CN114338313B

CN114338313B - Frequency offset acquisition method and device, electronic equipment and storage medium

Info

Publication number: CN114338313B
Application number: CN202011052576.4A
Authority: CN
Inventors: 鲁大顺; 金晓成; 李丹妮
Original assignee: Datang Mobile Communications Equipment Co Ltd
Current assignee: Datang Mobile Communications Equipment Co Ltd
Priority date: 2020-09-29
Filing date: 2020-09-29
Publication date: 2024-04-12
Anticipated expiration: 2040-09-29
Also published as: CN114338313A

Abstract

The embodiment of the application provides a frequency offset acquisition method, a device, electronic equipment and a storage medium, wherein the method comprises the steps of clustering data points of data to be transmitted based on a Gaussian mixture model comprising four Gaussian models, and acquiring phase offsets corresponding to the data points; acquiring frequency offset of the data to be transmitted based on the phase offset corresponding to the data point; according to the embodiment of the application, four clusters are obtained based on a Gaussian mixture model aiming at data points of data to be sent, and then frequency offset of the data to be sent is obtained; the frequency offset estimation is carried out based on the data points, the number of the pilot frequency columns is not dependent to be configured for correlation, the method is suitable for scenes under single-column pilot frequency or even pilot frequency-free configuration, and the problem that a frequency offset solving method by a two-column correlation method is limited when the single-column pilot frequency is configured in the prior art is solved; and the frequency offset estimation is carried out based on the Gaussian mixture model, so that the frequency offset estimation error is effectively reduced, the method is more suitable for actual application scenes, and the robustness is good.

Description

Frequency offset acquisition method and device, electronic equipment and storage medium

Technical Field

The present disclosure relates to the field of communications technologies, and in particular, to a method and apparatus for obtaining frequency offset, an electronic device, and a storage medium.

Background

Frequency synchronization is an important issue in communications. Frequency offset caused by frequency asynchronization can bring interference among carriers, so that the error rate is increased.

In the prior art, when two columns of pilot symbols are configured, two columns of pilots are generally used for carrying out correlation operation to obtain a frequency offset value; in order to increase the transmission rate, a case may occur in which only a single column of pilots is configured. In this case, the conventional frequency offset acquisition method similar to the two-column pilot correlation method is no longer applicable. When configuring single-column pilot symbols, frequency offset values are usually obtained based on a Kmeans clustering method, but the robustness of the method is not good enough.

Therefore, how to propose a frequency offset obtaining method which can be applied to more scenes and has strong robustness becomes a problem to be solved.

Disclosure of Invention

The embodiment of the application provides a frequency offset acquisition method, a device, electronic equipment and a storage medium, which are used for solving the defect of insufficient robustness in the prior art and realizing that frequency offset can be acquired in more scenes with high robustness.

In a first aspect, an embodiment of the present application provides a method for acquiring a frequency offset, including:

clustering data points of data to be transmitted based on a Gaussian mixture model to obtain phase deviations corresponding to the data points;

Acquiring frequency offset of the data to be transmitted based on the phase offset corresponding to the data point;

the Gaussian mixture model comprises four Gaussian models, and the four Gaussian models correspond to four clusters respectively.

According to an embodiment of the present application, the method for obtaining a frequency offset for a data point of data to be sent, clusters based on a gaussian mixture model, and obtains a phase offset corresponding to the data point, specifically includes:

and clustering data points of any symbol in any time slot of data to be transmitted based on a Gaussian mixture model to acquire phase bias of the symbol.

According to an embodiment of the present application, the method for obtaining a frequency offset of the data to be sent based on a phase offset corresponding to the data point includes:

determining the frequency offset of any symbol in any time slot of data to be transmitted based on the phase offset corresponding to the data point of the symbol and the time difference of the symbol relative to the symbol of the DMRS (Demodulation Reference Sgnal, demodulation reference signal);

and acquiring the frequency offset of the data to be transmitted based on the frequency offset of at least one symbol in the time slot.

According to an embodiment of the present application, the method for obtaining a frequency offset of a symbol in any time slot of data to be transmitted, based on a gaussian mixture model, performs clustering, and obtains a phase offset of the symbol, includes:

Clustering is carried out on the data points in the symbol based on the four Gaussian models, and four types of data points are correspondingly acquired;

and determining the phase offset of the symbol according to the position of the clustering center of any type of data points in the constellation diagram corresponding to the symbol.

According to an embodiment of the present application, the clustering is performed on the data points in the symbol based on the four gaussian models, and four types of data points are correspondingly acquired, including:

in each clustering process, based on four Gaussian models updated in the last clustering process, respectively obtaining data points included in each cluster;

for each cluster, updating parameters of a corresponding Gaussian model based on the data points included by the cluster; the parameters comprise mean parameters for describing a clustering center of the cluster;

and after the clustering is finished, acquiring the four types of data points acquired in the last clustering process.

According to the frequency offset obtaining method of one embodiment of the application, in each clustering process, after updating parameters in the corresponding gaussian model based on data points included in each cluster, the method further comprises:

And obtaining likelihood values in the current clustering process according to likelihood functions based on parameters of the four updated Gaussian models.

According to an embodiment of the present application, the method for obtaining the frequency offset includes:

the clustering times exceed a preset value or the likelihood value converges.

According to an embodiment of the present application, the method for obtaining the frequency offset determines the phase offset of the symbol according to the position of the cluster center of any type of data point in the constellation diagram corresponding to the symbol, including:

and according to the position of the clustering center of any type of data point in the constellation diagram, referring to the phase offset of at least one other symbol in the time slot to which the symbol belongs, and determining the phase offset of the symbol.

According to an embodiment of the present application, the clustering is performed on the data points of one symbol in the time slot based on a gaussian mixture model, specifically:

and clustering data points with the amplitude of one symbol smaller than a preset amplitude in the time slot based on a Gaussian mixture model.

