have great randomness, which cannot fully describe the channel characteristics of the link. .

Contents of the invention

The present invention provides a short-wave channel model modeling method, which is used to solve the problems in the prior art that are affected by the time-varying characteristics of short-wave channel parameters, the model error obtained by the existing channel modeling technology is relatively large, and the link channel characteristics cannot be accurately described. question.

In order to solve the above problems, the technical scheme adopted in the present invention is as follows:

The invention provides a shortwave channel model modeling method, comprising:

Step 1, extracting M samples of a certain link in the database storing HF channel parameter samples, and obtaining the multipath broadening parameters and Doppler broadening parameters in each sample;

Step 2, using the multipath broadening parameters and Doppler broadening parameters of each sample as column vectors, constructing a channel parameter matrix with 2 rows and M columns, and normalizing each vector in the matrix;

Step 3, define each column of the channel parameter matrix as an array point, and use the two-dimensional clustering combination algorithm to calculate the cluster center in each array point;

Step 4, detect whether the number of array points covered within the neighborhood radius of each cluster center meets the clustering requirements, and if so, denormalize the cluster centers to obtain The multipath broadening parameters and Doppler broadening parameters corresponding to the cluster centers are used as the shortwave channel model.

Optionally, in step 1 of the method of the present invention, the shortwave channel parameter samples in the database are parameter samples obtained through shortwave channel measurement experiments.

Optionally, in step 1 of the method of the present invention, in the database, M samples that meet the conditions of a certain link are extracted according to preset time conditions, signal-to-noise ratio conditions, and sunspot number conditions.

Optionally, in step 2 of the method of the present invention, normalization processing is performed on each vector in the matrix, specifically including:

以及将各多径展宽参数进行相互比较，获取满足条件2时τ_y的最小值

Compare each Doppler broadening parameter with each other to obtain the minimum value of f _x when condition 1 is satisfied

And compare the multipath broadening parameters with each other to obtain the minimum value of τ _y when condition 2 is satisfied

和

为多普勒展宽参数和多径展宽参数的归一化基数，对矩阵中的相应参数向量进行归一化处理；respectively as described

and

is the normalized base of the Doppler broadening parameter and the multipath broadening parameter, and normalizes the corresponding parameter vector in the matrix;

大于f_x的参数数量满足预先设定的阈值；所述条件2为：多径展宽参数

大于τ_y的参数数量满足设定的阈值。Wherein, the condition 1 is: Doppler broadening parameter

The number of parameters greater than f _x satisfies a preset threshold; the second condition is: multipath broadening parameters

The number of parameters larger than τ _y satisfies a set threshold.

Optionally, in step 3 of the method of the present invention, a two-dimensional clustering combination algorithm is used to calculate the cluster centers in each array point, specifically including:

Step 31, calculating the density index of each array point, obtaining the array point corresponding to the highest density index in each density index, and determining it as the first cluster center;

Step 32, using the density index of the kth cluster center to correct the density index of each array point, and obtain the array point corresponding to the highest value in the corrected density index;

Step 33, according to the set clustering judgment threshold, judge the clustering center of the obtained array point, when it is judged that the array point is the clustering center, get the k+1th clustering center, and set k=k +1, back to step 32.

Wherein, the step 33 specifically includes:

大于

是否成立，若是，判定该数组点为一个聚类中心；否则，执行步骤332；Step 331, determine the array point corresponding

more than the

Whether it is established, if so, determine that the array point is a cluster center; otherwise, execute step 332;

小于

是否成立，若是，判定该数组点不是聚类中心，终止聚类过程；否则，执行步骤333；Step 332, determine the array point corresponding

less than

Whether it is established, if so, determine that the array point is not a cluster center, and terminate the clustering process; otherwise, execute step 333;

大于小于

时，判断

是否成立，若是，判定该数组点为一个聚类中心；否则，判定该数组点不是聚类中心，将该数组点对应的密度指标设为零，并选择余下数组点中具有最高密度指标对应的数组点为待确认点，返回步骤331；Step 333, in

more than the less than

time, judge

Whether it is true, if so, determine that the array point is a cluster center; otherwise, determine that the array point is not a cluster center, set the density index corresponding to the array point to zero, and select the one with the highest density index corresponding to the remaining array points Array points are points to be confirmed, return to step 331;

ε分别为预先设定的判决门限上限和下限，

为第一个聚类中心的密度指标；d_min为当前待确认数组点距离第一个聚类中心的距离，r_a为设定的数组点的邻域半径。in,

ε are the preset upper and lower limits of the decision threshold, respectively,

is the density index of the first cluster center; d _min is the distance between the current array point to be confirmed and the first cluster center, and r _a is the neighborhood radius of the set array point.

