CN112566214B - User clustering method, system, storage medium, computer device and application - Google Patents
User clustering method, system, storage medium, computer device and application Download PDFInfo
- Publication number
- CN112566214B CN112566214B CN202011267021.1A CN202011267021A CN112566214B CN 112566214 B CN112566214 B CN 112566214B CN 202011267021 A CN202011267021 A CN 202011267021A CN 112566214 B CN112566214 B CN 112566214B
- Authority
- CN
- China
- Prior art keywords
- user
- initial
- users
- sample
- clustering
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W40/00—Communication routing or communication path finding
- H04W40/24—Connectivity information management, e.g. connectivity discovery or connectivity update
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention belongs to the technical field of wireless communication, and discloses a user clustering method, a system, a storage medium, computer equipment and application, wherein initial user clustering is carried out according to channel gains of all users in the system and channel similarity among the users; calculating initial model parameters of the Gaussian mixture model according to an initial user clustering result; calculating the posterior probability of each sample generated by each mixed component and iteratively updating the parameters of the Gaussian mixture model; and judging whether the iteration times reach a set termination condition, if so, clustering the users according to the cluster marks, otherwise, calculating the posterior probability generated by each mixed component of each sample and updating the parameters of the Gaussian mixture model in an iterative manner. The method optimizes the initial model parameters of the Gaussian mixture model, improves the system throughput compared with the traditional user clustering method, and can be used for a multi-antenna non-orthogonal multiple access system.
Description
Technical Field
The invention belongs to the technical field of wireless communication, and particularly relates to a user clustering method, a user clustering system, a user clustering storage medium, a computer device and application.
Background
At present: the user clustering method is a key technology which has a large influence on the throughput of a multi-antenna non-orthogonal multiple access system. The method of traversing search clustering is a method for pursuing the maximization of system performance. BenJEBBOVU A et al, in "Concept and Practical Considerations of Non-Orthogonal Multiple Access (NOMA) for Future Radio Access" (International Symposium on Intelligent Signal Processing and Communications systems. IEEE 2014 770-774), propose a traversal search clustering method that traverses all possible combinations of user clusters, so that an optimal solution or a locally optimal solution of the target function can always be obtained. However, the traversal process of the traversal search clustering method is very complicated and is not suitable for an actual system.
The random user clustering method is a user clustering method which pursues simplicity. Al-Obiedollah et Al, in "Energy Efficient Beamforming Design for MISO Non-Orthogonal Multiple Access Systems" (IEEE Transactions on Communications,2019, 67 (6): 4117-4131), mention a random user clustering method, which allocates users on appropriate sub-bands according to their required bandwidths and their target information rates, and only considers the amount of resources required by the users themselves in the user deployment process.
The pre-grouping clustering method clusters users according to the long-term signal-to-interference-and-noise ratios of the users, and then randomly selects a plurality of users from the user clusters as the users in the same cluster. Ding Z et al, in "Impact of User Pairing on 5G Non-orthogonal Multiple-Access Downlink Transmissions" (IEEE Transactions on Vehicular Technology,2016,65 (8): 6010-6023), mention that there may be problems of uneven distribution of strong and weak User sets or more concentrated users at interval values during the implementation of the pre-grouping clustering method, which further causes the selected users in the User cluster to be unable to reach a certain channel quality interval and affects the system to obtain higher system throughput.
The fixed grouping algorithm firstly groups users according to the channel gain of the users, and then pairs the users with large channel gain difference. Sun Q et al, in the "On the iterative capability of MIMO NOMA Systems" (IEEE Wireless Communications Letters,2015,4 (4): 405-408) uses a fixed grouping algorithm to first rank the channel gains of the users and divide them into two categories. Taking the number of multiplexing users of two users as an example, a user with higher channel gain is divided into an A user set, a user with lower channel gain is divided into a B user set, then two users with the largest channel gain difference are paired, and a user with the second highest channel gain is paired with a user with the second lowest channel gain, and so on. Thus, the users are distributed with large power difference to facilitate correct demodulation at the receiving end. In addition, users with higher channel gain in the grouping can obtain more throughput by the users with lower channel gain but higher power division, so that the total throughput performance of the system is improved.