In a second aspect, embodiments of the present application provide an electronic device including a memory, a transceiver, and a processor:

a memory for storing a computer program; a transceiver for transceiving data under control of the processor; a processor for reading the computer program in the memory and performing the following operations:

According to an embodiment of the present application, the clustering is performed on data points of data to be sent based on a gaussian mixture model, and phase bias corresponding to the data points is obtained, which specifically includes:

According to an embodiment of the present application, the electronic device obtains a frequency offset of the data to be sent based on a phase offset corresponding to the data point, including:

determining the frequency offset of a symbol based on the phase offset corresponding to the data point of any symbol in any time slot of data to be transmitted and the time difference of the symbol relative to a DMRS symbol;

According to an embodiment of the present application, the clustering is performed on data points of any symbol in any time slot of data to be transmitted based on a gaussian mixture model, and phase offset of the symbol is obtained, including:

According to an embodiment of the present application, the clustering, for the data points in the symbol, based on the four gaussian models, correspondingly obtains four types of data points, including:

According to the electronic device of one embodiment of the present application, in each clustering process, after updating parameters in the corresponding gaussian model based on the data points included in each cluster, the operations further include:

According to an embodiment of the present application, the determining that the clustering is finished includes:

the clustering times exceed a preset value or the likelihood value converges.

According to an embodiment of the present application, the determining, according to a position of a cluster center of any type of data point in a constellation diagram corresponding to the symbol, a phase offset of the symbol includes:

In a third aspect, an embodiment of the present application provides a frequency offset obtaining apparatus, including:

the clustering module is used for clustering data points of data to be transmitted based on a Gaussian mixture model to obtain phase deviations corresponding to the data points;

the acquisition module is used for acquiring the frequency offset of the data to be transmitted based on the phase offset corresponding to the data point;

In a fourth aspect, embodiments of the present application provide a non-transitory computer readable storage medium storing a computer program for causing the processor to perform the method provided in the first aspect.

According to the frequency offset obtaining method, the device, the electronic equipment and the storage medium, four clusters are obtained based on the Gaussian mixture model aiming at the data points of the data to be sent, and then the phase offset corresponding to the data points is obtained; based on the phase offset corresponding to the data point, obtaining the frequency offset of the data to be sent; the frequency offset estimation is carried out based on the data points, the number of the pilot frequency columns is not dependent on configuration, the method is suitable for a scene under single-column pilot frequency or even pilot frequency-free configuration, and the problem that a frequency offset solving method by a two-column correlation method is limited when the single-column pilot frequency is configured in the prior art is solved; and the frequency offset estimation is carried out based on the Gaussian mixture model, so that the frequency offset estimation error is effectively reduced, the method is more suitable for actual application scenes, and the robustness is good.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief description will be given below of the drawings that are needed in the embodiments or the prior art descriptions, and it is obvious that the drawings in the following description are some embodiments of the present application, and that other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a schematic diagram of a frequency offset acquisition flow in a two-column pilot frequency scenario according to an embodiment of the present application;

FIG. 2 is a schematic diagram of Kmeans cluster-based data point selection provided in an embodiment of the present application;

FIG. 3 is a schematic diagram of clustered data points based on Kmeans clustering according to an embodiment of the present application;

FIG. 4 is a schematic phase-bias diagram of data points based on Kmeans clusters provided in an embodiment of the present application;

fig. 5 is a schematic diagram of an application scenario of frequency offset estimation provided in an embodiment of the present application;

fig. 6 is a flowchart of a method for obtaining frequency offset according to an embodiment of the present application;

FIG. 7 is a schematic diagram of clustered data points of a Gaussian mixture model according to an embodiment of the disclosure;

FIG. 8 is a schematic diagram of phase bias based on Gaussian mixture model clustering according to an embodiment of the present application;

fig. 9 is a schematic phase deviation diagram based on gaussian mixture model clustering according to another embodiment of the present application;

fig. 10 is a schematic diagram of data points selected for frequency offset estimation provided by an embodiment of the present application;

FIG. 11 is a simulated schematic diagram of data points selected for frequency offset estimation provided by an embodiment of the present application;

fig. 12 is a flowchart of a method for obtaining frequency offset according to another embodiment of the present application;

FIG. 13 is a schematic diagram of simulation of Gaussian mixture model clustering and Kmeans clustering according to an embodiment of the present application;

FIG. 14 is a schematic diagram of simulation of deviation correction after obtaining frequency offset based on Gaussian mixture model clustering and Kmeans clustering according to an embodiment of the present application;

fig. 15 is a schematic structural diagram of an electronic device according to an embodiment of the present application;

fig. 16 is a schematic structural diagram of a frequency offset obtaining apparatus according to an embodiment of the present application.

Detailed Description

In the embodiment of the application, the term "and/or" describes the association relationship of the association objects, which means that three relationships may exist, for example, a and/or B may be represented: a exists alone, A and B exist together, and B exists alone. The character "/" generally indicates that the context-dependent object is an "or" relationship.

The term "plurality" in the embodiments of the present application means two or more, and other adjectives are similar thereto.

The following description of the technical solutions in the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.

Frequency synchronization is an important issue in communications. Frequency offset caused by frequency non-synchronization can cause inter-carrier interference (Carrier Interference, ICI) and lead to an increase in error rate, so that frequency offset estimation is required in the information transmission process. The source of the frequency offset is mainly 2: the first is the crystal error between the transmitter and the receiver. The second is the doppler shift due to the high speed movement of the receiver.

Fig. 1 is a schematic diagram of a frequency offset acquisition flow in a two-column pilot frequency scenario according to an embodiment of the present application, where, as shown in fig. 1, when two columns of pilot frequency symbols are configured, two columns of pilot frequencies may be used to perform correlation operation to obtain a frequency offset value. In actual operation, the influence of noise can be reduced by adopting a plurality of sampling points to perform the operation of correlation averaging, and the frequency offset estimation is more accurate. When more than two columns of pilot symbols are configured, the pilot symbols in multiple columns can be subjected to pairwise correlation averaging, and estimation accuracy can be improved.

In order to increase the transmission rate, a case where only a single column pilot is configured may also occur. In this case, two columns of pilot correlation methods are limited in use. For this case, an embodiment of the present application proposes a scheme based on Kmeans clustering. The method mainly comprises the following steps:

Step1: selecting a data point for acquiring frequency offset; FIG. 2 is a schematic diagram of Kmeans cluster-based data point selection provided in an embodiment of the present application; as shown in fig. 2, constellation points at the periphery of the 16QAM constellation are selected. It will be appreciated that, in order to save computing resources, only some of the points may be selected for clustering.