实现对每个数组点的密度指标进行修正；式中，

为第k个聚类中心的密度指标，

和

分别为经过归一化处理后的第i个样本的多普勒展宽参数和多径展宽参数，

和

分别为经过归一化处理后的第k个聚类中心的多普勒展宽参数和多径展宽参数，r_b为正数，且满足r_b大于数组点的邻域半径。Further, in the step 32, using the formula

Realize the correction of the density index of each array point; where,

is the density index of the kth cluster center,

and

are the Doppler broadening parameters and multipath broadening parameters of the i-th sample after normalization, respectively,

and

are the Doppler broadening parameters and multipath broadening parameters of the k-th cluster center after normalization, respectively, r _b is a positive number, and r _b is greater than the neighborhood radius of the array point.

Optionally, step 4 of the method of the present invention specifically includes:

是否大于等于设定的聚类阈值，以及

是否大于等于设定的平均聚类阈值，当二者均大于等于对应的阈值时，将所述聚类中心进行去归一化处理，以去归一化处理后的聚类中心对应的多径展宽参数和多普勒展宽参数作为短波信道模型；其中，K为聚类中心的总个数，

表示任意给定的。Get the number of array points N _ck falling within the neighborhood radius of each cluster center, and detect

Is it greater than or equal to the set clustering threshold, and

Whether it is greater than or equal to the set average clustering threshold, when both are greater than or equal to the corresponding threshold, the clustering center is denormalized, and the multipath corresponding to the denormalized clustering center The broadening parameter and the Doppler broadening parameter are used as the shortwave channel model; where K is the total number of cluster centers,

means any given.

Among them, the array points falling within the neighborhood radius of the cluster center are the array points satisfying the following conditions: ${(f_{{\mathrm{\σ}}_i}^{\′}-f_{{\mathrm{\σ}}_k}^{\′})}^2+{({\mathrm{\τ}}_{{\mathrm{\σ}}_i}^{\′}-{\mathrm{\τ}}_{{\mathrm{\σ}}_k}^{\′})}^2\≤{\left(r_a\right)}^2;$

和

分别为归一化处理后的第i个数组点的多普勒展宽参数和多径展宽参数，

和

分别为归一化处理后的第k个聚类中心的多普勒展宽参数和多径展宽参数，r_a为设定的数组点的邻域半径。In the formula,

and

are the Doppler broadening parameters and multipath broadening parameters of the i-th array point after normalization, respectively,

and

are the Doppler broadening parameter and the multipath broadening parameter of the k-th cluster center after normalization, respectively, and r _a is the neighborhood radius of the set array point.

Optionally, step 4 of the method of the present invention further includes:

和ε。When it is detected that the number of array points covered within the radius of each cluster center does not meet the clustering requirements, the modeling process is ended, or, the variable parameters used by the two-dimensional clustering combination algorithm are adjusted, and step 3 is re-executed; wherein, the The variable parameters include: the set neighborhood radius r _a of the array points, the set neighborhood r _b whose density index function is significantly reduced, and the set upper and lower limits of the cluster decision threshold

and ε .

The beneficial effects of the present invention are as follows:

The present invention establishes a method for modeling short-wave channel models by effectively classifying discrete channel parameters. Using the method provided by the present invention, short-wave channel models under specific time domains, air domains, and specific sunspot numbers can be established. Compared with the channel model provided by ITU-RF.1487, it is closer to the actual channel characteristics, and the channel model established according to this method can provide effective support for short-wave communication and short-wave frequency scoring.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description These are only some embodiments of the present invention, and those skilled in the art can also obtain other drawings based on these drawings without any creative effort.

Fig. 1 is the flow chart of a kind of shortwave channel model modeling method that the embodiment of the present invention provides;

FIG. 2 is a flow chart of an algorithm for performing two-dimensional clustering and combination of channel parameters in an embodiment of the present invention.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

In order to solve the problem that the model error obtained by the existing channel modeling technology in the prior art is relatively large, and the channel characteristics of the link cannot be accurately described, the embodiment of the present invention provides a shortwave channel model modeling method. From the measurement data on the road, the specified channel parameters are extracted in time division and segment, and the two-dimensional clustering combination algorithm is used to effectively classify the extracted discrete channel parameters, so as to obtain the channel on the specific time domain and air domain link. Model. The technical solution of the present invention will be described in detail below through several specific examples.