Through the above analysis, the problems and defects of the prior art are as follows:
(1) The current random user clustering method does not consider the interference problem among users in the whole implementation process, and the system performance is poor;
(2) The pre-grouping clustering method has the problems that the strong and weak user sets are not uniformly distributed or users at interval values are concentrated in the implementation process, so that the selected users in the user clusters cannot reach a certain channel quality interval, and the system is influenced to obtain higher system throughput;
(3) The fixed grouping algorithm is relatively easy to implement, but the algorithm is researched aiming at the power pairing condition that only two users exist in a cluster, and the advantages of a multi-antenna non-orthogonal multiple access system are not fully utilized, so that the improvement of the system throughput is not obvious enough.
The difficulty in solving the above problems and defects is: for a multi-antenna non-orthogonal multiple access system, high channel gain users are allocated as cluster heads of different user clusters, so that the capacity of the whole system is maximized. In addition, cluster members with larger channel similarity with cluster heads in the system can effectively eliminate inter-cluster interference, thereby improving the throughput of the system. The traditional user clustering algorithm only considers the difference of user channel gains or the size of channel similarity, and does not fully utilize channel state information, so that the overall performance of the system is not good enough. Therefore, in the design of the system user clustering scheme, how to find a new method, fully utilize the advantages of the multi-antenna non-orthogonal multiple access system, and comprehensively consider the channel gain of the users and the channel similarity among the users to improve the system performance is the difficulty of solving the problems and the defects.
The significance for solving the problems and the defects is as follows: in the multi-antenna non-orthogonal multiple access system, a plurality of users can share the same time and frequency band resource, and various performance requirements of the users are met through user clustering and power distribution. From the perspective of information theory, through power distribution, the non-orthogonal multiple access system can obtain remarkable improvement in throughput, fairness and spectrum efficiency. Compared with simple power control in an orthogonal multiple access system, reasonable user clustering and power distribution strategies are adopted in a non-orthogonal multiple access system to meet user requirements. In conclusion, the selection of the user clustering scheme directly affects the performance of the non-orthogonal multiple access system, the problems and the defects existing in the traditional user clustering algorithm are solved, a more reasonable user clustering scheme is provided, and the method has important significance for improving the overall performance of the system.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a user clustering method, a user clustering system, a storage medium, computer equipment and application.
The invention is realized in this way, a user clustering method, the user clustering method includes:
and performing initial user clustering according to the channel gain of the user sample set in the system and the channel similarity between users. The specific method comprises the following steps: and (3) taking the front users after the channel gains of all the users are arranged in a descending order as the central point users, and dividing the rest users into user clusters corresponding to different central point users according to the channel similarity maximization principle of the users and the central point users to perform initial user clustering. The aim of this is that the high channel gain users are distributed as cluster heads of different user clusters to maximize the throughput of the whole system, and cluster members with high channel similarity with the cluster heads can effectively eliminate inter-cluster interference, thereby improving the throughput of the system;
and calculating initial model parameters of the Gaussian mixture model according to the initial user clustering result. In the process of calculating the initial value of the model parameter, compared with randomly selected samples, the initial user clustering result is selected as the sample to initialize the model parameter, so that a good operation result is ensured;
and calculating the posterior probability generated by each mixed component of each sample and iteratively updating the parameters of the Gaussian mixture model. Fitting different user samples to enable the model to more effectively perform user clustering according to the characteristics of the users;
and judging whether the iteration times reach a set termination condition, if so, clustering the users according to the cluster marks, otherwise, returning to calculate the posterior probability of each sample generated by each mixed component and iteratively updating the parameters of the Gaussian mixture model. Setting the termination condition of the algorithm as the total iteration times of the algorithm or the function value increment of the likelihood function is smaller than a set threshold value, and ensuring the ideal operation result of the algorithm.