Step2, clustering.

Kmeans belongs to a relatively simple distance-dependent clustering algorithm. The principle can be expressed as follows: the total class of the classification is specified in advance, and an initial cluster center (center). The point to be classified is then found to be the distance from each class of center point (center). The closest is the corresponding category. And (5) recalculating the clustering center after each training. Intuitively, the Kmeans clustering process is training to get a series of circles, such that each circle covers all points of the class. FIG. 3 is a schematic diagram of clustered data points based on Kmeans clustering according to an embodiment of the present application; as shown in fig. 3, where 4 large circles represent 4 clusters, small circles are constellation points, i.e. data points, and triangles are the center of each cluster.

Step3, in the case of no frequency offset, the cluster center should be the 4 corners of the constellation point of the comparison standard. The angle between the center line of the origin and the coordinate axis should be 45 degrees. FIG. 4 is a schematic phase-bias diagram of data points based on Kmeans clusters provided in an embodiment of the present application; as shown in fig. 4, in the case of a frequency offset, the constellation diagram is deflected as a whole, and the deflection angle is phase offset a. The conversion relation between the phase offset and the frequency offset is as follows: a=2×pi×f×t, where f is a normalized frequency offset value, t is a time difference between a symbol where a data point in a constellation is located and a DMRS symbol, and pi is a circumferential rate pi. Therefore, in the case of frequency offset, the center point of the cluster will also rotate by the same angle. In this embodiment, the phase offset can be calculated by using the angle and the standard 45 degrees, and then the frequency offset value of the symbol where the data point in the constellation diagram is located can be calculated.

However, since Kmeans clustering is a clustering scheme that covers as many data points as possible in one circle. When the shape of the data points of the constellation diagram changes, the center point (center) of the constellation diagram sometimes shifts, and then a measurement error is introduced when the frequency offset is calculated with a standard point, so that the performance is not robust enough. And Kmeans clustering requires more data points to perform training iteration to determine the position of the central point of the cluster, and has higher calculation complexity.

In addition, scheme constellation points based on Kmeans clusters exceeding pi/4 can cause confusion. For example, when configuring single-column pilot, the DMRS is on symbol 3, so that the most distant symbol (14 th symbol in one slot) does not exceed the upper limit, the normalized frequency offset estimation range is:

taking a subcarrier spacing of 30K as an example, the maximum frequency offset estimation range is: 342HZ. Therefore, the frequency offset estimation range is limited.

Therefore, the embodiment of the application provides a frequency offset acquisition method and device, which mainly relate to a frequency offset estimation scheme based on a Gaussian mixture model under single-column pilot frequency configuration. The scheme has a larger frequency offset estimation range and more robust estimation performance, and is used for ensuring that more accurate frequency offset is acquired in more scenes. Fig. 5 is a schematic diagram of an application scenario of frequency offset estimation provided in an embodiment of the present application, as shown in fig. 5, where the frequency offset estimation provided in the embodiment of the present application may be generally applied to channel estimation, so that a receiving end may correctly demodulate an initially transmitted signal.

In the embodiments of the present application, the methods and apparatuses are based on the same application conception, and since the principles of solving the problems by the methods and apparatuses are similar, the implementation of the apparatuses and the methods may be referred to each other, and the repetition is not repeated.

Fig. 6 is a flowchart of a method for obtaining frequency offset according to an embodiment of the present application, as shown in fig. 6, where the method includes the following steps:

step 600, clustering is carried out on data points of data to be transmitted based on a Gaussian mixture model, and phase deviations corresponding to the data points are obtained;

specifically, the neural network and the machine learning algorithm can be classified into supervised learning and unsupervised learning. Both the Gaussian mixture model and the kmeans cluster belong to an unsupervised learning method. Gaussian mixture models different numbers of gaussian models are used to fit the data distribution. In the training process, the mean value and the variance of each Gaussian model are obtained through the features of the learning data, and finally the classifier is formed. According to this classifier, data points can be classified according to mean and variance during the test phase (classification phase).

Intuitively, the Keans training process chooses as much as possible a radius of a circle so that the circle encompasses all the data points of the training process; the Gaussian mixture model training process is to choose an ellipse as far as possible to fit the data points, and the round-flattening degree of the ellipse depends on the mean value and the variance of the Gaussian function. Therefore, the ellipse shape has more flexibility than circular fitting, and can be better adapted to the shapes of constellation points generated under different channel conditions.

Specifically, in this embodiment, a gaussian mixture model may be used to perform data feature fitting, and data points of data to be sent are clustered, and further based on the clustered data points, a corresponding phase offset is obtained.

Specifically, the present embodiment performs frequency offset estimation based on data points, does not depend on configuration pilot frequency columns to perform correlation, and can be used in single-column pilot frequency or even in a non-pilot frequency extreme scene.

Step 610, obtaining a frequency offset of the data to be sent based on the phase offset corresponding to the data point;

specifically, after the phase offset corresponding to the data point is obtained, the frequency offset of the time slot where the data point is located can be obtained based on the phase offset corresponding to the data point and used as the frequency offset of the data to be sent.

Specifically, the conversion relation between the phase offset and the frequency offset is as follows: a=2×pi×f×t, where f is a normalized frequency offset value, t is a time difference between a symbol where a data point in a constellation is located and a DMRS symbol, and pi is a circumferential rate pi.

For example, if the phase offset of the time slot where the data point is located is a1, the frequency offset of the time slot where the data point is located can be obtained by calculating based on the conversion relation a=2×pi×f×t of the phase offset and the frequency offset, and it can be understood that for the data to be sent, the frequency offset of each time slot is equal, so that after the frequency offset of any time slot is obtained, the frequency offset of the data to be sent can be used as the frequency offset of the data to be sent;

Specifically, the frequency offset of more than one time slot can be obtained, and then the average value of the frequency offset of each time slot is taken as the frequency offset of the data to be transmitted, so that the error is reduced.

Specifically, in this embodiment, in order to correspond to the number of constellations in the constellation diagram, the gaussian mixture model may be set to four gaussian models, which respectively correspond to four clusters, that is, when the data points to be transmitted are clustered, the data points may be divided into four clusters.