Embodiment one

As shown in Figure 1, an embodiment of the present invention provides a shortwave channel model modeling method, including:

Step S101, extracting M samples of a link from a database storing shortwave channel parameter samples, and obtaining multipath broadening parameters and Doppler broadening parameters in each sample;

In this step, the shortwave channel parameter samples in the database are parameter samples obtained through shortwave channel measurement experiments.

Further, in this step, in the database, according to preset time conditions, signal-to-noise ratio conditions, and sunspot number conditions, M samples that meet the conditions of a certain link are extracted. Among them, each condition can be flexibly set according to specific modeling requirements.

Step S102, using the multipath spreading parameters and Doppler spreading parameters of each sample as column vectors, constructing a channel parameter matrix with 2 rows and M columns, and normalizing each vector in the matrix;

In this step, the way of normalizing each vector in the matrix includes:

method one:

(1) Compare the Doppler broadening parameters with each other to obtain the minimum value of f _x when condition 1 is satisfied And compare the multipath broadening parameters with each other to obtain the minimum value of τ _y when condition 2 is satisfied

和

为多普勒展宽参数和多径展宽参数的归一化基数，对矩阵中的相应参数向量进行归一化处理；(2) Respectively as described

and

is the normalized base of the Doppler broadening parameter and the multipath broadening parameter, and normalizes the corresponding parameter vector in the matrix;

大于τ_y的参数数量满足设定的阈值。In the above method, condition 1 is: Doppler broadening parameter The number of parameters greater than f _x satisfies a preset threshold; condition 2 is: multipath broadening parameter

The number of parameters larger than τ _y satisfies a set threshold.

Method 2:

i＝1,2,...n，计算满足条件

时，多普勒展宽参数的个数L_f，并求出L_f满足大于等于设定阈值的各f_x中的最小值

以所述

为多普勒展宽参数的归一化基数，对矩阵中的多普勒展宽参数进行归一化处理；Set the judgment threshold of Doppler broadening parameters according to a specific step size

i=1,2,...n, the calculation satisfies the condition

, the number of Doppler broadening parameters L _f , and find the minimum value of each f _x that L _f satisfies greater than or equal to the set threshold

as stated

is the normalization base of the Doppler broadening parameter, and normalizes the Doppler broadening parameter in the matrix;

时，多普勒展宽参数的个数τ_l，并求出τ_l满足大于设定阈值的各τ_y中的最小值

以所述为多径展宽参数的归一化基数，对矩阵中的多径展宽参数进行归一化处理。According to a specific step size, set the multipath broadening parameter judgment threshold i=1,2,...n, the calculation satisfies the condition

, the number of Doppler broadening parameters τ _l , and the minimum value of each τ _y that τ _l satisfies greater than the set threshold

as stated is the normalized base of the multipath broadening parameters, and normalizes the multipath widening parameters in the matrix.

Regarding the setting of the above threshold, it is suggested in the present invention that the threshold is greater than or equal to M·0.95. Of course, those skilled in the art can also flexibly set it according to specific modeling requirements.

Step S103, defining each column of the channel parameter matrix as an array point, and using a two-dimensional clustering combination algorithm to calculate the cluster centers in each array point;

Step S104, detecting whether the number of array points covered within the neighborhood radius of each cluster center satisfies the clustering requirements, and if so, denormalizes the cluster centers to obtain The multipath broadening parameters and Doppler broadening parameters corresponding to the cluster centers are used as the shortwave channel model.

是否大于等于设定的平均聚类阈值，当二者均大于等于对应的阈值时，将所述聚类中心进行去归一化处理，以去归一化处理后的聚类中心对应的多径展宽参数和多普勒展宽参数作为短波信道模型；其中，K为聚类中心的总个数，

表示任意给定的。This step is specifically as follows: obtain the number N _{ck of} the array points falling within the neighborhood radius of each cluster center, and detect Is it greater than or equal to the set clustering threshold, and

Whether it is greater than or equal to the set average clustering threshold, when both are greater than or equal to the corresponding threshold, the clustering center is denormalized, and the multipath corresponding to the denormalized clustering center The broadening parameter and the Doppler broadening parameter are used as the shortwave channel model; where K is the total number of cluster centers,

means any given.

Among them, the settings of the clustering threshold and the average clustering threshold indicate whether the calculated cluster centers are typical, and the specific values of the two are not unique. This embodiment suggests that the average clustering threshold should be greater than or equal to 60%, and the clustering threshold It must be greater than or equal to 12%.