Further, the user clustering method comprises the following steps:
1. according to users in the system sample set D = { x = 1 ,x 2 ,...,x m Channel gain h = { h } 1 ,h 2 ,...,h m The channel similarity between users
Carrying out initial user clustering, wherein m is the total number of users in the system, and i, j belongs to {1, 2.., m };
2. initial model parameters of Gaussian mixture model according to initial user clustering result
Calculating;
3. calculating the posterior probability gamma generated by each sample from each mixed component pi And iteratively updating parameters of the Gaussian mixture model
All user samples in the system are divided into k user clusters, and superscript t represents iteration times;
4. judging the number of iterations to beIf not, the set termination condition is reached, and if so, the cluster mark C = { C is marked 1 ,C 2 ,...,C k And (4) clustering the users, and otherwise, returning to the step (3).
Further, in the step 1, initial user clustering is performed according to channel gains of user sample sets in the system and channel similarities among users, and the following is implemented:
(1) There are m users in the system, and the sample set of users D = { x = { [ x ] 1 ,x 2 ,...,x m H, each user sample x i Is gained by the user channel gain h i And the physical deviation angle alpha of the user and the base station i Constituent two-dimensional column vectors, i.e. x i ＝(h i ,α i ) Wherein i belongs to {1, 2.,. M }, assuming that all user samples in the system are divided into k user clusters, performing descending sorting on the channel gain values of all the user samples, and selecting k user samples before sorting as a central user sample;
(2) Computing user samples x i And center user sample x j Channel similarity of
And if the channel similarity between the user sample and a certain central user sample is maximum, dividing the user into user clusters corresponding to the central user.
Further, in the step 2, the initial model parameters of the gaussian mixture model are calculated according to the initial user clustering result, and the following is realized:
(1) The initial mean vector of the pth user cluster is defined as:
where p is in {1, 2.. K }, and j is in {1, 2.. K }, i.e., the central user sample x j As an initial mean vector
(2) Suppose the number of users in the p-th user cluster is m p Where p ∈ {1, 2.., k }, then the initial covariance matrix for the pth user cluster definesComprises the following steps:
wherein x is p,i The sample is the ith user sample in the p user cluster;
(3) The initial mixing coefficient for the p-th user cluster is defined as:
further, in the step 3, the posterior probability generated by each mixed component of each sample is calculated, and parameters of the gaussian mixture model are updated iteratively, which is implemented as follows:
(1) Defining a random variable z i E {1, 2.., k } represents the generation sample x i Is mixed with the components of (1), thereby z i Is a priori probability P (z) i = p) corresponds to α p ，z i The posterior probability distribution of (a) is expressed according to bayes' theorem as:
wherein, P (z) i ＝p|x i ) Represents the user sample x i The posterior probability generated by the p-th Gaussian mixture component is abbreviated as gamma pi Wherein p belongs to {1, 2.., k }, and i belongs to {1, 2.., m };
(2) Calculating and updating the mean vector:
(3) Calculating and updating the covariance matrix:
(4) Calculating and updating the mixing coefficient:
it is a further object of the invention to provide a computer device comprising a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to perform the steps of:
performing initial user clustering according to the channel gain of a user sample set in the system and the channel similarity between users;
calculating initial model parameters of the Gaussian mixture model according to an initial user clustering result;
calculating posterior probability of each sample generated by each mixed component and iteratively updating parameters of the Gaussian mixture model;
judging whether the iteration times reach the set termination condition, if so, marking according to the clusters
Sampling user x i Cut into the corresponding cluster:
finally completing the division of user clusters C = { C 1 ,C 2 ,...,C k And otherwise, returning to calculate the posterior probability generated by each mixed component of each sample and updating the parameters of the Gaussian mixture model in an iterative manner.