According to the frequency offset acquisition method, four clusters are obtained based on a Gaussian mixture model aiming at data points of data to be transmitted, and then phase offsets corresponding to the data points are obtained; based on the phase offset corresponding to the data point, obtaining the frequency offset of the data to be sent; the frequency offset estimation is carried out based on the data points, the number of the pilot frequency columns is not dependent on configuration, the method is suitable for a scene under single-column pilot frequency or even pilot frequency-free configuration, and the problem that a frequency offset solving method by a two-column correlation method is limited when the single-column pilot frequency is configured in the prior art is solved; and the frequency offset estimation is carried out based on the Gaussian mixture model, so that the frequency offset estimation error is effectively reduced, the method is more suitable for actual application scenes, and the robustness is good.

Optionally, in the foregoing embodiments, the clustering, based on a gaussian mixture model, of the data points of the data to be sent, and obtaining a phase offset corresponding to the data points specifically includes:

Specifically, since the frequency offsets of all the time slots on the data to be transmitted are equal, and the frequency offsets of all the symbols in one time slot are also equal under ideal conditions, when the frequency offset is acquired, the frequency offset of any symbol in any time slot in the data to be transmitted can be acquired as the frequency offset of the time slot, and further as the frequency offset of the data to be transmitted.

Therefore, in this embodiment, for the data point of any symbol in any time slot of the data to be sent, clustering may be performed based on the gaussian mixture model, so as to obtain the phase offset of the symbol, and be used for subsequently obtaining the frequency offset of the symbol, so as to further determine the frequency offset of the time slot where the symbol is located.

Optionally, in the foregoing embodiments, obtaining the frequency offset of the data to be sent based on the phase offset corresponding to the data point includes:

Specifically, after the phase offset corresponding to the data point of any symbol in any time slot of the data to be sent is obtained, the frequency offset f of the symbol can be obtained by calculation based on the phase offset a of the symbol and the time difference t of the symbol relative to the DMRS symbol;

specifically, the conversion relation between the phase offset and the frequency offset is as follows: a=2×pi×f×t, where f is a normalized frequency offset value, t is a time difference between a symbol where a data point in a constellation is located and a DMRS symbol, and pi is a circumferential rate pi. The frequency offset f of the symbol can be obtained by calculation according to the conversion formula, the phase offset a brought into the symbol and the time difference t of the symbol relative to the DMRS symbol;

Specifically, since the frequency offsets of all symbols in a slot are ideally identical, the frequency offset of any symbol in the slot can be taken as the frequency offset of the slot.

It can be understood that when calculating the frequency offset, the frequency offset is inevitably interfered by external factors, so that the frequency offset is obtained inaccurately, namely the frequency offset is not the frequency offset under the ideal condition, such as inaccurate measurement when measuring the phase offset value, etc.; therefore, in order to obtain more accurate frequency offset of a time slot, the frequency offset of the time slot can be determined based on the average value of the frequency offset of a plurality of symbols in the time slot; specifically, the frequency offset values of all symbols in the time slot can be obtained and then averaged to be used as the frequency offset of the time slot.

Optionally, in the foregoing embodiments, the clustering, based on a gaussian mixture model, of the data points of any symbol in any time slot of the data to be sent, to obtain phase offset of the symbol includes:

Specifically, when obtaining the frequency offset of any symbol in any time slot of the data to be sent, the data points in the symbol may be clustered based on the gaussian mixture model to obtain four clusters, and fig. 7 is a schematic diagram of the clustered data points of the gaussian mixture model provided in an embodiment of the present application, as shown in fig. 7, each cluster corresponds to an ellipse, and the center of each ellipse is the mean value of the gaussian model, that is, the cluster center point. The phase bias for the symbol may then be obtained based on the cluster center.

When the phase bias of the symbol is obtained based on the clustering center, because there are four clusters in total, there are four clustering centers, and the mean parameter mu of the Gaussian model can be directly used _i And (3) representing. Because the data points are offset integrally when phase offset occurs, the position of the clustering center of any clustering data point in the constellation diagram can be selected, and the phase offset of the symbol can be determined. For example, fig. 8 is a schematic phase diagram of gaussian mixture model clustering according to an embodiment of the present application, as shown in fig. 8, an included angle between a line between any cluster center and origin of coordinates and a standard constellation point may be obtained, and it may be understood that an angular bisector of the standard constellation point between coordinate axesTherefore, the included angle a between the connecting line of any cluster center and the origin of the constellation diagram and the angular bisector of the coordinate axis, namely the straight line which keeps standard 45 degrees with the coordinate axis, can be obtained, namely the phase deviation value of the symbol where the data point is located.

Optionally, in the foregoing embodiments, the clustering, for the data points in the symbol, based on the four gaussian models, correspondingly acquires four types of data points, including:

specifically, the gaussian mixture model training data has no label data in advance, i.e. it is not known in advance from which constellation the data points come. Training the gaussian mixture model at this time requires the use of the EM algorithm (Expectation-Maximization algorithm, maximum Expectation algorithm). The core idea of the EM algorithm Jie Gaosi hybrid model is that during the training process, likelihood probabilities from different constellation points are first calculated from the features of the training data, and the mean, variance, and mixing coefficients of the gaussian hybrid model are calculated using this likelihood probability. And then calculating a probability density function of the Gaussian mixture model, and carrying out iteration to obtain the optimal mean value, variance and mixture coefficient.

Taking 16QAM (Quadrature Amplitude Modulation ) as an example, modulating data points in a certain symbol by 16QAM to obtain a constellation diagram, and then selecting partial data points for frequency offset estimation; and then carrying out a Gaussian mixture model and EM algorithm training process on the selected data points to form a classifier for classification, and obtaining four clustered data points.

Specifically, each clustering process corresponds to one training process, and in each clustering process, data points included in each cluster can be respectively obtained based on four Gaussian models updated in the last clustering process; it can be understood that if the current clustering process is the first clustering, clustering is directly performed based on the initialized four gaussian models;

specifically, after updating based on the last clustering processRespectively obtaining data points included in each cluster; specifically, for each data point, four probabilities that the data point may belong to four clusters, such as the probability ω of the data j belonging to the class i, may be calculated based on the E-step of the EM algorithm first _ji Wherein i=1, 2,3,4; and determining the cluster corresponding to the maximum probability, namely the cluster to which the data point belongs in the current clustering process, and determining the data point included in each of the four clusters after all the data points determine the cluster to which the data point belongs in the current clustering process.