和ε。Further, in the step S104, when it is detected that the number of array points covered within the radius of each cluster center does not meet the clustering requirements, the modeling process is ended, or the variable parameters used by the two-dimensional clustering combination algorithm are adjusted, Re-execute step S103; wherein, the variable parameters include but are not limited to: the set neighborhood radius r _a of the array points, the set neighborhood r _b where the density index function is significantly reduced, and the set clustering decision threshold Upper and lower limits

and ε .

Embodiment two

The embodiment of the present invention provides a shortwave channel model modeling method, the principle of which is the same as that of the method described in Embodiment 1. It is a further detailed elaboration of the method described in Embodiment 1 in combination with specific implementation details. The method includes:

Step A: Obtain channel parameter samples;

Step B: performing two-dimensional clustering combination on channel parameters;

Step C: Corresponding the cluster center to the channel model;

For the above step A, the specific implementation process is as follows:

According to modeling requirements, two types of uncorrelated channel parameters that can fully reflect channel characteristics are obtained: multipath broadening and Doppler broadening. Among them, the source of channel parameters: after completing the channel measurement experiment on a specific link, the channel parameters are analyzed and calculated from the measured data. Each measured data sample corresponds to a set of channel parameters, as shown in Table 1. The channel parameter table .

Table 1

Step A1: According to the characteristics of the shortwave channel, set the channel parameter extraction conditions;

(1) Set the measurement time condition: T _min <t<T _max , in general, the time interval is preferably 1 month <T _max -T _min <3 months

(2) Set the signal-to-noise ratio condition: Snr>xdB, under normal circumstances, select 5dB for x.

(3) Set the sunspot number condition: N _min <n<N _max , in general, the sunspot number interval N _max -N _min is 25.

Step A2: extracting channel parameters;

Use the three conditions in step A1 to find the intersection of data samples that meet the conditions. Assuming that the intersection of data samples is M, put the corresponding Doppler broadening parameter f _σ and multipath broadening channel parameter τ _σ into the channel parameter matrix middle.

$\mathrm{<span\; class="notranslate">\; Channel</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Paramet</span>}<span\; class="notranslate">\; =</span>\left[\begin{array}{c}{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}\\ {\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}\end{array}\right]$

Step A3: Normalize the channel parameters, the specific steps are as follows:

τ_σ>τ_y时多普勒展宽和多径展宽参数的个数L_f和τ_l，生成参数数值与参数个数的对应表，如表2所示。Step A31: Computing to satisfy the condition

When τ _σ >τ _y , the number of Doppler broadening and multipath broadening parameters L _f and τ _l , generate a corresponding table of parameter values and parameter numbers, as shown in Table 2.

Table 2

同时求出满足τ_l≥M·0.95的τ_y最小值，设为

Step A32: Find the minimum value of f _x that satisfies L _f ≥ M·0.95 from Table 2, and set it as

At the same time, find the minimum value of τ _y that satisfies τ _l ≥ M·0.95, set

和

做归一化处理，得到:Step A33: Use the numbers in the first row and the second row in the channel parameter matrix respectively by

and

After normalization, we get:

$\mathrm{<span\; class="notranslate">\; Channel</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Paramet</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; unitary</span>}<span\; class="notranslate">\; =</span>\left[\begin{array}{c}{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}\\ {\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}\end{array}\right]$

For step B, the specific implementation process is shown in Figure 2, including:

First, calculate the density index of each column data group in the Channel-Paramet-Unitary matrix.

都作为聚类中心的侯选数组点。计算每列

的密度指标：Group each column of data in the Channel-Paramet-Unitary matrix

All of them are used as the candidate array points of the cluster centers. Calculate each column

The density index of:

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\frac{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{\mathrm{<span\; class="notranslate">\; \&sigma;j</span>}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; ]</span>$

周围有多个邻近的数组点，则具有高密度值，半径以外的数组点对该点的密度指标贡献非常小。r_a的取值与希望得到的聚类中心个数有关，当r_a越大，得到的聚类中心个数越少；相反的，当r_a越小，得到的聚类中心个数越多。所以，r_a可以根据需求进行灵活设置。Among them, r _a is a positive number that defines the neighborhood radius of the point, if the array

There are many adjacent array points around, it has a high density value, and the array points outside the radius contribute very little to the density index of the point. The value of r _a is related to the number of cluster centers you want to obtain. When r _a is larger, the number of cluster centers obtained is less; on the contrary, when r _a is smaller, the number of cluster centers obtained is more . Therefore, r _a can be flexibly set according to requirements.