It is another object of the present invention to provide a computer-readable storage medium storing a computer program which, when executed by a processor, causes the processor to perform the steps of:
performing initial user clustering according to the channel gain of a user sample set in the system and the channel similarity between users;
calculating initial model parameters of the Gaussian mixture model according to an initial user clustering result;
calculating posterior probability of each sample generated by each mixed component and iteratively updating parameters of the Gaussian mixture model;
and judging whether the iteration times reach a set termination condition, if so, clustering the users according to the cluster marks, otherwise, returning to calculate the posterior probability of each sample generated by each mixed component and iteratively updating the parameters of the Gaussian mixture model.
Another object of the present invention is to provide a data processing terminal, which is used for implementing the user clustering method.
Another object of the present invention is to provide a user clustering system implementing the user clustering method, the user clustering system comprising:
the initial user clustering module is used for performing initial user clustering according to the channel gains of all users in the system and the channel similarity among the users;
the initial model parameter calculation module is used for calculating initial model parameters of the Gaussian mixture model according to an initial user clustering result;
the Gaussian mixture model parameter updating module is used for calculating the posterior probability of each sample generated by each mixture component and iteratively updating the parameters of the Gaussian mixture model;
and the termination condition judging module is used for judging whether the iteration times reach the set termination condition.
Another object of the present invention is to provide a multi-antenna non-orthogonal multiple access system, which operates the user clustering method.
By combining all the technical schemes, the invention has the advantages and positive effects that: in the user clustering process, the invention initializes the model parameters of the Gaussian mixture model by combining the channel gains of all users and the channel similarity among the users, can effectively improve the system throughput and is more suitable for an actual system. The invention is based on the Gaussian mixture model, so that the method can be used for fitting any user distribution condition, can be used for more effectively clustering the users according to the characteristics of the users, and has wider application range.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments of the present application will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained from the drawings without creative efforts.
FIG. 1 is a flowchart of a user clustering method according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a user clustering system according to an embodiment of the present invention;
in fig. 2: 1. an initial user clustering module; 2. an initial model parameter calculation module; 3. a Gaussian mixture model parameter updating module; 4. and a termination condition judgment module.
FIG. 3 is a diagram of a multi-antenna non-orthogonal multiple access system model used in accordance with an embodiment of the present invention;
fig. 4 is a flowchart of an implementation of a user clustering method according to an embodiment of the present invention;
fig. 5 is a comparison graph of simulated system throughput between the present invention and the conventional user clustering method according to the embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the invention.
In view of the problems in the prior art, the present invention provides a user clustering method, system, storage medium, computer device and application, and the present invention is described in detail below with reference to the accompanying drawings.
As shown in fig. 1, the user clustering method provided by the present invention includes the following steps:
s101: performing initial user clustering according to the channel gain of a user sample set in the system and the channel similarity between users;
s102: calculating initial model parameters of the Gaussian mixture model according to an initial user clustering result;
s103: calculating posterior probability of each sample generated by each mixed component and iteratively updating parameters of the Gaussian mixture model;
s104: and judging whether the iteration times reach a set termination condition, if so, clustering the users according to the cluster marks, otherwise, returning to the S103.
Those skilled in the art of the user clustering method provided by the present invention can also implement the user clustering method by adopting other steps, and the user clustering method provided by the present invention in fig. 1 is only a specific embodiment.
As shown in fig. 2, the user clustering system provided by the present invention includes:
an initial user clustering module 1, configured to perform initial user clustering according to channel gains of all users in the system and channel similarities among the users;
the initial model parameter calculation module 2 is used for calculating initial model parameters of the Gaussian mixture model according to an initial user clustering result;
the Gaussian mixture model parameter updating module 3 is used for calculating the posterior probability generated by each mixture component of each sample and iteratively updating the parameters of the Gaussian mixture model;
and the termination condition judging module 4 is used for judging whether the iteration times reach the set termination condition.
The technical solution of the present invention is further described below with reference to the accompanying drawings.
As shown in fig. 3, the multi-antenna non-orthogonal multiple access system to which the present invention is applied includes a base station equipped with multiple antennas and a plurality of single-antenna users. The base station end adopts the beam forming technology to form a plurality of beams, and each beam of the base station end serves one user cluster after the user clustering process is finished.