Specifically, in calculating the probability ω of the data j belonging to the category i _ji When (1):

wherein z is _j Indicating the category to which the data j belongs, x _j The data j is represented by a representation of,is the mixing coefficient of the last iteration, mu _i Is the mean value parameter of the Gaussian model corresponding to the class i, sigma _i Covariance parameters of the Gaussian model corresponding to class i,/->Is the mixing coefficient of the gaussian model corresponding to class i.

specifically, in the current clustering process, after determining the data points included in each category i, the parameters of the gaussian mixture model can be updated based on the M steps of the EM algorithm; i.e. all data points included in each category i and the probability of each data point in the category are brought into the Gaussian model corresponding to the category, and the parameters of the Gaussian model corresponding to the category are updated;

specifically, the parameters of the gaussian model corresponding to the class i may be updated as:

it can be understood that the updated gaussian models of four categories obtained after updating the parameters in the current clustering process are used for obtaining the probability of each data point in each category in the next clustering process.

Specifically, after all the clustering processes are determined to be finished, four clusters, namely four types of data points, obtained in the last clustering process can be obtained; it can be understood that, in this embodiment, the optimal gaussian mixture model is obtained through a clustering process of updating the gaussian mixture model for a plurality of times, that is, through a plurality of iterations, and based on the optimal gaussian mixture model, four optimal clusters, that is, four types of data points with optimal clustering results, can be obtained.

Optionally, in the foregoing embodiments, in each clustering process, after updating, for each cluster, parameters in its corresponding gaussian model based on the data points included in the cluster, the method further includes:

Specifically, in order to better judge the iterative updating effect of the Gaussian mixture model, the iterative updating effect can be described by a likelihood function;

in this embodiment, after parameters of gaussian models corresponding to four categories respectively are obtained in each clustering process, whether further iteration is needed or not may be determined based on the likelihood function; for example, if likelihood values LL obtained in a continuous multi-clustering process are not significantly increased any more, it can be considered that the gaussian mixture model has been updated to be a classifier with better effect; wherein the likelihood function is:

Optionally, in the foregoing embodiments, the determining that the clustering is ended includes:

the clustering times exceed a preset value or the likelihood value converges.

Specifically, when it is determined that all the clustering processes are finished, it may be determined based on likelihood functions, that is, likelihood values LL obtained in consecutive clustering processes do not increase significantly any more, and it may be determined that the clustering processes may stop iterating;

specifically, in this embodiment, the iteration number of the clustering process may also be preset, for example, 1000 clustering processes may be preset, and then all the clusters may be determined to be ended when the 1000 th clustering process is ended.

Optionally, in the foregoing embodiments, determining the phase offset of the symbol according to the position of the cluster center of any type of data point in the constellation diagram corresponding to the symbol includes:

Specifically, when determining the phase offset of a symbol based on the position of any cluster center in the constellation diagram corresponding to the symbol after the data point is clustered in a certain symbol, if the phase offset of the symbol is greater than pi/4, i.e. the symbol is offset to another quadrant, for example, the symbol is offset from the second quadrant to the third quadrant, but when calculating the phase offset of the symbol, the phase offset may be directly calculated based on an angular bisector, the wrong phase offset may be obtained, and further the frequency offset estimation error is caused; for example, fig. 9 is a schematic phase diagram of a gaussian mixture model clustering according to another embodiment of the present application, where the rotation angle does not cause ambiguity if the constellation area of a certain symbol rotates to a smaller angle as shown in fig. 9, but if the constellation area a spans a quadrant if the constellation area a rotates to a larger angle as shown in fig. 9, the true rotation angle a1 is easily mistaken as b1 that the third quadrant standard constellation point rotates counterclockwise.

Therefore, the conversion relationship between the phase offset and the frequency offset in this embodiment is: a=2×pi×f×t; the characteristic that the frequency offsets of all symbols in a time slot are basically consistent and the time difference t of each symbol relative to a DMRS symbol is in equal proportion increment is fully utilized, the rotation direction of a constellation area of the symbol currently carrying out frequency offset estimation and the position range in which a clustering center is possibly positioned are estimated based on the phase offset values of other symbols, and the real phase offset value of the symbol can be further obtained; for example, the real phase offset value of the symbol can be determined by combining the phase offset value of the last symbol of the symbol, so that the problem of limited frequency offset estimation range can be effectively solved, and the frequency offset estimation range is enlarged.

For example, on the PUSCH (Physical Uplink Shared Channel, physical layer uplink shared channel) channel, 1 frame of 14 symbols, single column pilot is configured on symbol 3 by default. Under the condition, kmeans clustering ensures that when the frequency offset of the farthest symbol 14 does not exceed pi/4 and the subcarrier spacing of 30KHZ, the maximum frequency offset estimation range of Kmeans is 341HZ; the gaussian mixture model in this embodiment still maintains better performance under the frequency offset 2600HZ after joint decision.

Optionally, in the foregoing embodiments, the data points for one symbol in the timeslot are clustered based on a gaussian mixture model, specifically:

Specifically, in this embodiment, when the frequency offset of a symbol is obtained based on a data point of the symbol, only a part of data points may be selected for obtaining the frequency offset because the constellation diagram is the overall offset; because the Gaussian mixture model selects an ellipse as much as possible to fit the data points, the frequency offset estimation can be performed based on only the data points with the amplitude smaller than the preset amplitude in one symbol. Taking 16QAM as an example, fig. 10 is a schematic diagram of data points selected for frequency offset estimation according to an embodiment of the present application; fig. 11 is a schematic diagram of simulation of data points selected for frequency offset estimation according to an embodiment of the present application, as shown in fig. 10 and 11, the data points in the circles may be selected as data points with a magnitude smaller than a preset magnitude.

It can be appreciated that, in this embodiment, the points in the constellation are selected for processing, so that the selection of the data points can be reduced theoretically, and the operand is reduced.