Then, select the array point with the highest density index to judge the cluster center, and when the array point is the cluster center, correct the density index of each array point with the density index of the array point, as follows:

作为第一个聚类中心，

为其密度指标。设当前选出了第k个聚类中心

其密度指标为

则用

修正每个数组点的密度指标：After calculating the density index of each array point, select the array point with the highest density index

As the first cluster center,

is its density index. Suppose the kth cluster center is currently selected

Its density index is

then use

Fix the density index for each array point:

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; D.</span>}}_{{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\frac{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; ]</span>$

Where r _b is a positive number, which defines a neighborhood where the density index function is significantly reduced, usually r _b >r _a to avoid cluster centers that are very close to each other, and r _b =1.5r _a in this embodiment . After the correction, the density index of the array points that are obviously close to the previous cluster center will be significantly reduced, making these points less likely to be selected as the next cluster center.

然后利用如下判决方式，判决该数组点是否符合聚类中心要求，并判决是否终止聚类；设判决门限的上限为

下限为ε＝0.15。Among the corrected density indexes, select the array point of the highest density index and calculate its density index

Then use the following judgment method to judge whether the array point meets the requirements of the clustering center, and judge whether to terminate the clustering; set the upper limit of the judgment threshold as

The lower limit is ε = 0.15.

认为

为一个聚类中心，以

对各组的密度指标进行修正后，继续下一判决过程。(a) when

think

as a cluster center, with

After correcting the density index of each group, continue to the next judgment process.

认为

不是聚类中心，终止聚类过程。(b) when

think

If it is not a cluster center, the clustering process is terminated.

时，判断式

是否成立，若成立，则认为

为一个聚类中心，以

对各组的密度指标进行修正后，继续下一判决过程；若不成立，认为

不是聚类中心，并将该数组点的密度指标设为0，选择余下数组点中具有最高密度指标的点为待确认的点，重新进行判决。(c) when

time, judgment

Whether it is established, and if it is established, it is considered that

as a cluster center, with

After correcting the density index of each group, continue to the next judgment process; if not established, consider

It is not the cluster center, and the density index of the array point is set to 0, and the point with the highest density index among the remaining array points is selected as the point to be confirmed, and the judgment is made again.

For step C, the specific implementation process is as follows:

落在每个聚类中心半径范围内的比例。落在第k个聚类中心半径范围内的数组点个数为：满足 ${(f_{{\mathrm{\&sigma;}}_i}/f_{x_L}-f_{{\mathrm{\&sigma;}}_{\mathrm{ck}}}/f_{x_L})}^2+{({\mathrm{\&tau;}}_{{\mathrm{\&sigma;}}_i}/{\mathrm{\&tau;}}_{y_L}-{\mathrm{\&tau;}}_{\mathrm{\&sigma;ck}}/{\mathrm{\&tau;}}_{y_L})}^2\&le;{\left(r_a\right)}^2$ 的数组点个数，记为N_ck。Step C1: Calculate all array points in the Channel-Paramet-Unitary matrix

The proportion that falls within the radius of each cluster center. The number of array points falling within the radius of the kth cluster center is: ${(f_{{\mathrm{\&sigma;}}_i}/f_{x_L}-f_{{\mathrm{\&sigma;}}_{\mathrm{ck}}}/f_{x_L})}^2+{({\mathrm{\&tau;}}_{{\mathrm{\&sigma;}}_i}/{\mathrm{\&tau;}}_{{\mathrm{the\; y}}_L}-{\mathrm{\&tau;}}_{\mathrm{\&sigma;ck}}/{\mathrm{\&tau;}}_{{\mathrm{the\; y}}_L})}^2\&le;{\left(r_a\right)}^2$ The number of array points in , denoted as N _ck .

并且 $\frac{\&ForAll;N_{\mathrm{ck}}}{M}\&times;100\%\&GreaterEqual;15\%$ 时，认为聚类中心

能表示信道模型。Step C2: Check the N _ck value within the radius of each cluster center, when

and $\frac{\&ForAll;N_{\mathrm{ck}}}{m}\&times;100\%\&Greater\; Equal;15\%$ When , it is considered that the cluster center

Can represent channel models.

Step C3: Denormalize the cluster centers that meet the requirements to obtain the channel model. The channel model is expressed as $(f_{{\mathrm{\&sigma;}}_{c_1}},{\mathrm{\&tau;}}_{{\mathrm{\&sigma;}}_{c_1}})...(f_{{\mathrm{\&sigma;}}_{c_k}},{\mathrm{\&tau;}}_{{\mathrm{\&sigma;}}_{c_k}}).$

In order to further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, a specific application example is given below to further explain the technical solution proposed by the present invention.