As shown in fig. 4, the user clustering method based on the gaussian mixture model according to the multi-antenna non-orthogonal multiple access system shown in fig. 3 of the present invention is implemented as follows:
firstly, performing initial user clustering according to user parameters;
(1.1) assuming there are m users in the system, the sample set of users D = { x = { x = 1 ,x 2 ,...,x m Every user sampleThis x i Is gained by the user channel gain h i And the physical deviation angle alpha of the user and the base station i A two-dimensional column vector of components, i.e. x i ＝(h i ,α i ) Wherein i belongs to {1, 2., m }, assuming that all user samples in the system are divided into k user clusters, performing descending sorting on the channel gain values of all user samples, and selecting the k user samples before sorting as center user samples;
(1.2) computing user samples x i And center user sample x j Channel similarity of
And if the channel similarity between the user sample and a certain central user sample is maximum, dividing the user into user clusters corresponding to the central user.
(1.3) completion of initial Cluster partitioning
Calculating initial model parameters;
(2.1) the initial mean vector of the p-th user cluster is defined as:
where p is in {1,2,. K }, and j is in {1,2,. K }, i.e., a sample x of the central user j As initial mean vector
(2.2) assume that the number of users in the p-th user cluster is m p Where p ∈ {1, 2., k }, then the initial covariance matrix for the pth user cluster is defined as:
wherein x is p,i The sample is the ith user sample in the p user cluster;
(2.3) of the p-th user clusterThe initial mixing coefficient is defined as:
calculating the posterior probability of each sample generated by each mixed component and iteratively updating the parameters of the Gaussian mixture model;
(3.1) defining a random variable z i E {1,2,. K } represents the generation sample x i Is mixed with the components of (1), thereby z i Is a priori probability P (z) i = p) corresponds to α p ，z i The posterior probability distribution of (a) is expressed as:
wherein, P (z) i ＝p|x i ) Represents the user sample x i The posterior probability generated by the p-th Gaussian mixture component is abbreviated as gamma pi Wherein p belongs to {1, 2.., k }, and i belongs to {1, 2.., m };
(3.2) calculating and updating the mean vector:
(3.3) calculating and updating covariance matrix:
(3.4) calculating and updating the mixing coefficient:
step four, judging whether the termination condition of the algorithm is met, and if the termination condition of the algorithm is met, performing step five; and if the terminal conditions of the algorithm are not met, repeating the step three.
Step five, according to each user sample x i Is marked with a subset
Sample x i Cut into the corresponding cluster:
finally completing the division of user clusters C = { C 1 ,C 2 ,...,C k }。
The technical effects of the present invention will be described in detail with reference to simulations.
1. Simulation conditions are as follows:
in the experiment, the transmitting power of a base station is set to be 1W-5.5W, the number of base station antennas is 10, the total number of users in the system is assumed to be 24, the path loss index of a millimeter wave channel is set to be 2, the transmission bandwidth is 1GHz, a fractional power distribution algorithm is adopted among users in the same cluster, the distribution coefficient is 0.8, and the noise power at the users is 3.98 multiplied by 10 -11 W is added. Setting the total iterative times of the algorithm to 25 times and the function value increment threshold value of the likelihood function to 10 -8 。
2. Simulation content and results:
under the above simulation conditions, the system throughput of the multi-antenna non-orthogonal multiple access system is simulated and compared respectively by using the method of the present invention and the conventional user clustering method, and the result is shown in fig. 5. In fig. 5, the abscissa is the transmission power of the base station in W, and the ordinate is the throughput of the system in Gbps. As can be seen from fig. 5, the system throughput value of the method of the present invention is significantly improved compared to the conventional user clustering method. It is worth noting that compared with the traditional user clustering method, the method of the present invention can fit any user distribution situation, and is more suitable for practical application.