Fig. 12 is a flowchart of a method for obtaining frequency offset according to another embodiment of the present application, as shown in fig. 12, where the method includes the following steps:

step 1200, selecting points in a constellation diagram;

specifically, when the frequency offset of a certain symbol is acquired based on the data point of the symbol, the data point is modulated first to obtain a corresponding constellation diagram.

The frequency offset can cause the overall offset of the constellation diagram, so that only partial data points can be selected for frequency offset acquisition; in this embodiment, the frequency offset estimation may be performed based on only one data point having an intra-symbol amplitude smaller than a preset amplitude.

Step 1210, training a Gaussian mixture model;

specifically, training is performed on the Gaussian mixture model, so that clustering can be performed based on four Gaussian models after training, and four clustered data points are correspondingly acquired;

specifically, during each training process, i.e. each clustering process, for each data point, four probabilities that such data point may belong to four clusters, such as the probability ω of the data j belonging to the category i, may be calculated based on the E-step of the EM algorithm first _ji Wherein i=1, 2,3,4; and determining the cluster corresponding to the maximum probability, namely the cluster to which the data point belongs in the current clustering process, and determining the data point included in each of the four clusters after all the data points determine the cluster to which the data point belongs in the current clustering process.

Parameters of its corresponding gaussian model may then be updated for each cluster based on the data points it includes: mean parameter mu of Gaussian model corresponding to class i _i Covariance parameter sigma of Gaussian model corresponding to class i _i And the mixing coefficient of the Gaussian model corresponding to the class iWherein the method comprises the steps ofi＝1，2，3，4。

Specifically, in the current training process, after determining the data points included in each category i, the parameters of the gaussian mixture model can be updated based on the M steps of the EM algorithm; i.e. all data points included in each category i and the probability of each data point in the category are brought into the Gaussian model corresponding to the category, and the parameters of the Gaussian model corresponding to the category are updated;

Step 1220, obtaining each center of the gaussian model;

each Gaussian model corresponds to a cluster, each cluster corresponds to a cluster center, and the mean parameter mu of the Gaussian model can be directly used _i And (3) representing. The cluster center point, i.e. the mean value parameter mu, can then be based on _i Representing the position on the constellation to obtain the phase offset of the symbol.

Step 1230, calculating frequency offset according to the center and the standard constellation points;

when the phase bias of the symbol is acquired based on the cluster center, because there are four clusters in total, there are four cluster centers, but because the data points are offset as a whole when phase bias occurs, the position of the cluster center of any cluster data point in the constellation diagram can be selected, and the phase bias of the symbol is determined. Specifically, an included angle between a connecting line of any cluster center and the origin of coordinates and a standard constellation point can be obtained, and it can be understood that the standard constellation point is on an angular bisector between coordinate axes, so that an angular bisector between a connecting line of any cluster center and the origin of a constellation diagram and the coordinate axes, namely a straight line which keeps standard 45 degrees with the coordinate axes, can be obtained, and the included angle a between the connecting line of any cluster center and the origin of the constellation diagram is the phase deviation value of a symbol where the data point is located.

Specifically, after the phase offset value of the symbol is obtained, the frequency offset of the symbol can be obtained according to the relation between the phase offset value and the frequency offset;

Specifically, the conversion relation between the phase offset and the frequency offset is as follows: a=2×pi×f×t, where f is a normalized frequency offset value, t is a time difference between a symbol where a data point in a constellation is located and a DMRS symbol, and pi is a circumferential rate pi. The frequency offset f of the symbol can be obtained by calculating according to the conversion formula, the phase offset a carried into the symbol, and the time difference t of the symbol relative to the DMRS symbol.

Specifically, after obtaining the frequency offset of a symbol, the frequency offset of the time slot can be determined based on the frequency offset of at least one symbol in the time slot to which the symbol belongs.

Step 1240, compensating the frequency offset value.

Specifically, after the frequency offset of a time slot is obtained, the time slot may be compensated based on the frequency offset value. It can be understood that in this embodiment, frequency offsets of all time slots in the same data to be transmitted are consistent; the frequency offset compensation can be performed on the data to be transmitted.

Fig. 13 is a schematic diagram of simulation of gaussian mixture model clustering and Kmeans clustering provided in an embodiment of the present application, fig. 14 is a schematic diagram of simulation of deviation correction after obtaining a frequency deviation based on gaussian mixture model clustering and Kmeans clustering provided in an embodiment of the present application, and as shown in fig. 13 and fig. 14, is a graph of comparison between Kmeans clustering and gaussian mixture model of frequency deviation 300HZ (subcarrier spacing 30kHZ normalized frequency deviation 0.01) with a signal-to-noise ratio of 15dB symbol 9. The data points selected by the gaussian mixture model are the data points in the circle in fig. 13, the data points with smaller amplitude, and the data points selected by Kmeans clusters are the data points outside the circle in fig. 13. It can be seen that the center points of the Keans clusters deviate, and the center points of the Gaussian mixture model are better selected. As shown in fig. 14, the deviation rectifying effect after obtaining the frequency offset based on the gaussian mixture model is better, that is, the frequency offset value obtained based on the gaussian mixture model is more accurate.

In the embodiment of the application, the Gaussian mixture model is used for fitting the data points, so that the method is more flexible compared with Kmeans clustering, can adapt to different constellation diagram shape distribution, and is accurate in estimation.

The technical scheme provided by the embodiment of the application can be suitable for various systems, in particular to a 5G system. For example, suitable systems may be global system for mobile communications (global system of mobile communication, GSM), code division multiple access (code division multiple access, CDMA), wideband code division multiple access (Wideband Code Division Multiple Access, WCDMA) universal packet Radio service (general packet Radio service, GPRS), long term evolution (long term evolution, LTE), LTE frequency division duplex (frequency division duplex, FDD), LTE time division duplex (time division duplex, TDD), long term evolution-advanced (long term evolution advanced, LTE-a), universal mobile system (universal mobile telecommunication system, UMTS), worldwide interoperability for microwave access (worldwide interoperability for microwave access, wiMAX), 5G New air interface (New Radio, NR), and the like. Terminal devices and network devices are included in these various systems. Core network parts such as evolved packet system (Evloved Packet System, EPS), 5G system (5 GS) etc. may also be included in the system.