Table 3 shows the channel parameter table. The channel parameter table is based on the channel parameters extracted from the measurement data transmitted by Qingdao and received by Beijing from January 12, 2012 to January 17, 2012.

table 3

The embodiment of the present invention is based on the shortwave channel model modeling method of above-mentioned measured data, comprises following concrete steps:

Step A: Obtain channel parameter samples. Obtain samples of Doppler broadening and multipath broadening parameters for channel modeling from Table 3, the specific steps are as follows:

A1: According to the characteristics of the short-wave channel, set the channel parameter extraction conditions; because the time span of the measurement data is small, it is considered that all the channel parameters extracted from the measurement data can be used. Set the signal-to-noise ratio condition as Snr>5dB. Since this measurement experiment is in winter, the time span is short, and the number of sunspots does not change much (the difference between the maximum value and the minimum value is not greater than 25), the sunspot number interval is not divided.

A2: Extract channel parameters; from the conditions of using step A1, the data sample set can be obtained, and the data sample set is 1135, and the corresponding Doppler broadening parameter f _σ and multipath widening channel parameter τ _σ are respectively put into the channel in the parameter matrix.

$\mathrm{<span\; class="notranslate">\; Channel</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Paramet</span>}<span\; class="notranslate">\; =</span>\left[\begin{array}{c}{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}\\ {\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}\end{array}\right]<span\; class="notranslate">\; =</span>\left[\begin{array}{cccc}\mathrm{<span\; class="notranslate">\; 2.52</span>}& \mathrm{<span\; class="notranslate">\; 2.41</span>}& \mathrm{<span\; class="notranslate">\; 3.23</span>}& <span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span>\\ \mathrm{<span\; class="notranslate">\; 0.25</span>}& \mathrm{<span\; class="notranslate">\; 0.36</span>}& \mathrm{<span\; class="notranslate">\; 0.21</span>}& <span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span>\end{array}\right]$

A3: Normalize the channel parameters; according to the characteristics of Doppler broadening and multipath broadening parameters, find out the values used for normalization, and normalize the two parameters respectively. The specific steps are as follows:

τ_σ>τ_y时多普勒展宽和多径展宽参数的个数L_f和τ_l。生成参数数值与参数个数的对应表，如表4所示；A31: Calculation satisfies the condition

When τ _σ >τ _y , the number of Doppler broadening and multipath broadening parameters L _f and τ _l . Generate a corresponding table of parameter values and parameter numbers, as shown in Table 4;

Table 4

A32: From Table 4, obtain the minimum value of f _x satisfying L _f ≥ M·0.98, which is 0.83 Hz, and obtain the minimum value of τ _y satisfying τ _l ≥ M·0.98, and set it to 79 ms.

A33: Normalize the first row and the second row of data in the channel parameter matrix with 0.83 and 79 respectively, and get:

$\mathrm{<span\; class="notranslate">\; Channel</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Paramet</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; unitary</span>}<span\; class="notranslate">\; =</span>\left[\begin{array}{c}{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}\\ {\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}\end{array}\right]<span\; class="notranslate">\; =</span>\left[\begin{array}{ccccc}\mathrm{<span\; class="notranslate">\; 0.3190</span>}& & \mathrm{<span\; class="notranslate">\; 0.3051</span>}& \mathrm{<span\; class="notranslate">\; 0.4089</span>}& <span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span>\\ \mathrm{<span\; class="notranslate">\; 0.3012</span>}& & \mathrm{<span\; class="notranslate">\; 0.4337</span>}& \mathrm{<span\; class="notranslate">\; 0.2530</span>}& <span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span>\end{array}\right]$

Step B, perform two-dimensional clustering combination on the matrix Channel-Paramet-Unitary according to the following process:

First, calculate the density index of each column data group in the Channel-Paramet-Unitary matrix.

都作为聚类中心的侯选数组点。计算每列

的密度指标：Group each column of data in the Channel-Paramet-Unitary matrix

All of them are used as the candidate array points of the cluster centers. Calculate each column

The density index of:

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\frac{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{\mathrm{<span\; class="notranslate">\; \&sigma;j</span>}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; ]</span>$

where r _a =0.17.

Then, pick the array point with the highest density index and update the density index for each data point.