Simulation results show that compared with a fixed grouping algorithm, the method can effectively improve the system throughput, in the traditional user clustering algorithm, the system throughput is lower when the user clustering is carried out by adopting a random clustering algorithm, and meanwhile, the system throughput adopting the random clustering algorithm has no obvious change along with the increase of the transmitting power of the base station. In addition, as can be seen from the simulation results, the difference between the system throughput obtained by adopting the four algorithms is obvious.
In addition, the invention analyzes the time complexity and the performance gain of the fixed grouping algorithm and the user clustering algorithm based on the Gaussian mixture model. As can be seen from the simulation result in fig. 5, when the transmission power of the base station is 3.5W, the scheme can further improve the system throughput by 30% compared with the fixed grouping algorithm. In terms of complexity, if a fixed grouping algorithm is adopted for user clustering for m users in the system, the time complexity of the algorithm is O (m) 2 ) Assuming that each user in the system has I characteristic parameters, the iteration times of the Gaussian mixture model-based user clustering algorithm is Q, the users in the system are divided into N clusters, and the time complexity of the Gaussian mixture model-based user clustering algorithm is O (m) 2 + (m + IQ) N). Therefore, aiming at the actual situation that the number m of the users of the multi-antenna non-orthogonal multiple access system is large, the scheme can obtain higher system performance gain by using smaller time complexity, and is more dominant in practical application.
It should be noted that the embodiments of the present invention can be realized by hardware, software, or a combination of software and hardware. The hardware portions may be implemented using dedicated logic; the software portions may be stored in a memory and executed by a suitable instruction execution system, such as a microprocessor or specially designed hardware. It will be appreciated by those skilled in the art that the apparatus and methods described above may be implemented using computer executable instructions and/or embodied in processor control code, for example such code provided on a carrier medium such as a diskette, CD-or DVD-ROM, a programmable memory such as read-only memory (firmware) or a data carrier such as an optical or electronic signal carrier. The apparatus of the present invention and its modules may be implemented by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, or software executed by various types of processors, or a combination of hardware circuits and software, e.g., firmware.
The above description is only for the purpose of illustrating the embodiments of the present invention, and the scope of the present invention should not be limited thereto, and any modifications, equivalents and improvements made by those skilled in the art within the technical scope of the present invention as disclosed in the present invention should be covered by the scope of the present invention.
Claims (8)
1. A user clustering method is characterized by comprising the following steps:
performing initial user clustering according to the channel gain of a user sample set in the system and the channel similarity between users;
calculating initial model parameters of the Gaussian mixture model according to an initial user clustering result;
calculating the posterior probability of each sample generated by each mixed component and iteratively updating the parameters of the Gaussian mixture model;
judging whether the iteration times reach a set termination condition, if so, clustering the users according to the cluster marks, otherwise, calculating the posterior probability generated by each mixed component of each sample and updating the parameters of the Gaussian mixture model in an iterative manner;
the user clustering method comprises the following steps:
(1) According to the in-system user sample set D = { x = { (x) 1 ,x 2 ,...,x m Channel gain h = { h } 1 ,h 2 ,...,h m Channel similarity between users and users
Carrying out initial user clustering, wherein m is the total number of users in the system, and i, j belongs to {1, 2.., m };
(2) Initial model parameters of Gaussian mixture model according to initial user clustering result
Calculating; mean vector
Covariance matrix
Coefficient of mixing
(3) Calculating the posterior probability gamma of each sample generated by each mixed component pi And iteratively updating parameters of the Gaussian mixture model
All user samples in the system are divided into k user clusters, and superscript t represents iteration times;
(4) Judging whether the iteration times reach the set termination condition, if so, marking C = { C according to the cluster 1 ,C 2 ,...,C k User clustering is carried out, and if not, the step (3) is returned;
in the step (2), the initial model parameters of the Gaussian mixture model are calculated according to the initial user clustering result, and the following steps are realized:
(a) The initial mean vector of the pth user cluster is defined as:
where p is in {1, 2.. K }, and j is in {1, 2.. K }, i.e., the central user sample x j As an initial mean vector
(b) Suppose the number of users in the p-th user cluster is m p Where p e {1,2,. K }, then the initial covariance matrix for the pth user cluster is defined as:
wherein x is p,i An ith user sample in the p user cluster;
(c) P-th user clusterThe initial mixing coefficient of (a) is defined as:
2. the user clustering method according to claim 1, wherein the initial user clustering in (1) is performed according to the channel gain of the user sample set in the system and the channel similarity between users, and is implemented as follows:
(1) There are m users in the system and, sample set D = { x ] for user 1 ,x 2 ,...,x m H, each user sample x i Is gained by the user channel gain h i And the physical deviation angle alpha of the user and the base station i Constituent two-dimensional column vectors, i.e. x i ＝(h i ,α i ) Wherein i belongs to {1, 2.,. M }, assuming that all user samples in the system are divided into k user clusters, performing descending sorting on the channel gain values of all the user samples, and selecting k user samples before sorting as a central user sample;
(2) Computing user samples x i And center user sample x j Channel similarity of
And if the channel similarity between the user sample and a certain central user sample is maximum, dividing the user into user clusters corresponding to the central user.