The terminal according to the embodiments of the present application may be a device that provides voice and/or data connectivity to a user, a handheld device with a wireless connection function, or other processing device connected to a wireless modem, etc. The names of terminals may also be different in different systems, for example in a 5G system, a terminal may be referred to as User Equipment (UE). The wireless terminal device may communicate with one or more Core Networks (CNs) via a radio access Network (Radio Access Network, RAN), which may be mobile terminal devices such as mobile phones (or "cellular" phones) and computers with mobile terminal devices, e.g., portable, pocket, hand-held, computer-built-in or vehicle-mounted mobile devices that exchange voice and/or data with the radio access Network. Such as personal communication services (Personal Communication Service, PCS) phones, cordless phones, session initiation protocol (Session Initiated Protocol, SIP) phones, wireless local loop (Wireless Local Loop, WLL) stations, personal digital assistants (Personal Digital Assistant, PDAs), and the like. The wireless terminal device may also be referred to as a system, subscriber unit (subscriber unit), subscriber station (subscriber station), mobile station (mobile), remote station (remote station), access point (access point), remote terminal device (remote terminal), access terminal device (access terminal), user terminal device (user terminal), user agent (user agent), user equipment (user device), and the embodiments of the present application are not limited.

The base station according to the embodiment of the application may include a plurality of cells for providing services for the terminal. A base station may also be called an access point or may be a device in an access network that communicates over the air-interface, through one or more sectors, with wireless terminal devices, or other names, depending on the particular application. The base station may be configured to exchange received air frames with internet protocol (Internet Protocol, IP) packets as a router between the wireless terminal device and the rest of the access network, which may include an Internet Protocol (IP) communication network. The base station may also coordinate attribute management for the air interface. For example, the network device according to the embodiments of the present application may be a network device (Base Transceiver Station, BTS) in a global system for mobile communications (Global System for Mobile communications, GSM) or code division multiple access (Code Division Multiple Access, CDMA), a network device (NodeB) in a wideband code division multiple access (Wide-band Code Division Multiple Access, WCDMA), an evolved network device (evolutional Node B, eNB or e-NodeB) in a long term evolution (long term evolution, LTE) system, a 5G base station (gNB) in a 5G network architecture (next generation system), a home evolved base station (Home evolved Node B, heNB), a relay node (relay node), a home base station (femto), a pico base station (pico), and the like. In some network structures, a base station may include a Centralized Unit (CU) node and a Distributed Unit (DU) node, which may also be geographically separated.

Fig. 15 is a schematic structural diagram of an electronic device according to an embodiment of the present application, as shown in fig. 15, the base station includes a memory 1501, a transceiver 1502 and a processor 1503, wherein:

the memory 1501 is used to store a computer program; the transceiver 1502 is configured to transmit and receive data under the control of the processor; the processor 1503 is configured to read the computer program in the memory and perform the following operations:

A transceiver 1502 for receiving and transmitting data under the control of a processor 1503.

Wherein in fig. 15, a bus architecture may comprise any number of interconnected buses and bridges, and in particular one or more processors represented by processor 1503 and various circuits of memory represented by memory 1502, linked together. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, power management circuits, etc., which are well known in the art and, therefore, will not be described further herein. The bus interface provides an interface. The transceiver 1502 may be a number of elements, i.e., include a transmitter and a receiver, providing a means for communicating with various other apparatus over a transmission medium, including wireless channels, wired channels, optical cables, and the like. The processor 1503 is responsible for managing the bus architecture and general processing, and the memory 1501 may store data used by the processor 1503 in performing operations.

The processor 1503 may be a Central Processing Unit (CPU), application specific integrated circuit (Application Specific Integrated Circuit, ASIC), field-programmable gate array (Field-Programmable Gate Array, FPGA), or complex programmable logic device (Complex Programmable Logic Device, CPLD), or it may employ a multi-core architecture.

It should be noted that, the above device provided in the embodiment of the present invention can implement all the method steps implemented in the method embodiment and achieve the same technical effects, and detailed descriptions of the same parts and beneficial effects as those in the method embodiment in this embodiment are omitted.

Based on any one of the above embodiments, the clustering is performed on the data points of the data to be sent based on a gaussian mixture model, and the phase offset corresponding to the data points is obtained, which specifically includes:

Specifically, the electronic device provided in the embodiment of the present application can implement all the method steps implemented in the embodiment of the method, and can achieve the same technical effects, and the parts and beneficial effects that are the same as those of the embodiment of the method in the embodiment are not described in detail herein.

Based on any of the above embodiments, obtaining the frequency offset of the data to be sent based on the phase offset corresponding to the data point includes:

Based on any of the above embodiments, the clustering is performed on the data points of any symbol in any time slot of the data to be sent based on a gaussian mixture model, and the phase offset of the symbol is obtained, including:

Based on any of the foregoing embodiments, the clustering, based on the four gaussian models, of the data points in the symbol, correspondingly obtains four types of data points, including:

Based on any of the above embodiments, in each clustering process, after updating parameters in its corresponding gaussian model based on the data points included in each cluster, the operations further include:

Specifically, the device provided in the embodiment of the present application can implement all the method steps implemented in the embodiment of the method, and can achieve the same technical effects, and the parts and beneficial effects that are the same as those of the embodiment of the method in the embodiment are not described in detail herein.

Based on any of the above embodiments, the determining that the clustering is ended includes:

the clustering times exceed a preset value or the likelihood value converges.

Based on any of the above embodiments, determining the phase offset of the symbol according to the position of the cluster center of any type of data point in the constellation diagram corresponding to the symbol includes:

Based on any of the foregoing embodiments, the clustering is performed on the data points of one symbol in the time slot based on a gaussian mixture model, specifically:

According to the electronic equipment provided by the embodiment of the application, four clusters are obtained based on the Gaussian mixture model aiming at the data points of the data to be sent, and then the phase offset corresponding to the data points is obtained; based on the phase offset corresponding to the data point, obtaining the frequency offset of the data to be sent; the frequency offset estimation is carried out based on the data points, the number of the pilot frequency columns is not dependent on configuration, the method is suitable for a scene under single-column pilot frequency or even pilot frequency-free configuration, and the problem that a frequency offset solving method by a two-column correlation method is limited when the single-column pilot frequency is configured in the prior art is solved; and the frequency offset estimation is carried out based on the Gaussian mixture model, so that the frequency offset estimation error is effectively reduced, the method is more suitable for actual application scenes, and the robustness is good.