作为第一个聚类中心，为其密度指标。设当前选出了第k个聚类中心

其密度指标为

则用

修正每个数组点的密度指标After calculating the density index of each array point, select the array point with the highest density index

As the first cluster center, is its density index. Suppose the kth cluster center is currently selected

Its density index is

then use

Fix density index for each array point

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; D.</span>}}_{{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\frac{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; ]</span>$

where r _b =1.5r _a .

然后利用如下判决方式，判决该数组点是否符合聚类中心要求，并判决是否终止聚类；设判决门限为

ε=0.15；Among the corrected density indexes, select the array point of the highest density index and calculate its density index

Then use the following judgment method to judge whether the array point meets the requirements of the clustering center, and judge whether to terminate the clustering; set the judgment threshold as

ε =0.15;

为一个聚类中心，并继续密度指标的修正过程。(a) when think

as a cluster center, and continue the correction process of the density index.

不是聚类中心，终止聚类过程。(b) when think

If it is not a cluster center, the clustering process is terminated.

时，判断式

是否成立，不成立时，认为

不是聚类中心，并将该数组点的密度指标设为0，选择余下数组点中具有最高密度指标的点为待确认的点，重新进行判决。(c) when

time, judgment

Whether it is established, if it is not established, it is considered

It is not the cluster center, and the density index of the array point is set to 0, and the point with the highest density index among the remaining array points is selected as the point to be confirmed, and the judgment is made again.

Through the above clustering, three cluster centers are obtained: (0.2530, 0.0759), (0.2651, 0.6329), (0.2410, 0.3418).

Step C, according to the following steps to correspond the cluster center with the channel model.

Step C1: Calculate the proportion of all array points in the Channel-Paramet-Unitary matrix falling within the radius of each cluster center. According to the constraint formula ${(f_{{\mathrm{\&sigma;}}_i}/f_{x_L}-f_{{\mathrm{\&sigma;}}_{\mathrm{ck}}}/f_{x_L})}^2+{({\mathrm{\&tau;}}_{{\mathrm{\&sigma;}}_i}/{\mathrm{\&tau;}}_{{\mathrm{the\; y}}_L}-{\mathrm{\&tau;}}_{\mathrm{\&sigma;ck}}/{\mathrm{\&tau;}}_{{\mathrm{the\; y}}_L})}^2\&le;{\left(r_a\right)}^2$ Calculate the number of array points falling in the three cluster centers to be 454, 219 and 170 respectively.

Step C2: After inspection, the ratio of the number of array points within the radius of all cluster centers to the total array points is $\frac{\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}{N}_{\mathrm{ck}}}{m}\&times;100\%=\frac{839}{1135}\&times;100\%=73.9\%\&Greater\; Equal;70\%,$ and $\frac{\&ForAll;N_{\mathrm{ck}}}{m}\&times;100\%\&Greater\; Equal;15\%,$ It is considered that the cluster centers (0.2530, 0.0759), (0.2651, 0.6329) and (0.2410, 0.3418) can represent the channel model.

Step C3: Denormalize the cluster centers that meet the requirements to obtain the channel model. The channel model of this link under the setting conditions of step A1 can be obtained as shown in Table 5.

table 5

channel parameters Doppler broadening (Hz) Multipath spread (ms) Channel Model 1 0.21 0.6 Channel Model 2 0.22 5.0 Channel Model 3 0.2 2.7

In summary, the embodiment of the present invention establishes a method for modeling shortwave channel models by effectively classifying discrete channel parameters. Using the method provided by the embodiment of the present invention, specific time domain, air domain and specific sun domain can be established. The HF channel model under the sunspot number is closer to the actual channel characteristics than the channel model provided by ITU-RF.1487, and the channel model established according to this method can provide effective support for HF communication and HF frequency scoring.

Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims (10)

1. A short wave channel model modeling method is characterized by comprising the following steps:

step 1, extracting M samples of a certain link from a database in which short wave channel parameter samples are stored, and acquiring a multipath broadening parameter and a Doppler broadening parameter in each sample;

step 2, taking the multipath broadening parameters and Doppler broadening parameters of each sample as column vectors, constructing a channel parameter matrix of 2 rows and M columns, and carrying out normalization processing on each vector in the matrix;

step 3, defining each column of the channel parameter matrix as an array point, and calculating a clustering center in each array point by using a two-dimensional clustering combination algorithm;

and 4, detecting whether the number of the array points covered in the neighborhood radius range of each clustering center meets the clustering requirement, if so, performing de-normalization processing on the clustering centers, and taking the multipath broadening parameters and the Doppler broadening parameters corresponding to the de-normalized clustering centers as short wave channel models.