3. The user clustering method according to claim 1, wherein in (3), the posterior probability generated by each mixture component of each sample is calculated and the parameters of the Gaussian mixture model are iteratively updated by:
(1) Defining a random variable z i E {1, 2.., k } represents the generation sample x i Is mixed with the components of (1), thereby z i Is a priori probability P (z) i = p) corresponds to α p ，z i The posterior probability distribution of (a) is expressed as:
wherein, P (z) i ＝p|x i ) Represents the user sample x i The posterior probability generated by the p-th Gaussian mixture component is abbreviated as gamma pi Wherein p belongs to {1, 2.., k }, and i belongs to {1, 2.., m };
(2) Calculating and updating the mean vector:
(3) Calculating and updating the covariance matrix:
(4) Calculating and updating the mixing coefficient:
4. a computer arrangement, characterized in that the computer arrangement comprises a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to carry out the steps of the method as claimed in claim 1.
5. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, causes the processor to carry out the steps of the method as claimed in claim 1.
6. A data processing terminal, characterized in that the data processing terminal is configured to implement the user clustering method according to any one of claims 1 to 3.
7. A user clustering system for implementing the user clustering method according to any one of claims 1 to 3, wherein the user clustering system comprises:
the initial user clustering module is used for performing initial user clustering according to the channel gains of all users in the system and the channel similarity among the users;
the initial model parameter calculation module is used for calculating initial model parameters of the Gaussian mixture model according to an initial user clustering result;
the Gaussian mixture model parameter updating module is used for calculating the posterior probability generated by each mixture component of each sample and iteratively updating the parameters of the Gaussian mixture model;
and the termination condition judging module is used for judging whether the iteration times reach the set termination condition.
8. A multi-antenna non-orthogonal multiple access system, characterized in that it operates the user clustering method according to any of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011267021.1A CN112566214B (en) | 2020-11-13 | 2020-11-13 | User clustering method, system, storage medium, computer device and application |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011267021.1A CN112566214B (en) | 2020-11-13 | 2020-11-13 | User clustering method, system, storage medium, computer device and application |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112566214A CN112566214A (en) | 2021-03-26 |
| CN112566214B true CN112566214B (en) | 2022-10-18 |
Family
ID=75042988
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011267021.1A Active CN112566214B (en) | 2020-11-13 | 2020-11-13 | User clustering method, system, storage medium, computer device and application |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112566214B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8488578B1 (en) * | 2010-09-27 | 2013-07-16 | Rockwell Collins, Inc. | Identifying a CDMA scrambling code |
| CN108768567A (en) * | 2018-06-01 | 2018-11-06 | 北京邮电大学 | A kind of multipath cluster-dividing method, device, electronic equipment and readable storage medium storing program for executing |
| CN109922487A (en) * | 2019-03-28 | 2019-06-21 | 南京邮电大学 | A kind of resource allocation methods under downlink MIMO-NOMA network |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9530417B2 (en) * | 2013-01-04 | 2016-12-27 | Stmicroelectronics Asia Pacific Pte Ltd. | Methods, systems, and circuits for text independent speaker recognition with automatic learning features |