The frequency offset obtaining device provided in the embodiment of the present application is described below, and the frequency offset obtaining device described below and the frequency offset obtaining method described above may be referred to correspondingly.

Fig. 16 is a schematic structural diagram of a frequency offset obtaining apparatus according to an embodiment of the present application, as shown in fig. 16, where the frequency offset obtaining apparatus includes: a clustering module 1610 and an acquisition module 1620; wherein:

the clustering module 1610 is configured to perform clustering based on a gaussian mixture model for data points of data to be sent, and obtain phase offsets corresponding to the data points;

the obtaining module 1620 is configured to obtain a frequency offset of the data to be sent based on the phase offset corresponding to the data point;

Specifically, the frequency offset obtaining device performs clustering on data points of data to be sent based on a Gaussian mixture model through a clustering module 1610 to obtain phase offsets corresponding to the data points; and then obtaining the frequency offset of the data to be sent based on the phase offset corresponding to the data point by the obtaining module 1620.

According to the frequency offset acquisition device provided by the embodiment of the application, four clusters are obtained based on the Gaussian mixture model aiming at the data points of the data to be transmitted, so that the phase offset corresponding to the data points is acquired; based on the phase offset corresponding to the data point, obtaining the frequency offset of the data to be sent; the frequency offset estimation is carried out based on the data points, the number of the pilot frequency columns is not dependent on configuration, the method is suitable for a scene under single-column pilot frequency or even pilot frequency-free configuration, and the problem that a frequency offset solving method by a two-column correlation method is limited when the single-column pilot frequency is configured in the prior art is solved; and the frequency offset estimation is carried out based on the Gaussian mixture model, so that the frequency offset estimation error is effectively reduced, the method is more suitable for actual application scenes, and the robustness is good.

It should be noted that, the above device provided in this embodiment of the present application can implement all the method steps implemented in the method embodiment and achieve the same technical effects, and detailed descriptions of the same parts and beneficial effects as those of the method embodiment in this embodiment are omitted.

It should be noted that, in the embodiment of the present application, the division of the units is schematic, which is merely a logic function division, and other division manners may be implemented in actual practice. In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a processor-readable storage medium. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution, in the form of a software product stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (processor) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

Based on any of the foregoing embodiments, the present application further provides a processor-readable storage medium, where the processor-readable storage medium stores a computer program, where the computer program is configured to cause the processor to perform the method provided in the foregoing embodiments, for example, including:

The computer program stored on the processor readable storage medium according to this embodiment enables the processor to implement all the method steps implemented by the method embodiment and achieve the same technical effects, and details of the same parts and advantages as those of the method embodiment in this embodiment are not described herein.

The processor-readable storage medium may be any available medium or data storage device that can be accessed by a processor, including, but not limited to, magnetic storage (e.g., floppy disks, hard disks, magnetic tape, magneto-optical disks (MOs), etc.), optical storage (e.g., CD, DVD, BD, HVD, etc.), semiconductor storage (e.g., ROM, EPROM, EEPROM, nonvolatile storage (NAND FLASH), solid State Disk (SSD)), and the like.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, magnetic disk storage, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer-executable instructions. These computer-executable instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These processor-executable instructions may also be stored in a processor-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the processor-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These processor-executable instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the spirit or scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims and the equivalents thereof, the present application is intended to cover such modifications and variations.

Claims

1. The frequency offset obtaining method is characterized by comprising the following steps:

the Gaussian mixture model comprises four Gaussian models, and the four Gaussian models correspond to four clusters respectively;

the clustering is performed on the data points of the data to be sent based on a Gaussian mixture model, and the phase offset corresponding to the data points is obtained, specifically including:

clustering data points of any symbol in any time slot of data to be transmitted based on a Gaussian mixture model to acquire phase bias of the symbol;

clustering the data points of any symbol in any time slot of the data to be sent based on a Gaussian mixture model to obtain the phase offset of the symbol, wherein the clustering comprises the following steps:

determining phase offset of the symbol according to the position of the clustering center of any type of data point in the constellation diagram corresponding to the symbol;

clustering is performed on the data points in the symbol based on the four gaussian models, and four types of data points are correspondingly acquired, including:

2. The method of obtaining a frequency offset according to claim 1, wherein obtaining the frequency offset of the data to be transmitted based on the phase offset corresponding to the data point includes:

3. The method of claim 1, wherein during each clustering, after updating parameters in its corresponding gaussian model based on the data points included in each cluster, the method further comprises:

4. The method for acquiring frequency offset according to claim 3, wherein: the determining that the clustering is finished comprises:

the clustering times exceed a preset value or the likelihood value converges.

5. The method of claim 1, wherein determining the phase offset of the symbol according to the location of the cluster center of any type of data point in the constellation corresponding to the symbol comprises:

6. The method for obtaining a frequency offset according to any one of claims 1 to 5, wherein the clustering is performed on the data points of one symbol in the time slot based on a gaussian mixture model, specifically:

7. An electronic device comprising a memory, a transceiver, and a processor:

8. The electronic device of claim 7, wherein obtaining the frequency offset of the data to be transmitted based on the phase offset corresponding to the data point comprises:

9. The electronic device of claim 7, wherein during each clustering, after updating parameters in its corresponding gaussian model based on the data points it includes for each cluster, the operations further comprise:

10. The electronic device of claim 9, wherein: the determining that the clustering is finished comprises:

the clustering times exceed a preset value or the likelihood value converges.

11. The electronic device of claim 7, wherein determining the phase bias of the symbol based on the location of the cluster center of any type of data point in the constellation corresponding to the symbol comprises:

12. The electronic device according to any of the claims 7 to 11, wherein the data points for one symbol in the time slot are clustered based on a gaussian mixture model, in particular:

13. A frequency offset acquisition apparatus, comprising:

the clustering module is specifically configured to:

14. A processor-readable storage medium storing a computer program for causing the processor to perform the frequency offset acquisition method of any one of claims 1 to 6.