2. The method of claim 1, wherein in the step 1, the short wave channel parameter samples in the database are parameter samples obtained through a short wave channel measurement experiment.

3. The method according to claim 1 or 2, wherein in the step 1, M samples meeting the conditions of a certain link are extracted from the database according to preset time conditions, signal-to-noise ratio conditions and sun blackness conditions.

4. The method according to claim 1, wherein in the step 2, normalizing the vector quantities in the matrix specifically includes:

comparing the Doppler broadening parameters with each other to obtain f when the condition 1 is met_xMinimum value of (2)

And comparing the multipath broadening parameters with each other to obtain the tau satisfying the condition 2_yMinimum value of (2)

Are respectively provided with

Andnormalizing the corresponding parameter vector in the matrix for the normalization cardinality of the Doppler broadening parameter and the multipath broadening parameter;

wherein the condition 1 is: doppler spread parameter

Greater than f_xThe number of the parameters satisfies a preset threshold; the condition 2 is as follows: multipath broadening parameter

Greater than τ_yThe number of parameters of (2) satisfies a set threshold.

5. The method according to claim 1 or 4, wherein in the step 3, calculating the cluster center in each array point by using a two-dimensional cluster combination algorithm specifically comprises:

step 31, calculating the density index of each array point, acquiring the array point corresponding to the highest density index in each density index, and judging that the array point is a first clustering center;

step 32, correcting the density index of each array point by using the density index of the kth clustering center, and acquiring the array point corresponding to the highest value in the corrected density index;

and step 33, judging the clustering centers of the acquired array points according to the set clustering judgment threshold, obtaining the (k + 1) th clustering center when the array points are judged to be the clustering centers, and enabling k = k +1, and returning to the step 32.

6. The method according to claim 5, wherein the step 33 specifically comprises:

step 331, determining the array point correspondences

Is greater than

If yes, judging the array point as a clustering center; otherwise, go to step 332;

step 332, determining the array point correspondences

Is less than

If yes, judging that the array point is not a clustering center, and terminating the clustering process; otherwise, go to step 333;

step 333, in

Is greater thanIs less than

Time, judge

If yes, judging the array point as a clustering center; otherwise, it is determined that the array point is not the clustering center, the density index corresponding to the array point is set to zero, and the array point corresponding to the highest density index in the remaining array points is selected as the point to be confirmed, and the procedure returns to step 331;

wherein, εrespectively as the upper limit and the lower limit of a preset judgment threshold,

the density index of the first clustering center is obtained; d_minFor the array point currently to be confirmedDistance from the first cluster center, r_aIs the neighborhood radius of the set array point.

7. The method of claim 5, wherein in step 32, a formula is used

The density index of each array point is corrected; in the formula,is an index of the density of the k-th cluster center,

andrespectively are the Doppler broadening parameter and the multipath broadening parameter of the ith sample after normalization processing,

and

respectively are Doppler broadening parameter and multipath broadening parameter r of the k-th clustering center after normalization processing_bIs positive and satisfies r_bGreater than the neighborhood radius of the array point.

8. The method according to claim 1, wherein the step 4 specifically comprises:

obtaining the number N of array points falling in the neighborhood radius range of each cluster center_ckDetectingWhether or not it is greater than or equal to a set clustering threshold, an

Whether the average clustering threshold is larger than or equal to a set average clustering threshold or not, when both are larger than or equal to corresponding thresholds, the clustering center is subjected to normalization processing, and a multipath broadening parameter and a Doppler broadening parameter corresponding to the clustering center after the normalization processing are used as short wave channel models; wherein K is the total number of the clustering centers,any given is meant.

9. The method of claim 8, wherein the array points falling within the neighborhood radius of the cluster center are array points satisfying the following condition:

in the formula,

and

respectively are Doppler broadening parameters and multipath broadening parameters of the ith array point after normalization processing,

and

respectively a Doppler broadening parameter and a multipath broadening parameter r of the k-th clustering center after normalization processing_aIs the neighborhood radius of the set array point.

10. The method of claim 1, 8 or 9, wherein the step 4 further comprises:

when detecting that the number of the array points covered in the radius range of each cluster center does not meet the cluster requirement, ending the modeling process, or adjusting variable parameters used by the two-dimensional cluster combination algorithm, and executing the step 3 again; wherein the variable parameters include: neighborhood radius r of set array points_aNeighborhood r with significantly reduced set density index function_bAnd the upper and lower limits of the set clustering judgment thresholdAndε。

Images