| TWI738317B (en) * | 2019-05-08 | 2021-09-01 | 國立臺灣大學 | An communication system and method |
-
2020
- 2020-11-13 CN CN202011267021.1A patent/CN112566214B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8488578B1 (en) * | 2010-09-27 | 2013-07-16 | Rockwell Collins, Inc. | Identifying a CDMA scrambling code |
| CN108768567A (en) * | 2018-06-01 | 2018-11-06 | 北京邮电大学 | A kind of multipath cluster-dividing method, device, electronic equipment and readable storage medium storing program for executing |
| CN109922487A (en) * | 2019-03-28 | 2019-06-21 | 南京邮电大学 | A kind of resource allocation methods under downlink MIMO-NOMA network |
Non-Patent Citations (4)
| Title |
|---|
| Clustering in the wireless channel with a power weighted statistical mixture model in indoor scenario;Yupeng Li等;《China Communications》;20190719;全文 * |
| 基于分布式EM算法的目标跟踪;李佳亮;《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》;20180215;全文 * |
| 超密集组网下一种基于干扰增量降低的分簇算法;梁彦霞等;《电子与信息学报》;20200215(第02期);全文 * |
| 超密集网络以用户为中心及多维协作的用户分簇算法;李皓等;《电视技术》;20180305(第03期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112566214A (en) | 2021-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110932764B (en) | User matching and power distribution method of MIMO-NOMA downlink communication system | |
| CN111698045B (en) | Energy efficiency power distribution method in millimeter wave communication system based on non-orthogonal multiple access | |
| CN109104225A (en) | A kind of optimal extensive MIMO Beam Domain multicast transmission method of efficiency | |
| CN111132264B (en) | User clustering method for downlink of multi-user MIMO-NOMA system | |
| CN113708804B (en) | Whale algorithm-based user scheduling and simulated beam selection optimization method | |
| Cui et al. | The application of machine learning in mmWave-NOMA systems | |
| US8767671B2 (en) | Space division multiple access transmission method of statistical characteristic mode | |
| US20200366341A1 (en) | Software-defined massive multi-input multi-output (mimo) | |
| US9344904B2 (en) | Interference alignment method and device in cellular network | |
| WO2017118099A1 (en) | Method and apparatus for allocating uplink pilot and jointly optimizing received beamforming vectors | |
| CN113329418B (en) | User clustering method, system and medium based on large-scale MIMO-NOMA system | |
| CN105554780A (en) | Massive MIMO multi-cell cooperative beam distribution method under millimeter wave | |
| CN110677858B (en) | Transmission power and computing resource allocation method based on task migration period of Internet of things | |
| US20220076134A1 (en) | Two-stage deep learning based secure precoder for information and artificial noise signal in non-orthogonal multiple access system | |
| CN112566214B (en) | User clustering method, system, storage medium, computer device and application | |
| CN110149133B (en) | Large-scale uplink transmission method based on beam space | |
| CN109951219B (en) | Low-cost large-scale non-orthogonal multi-access method | |
| CN107346985B (en) | Interference alignment method combined with transmitting antenna selection technology | |
| CN107154815B (en) | Multi-user system hybrid pre-coding method | |
| CN116056210A (en) | IRS auxiliary ultra-dense network resource allocation method for capacity coverage | |
| CN112243283B (en) | Cell-Free Massive MIMO network clustering calculation method based on successful transmission probability | |
| CN105429925B (en) | Multi-antenna OFDMA signal decoding method based on Fast Independent Component Analysis | |
| Sabaawi et al. | Unconstrained quantum genetic algorithm for massive MIMO system | |
| CN114710185B (en) | AP selection method for honeycomb-removed large-scale MIMO system | |
| Maruyama et al. | Pilot Allocation Optimization using Digital Annealer for Multi-cell Massive MIMO |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |