CN113194031A - User clustering method and system combining interference suppression in fog wireless access network - Google Patents
User clustering method and system combining interference suppression in fog wireless access network Download PDFInfo
- Publication number
- CN113194031A CN113194031A CN202110444218.6A CN202110444218A CN113194031A CN 113194031 A CN113194031 A CN 113194031A CN 202110444218 A CN202110444218 A CN 202110444218A CN 113194031 A CN113194031 A CN 113194031A
- Authority
- CN
- China
- Prior art keywords
- user
- clustering
- cluster
- access network
- data space
- 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.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/46—Cluster building
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2137—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on criteria of topology preservation, e.g. multidimensional scaling or self-organising maps
- G06F18/21375—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on criteria of topology preservation, e.g. multidimensional scaling or self-organising maps involving differential geometry, e.g. embedding of pattern manifold
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23211—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with adaptive number of clusters
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y10/00—Economic sectors
- G16Y10/75—Information technology; Communication
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y20/00—Information sensed or collected by the things
- G16Y20/10—Information sensed or collected by the things relating to the environment, e.g. temperature; relating to location
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y40/00—IoT characterised by the purpose of the information processing
- G16Y40/20—Analytics; Diagnosis
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W16/00—Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
- H04W16/14—Spectrum sharing arrangements between different networks
-
- 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
- H04W40/32—Connectivity information management, e.g. connectivity discovery or connectivity update for defining a routing cluster membership
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W72/00—Local resource management
- H04W72/04—Wireless resource allocation
- H04W72/044—Wireless resource allocation based on the type of the allocated resource
- H04W72/0453—Resources in frequency domain, e.g. a carrier in FDMA
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W72/00—Local resource management
- H04W72/50—Allocation or scheduling criteria for wireless resources
- H04W72/54—Allocation or scheduling criteria for wireless resources based on quality criteria
- H04W72/541—Allocation or scheduling criteria for wireless resources based on quality criteria using the level of interference
-
- 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)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Environmental & Geological Engineering (AREA)
- Probability & Statistics with Applications (AREA)
- Toxicology (AREA)
- Quality & Reliability (AREA)
- Biomedical Technology (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a user clustering method and a system combining interference suppression in a fog wireless access network, wherein the method comprises the following steps: s1, obtaining the user position coordinate and the service type information; s2, carrying out normalization and weighted combination processing; s3, performing Laplace feature mapping on the processed data space; s4, randomly initializing a centroid coordinate corresponding to the cluster number k being 2; performing clustering matching according to a minimum distance principle to obtain a user cluster; calculating the average value of each dimension of the data of each user cluster and updating the original clustering center as the calculated centroid coordinate; s5, judging whether the distance between the devices in each user cluster exceeds the D2D communication threshold, if so, determining that the cluster number k is k +1 and executing the step S4, otherwise, skipping to execute the step S6; and S6, judging whether the iterated clustering center is not changed any more, if so, outputting a clustering result, otherwise, executing the step S4. The invention utilizes the characteristics of the F-RAN to carry out interference suppression in the user clustering process.
Description
Technical Field
The invention belongs to the field of mobile communication access network user clustering, relates to the fields of a 5G fog wireless access network, access network user clustering and machine learning, and particularly relates to a user clustering method and a system combining interference suppression in the fog wireless access network.
Background
Due to the exponential increase of the number of users accessing the mobile communication network in the 5G era, the co-channel interference between the devices and the scarcity of the frequency spectrum resources become the main brake of the user service quality. In this context, Fog Radio Access Network (F-RAN) is attracting much attention as one of the 5G Access Network solutions. Due to the abundant fog computing nodes and the distributed signal processing and resource management capabilities, the F-RAN framework can effectively reduce network delay and improve user service quality.
In addition, the contradiction between the increasing number of mobile devices and the scarce spectrum resources leads to the fact that in future mobile communication networks, radio resources must be shared among a plurality of devices, and a Non-orthogonal Multiple Access (NOMA) technology is an effective way to improve the utilization rate of the spectrum resources. The Device-to-Device (D2D) communication mode within the F-RAN can increase the multiplexing gain within the cell in the manner described above, which is significant for improving spectrum utilization. Because the co-channel interference problem caused by spectrum reuse greatly limits the performance of the network, a great deal of research is carried out on the interference suppression problem in the network. However, in the existing schemes, the distance between the devices is reduced and the number of clusters is increased in a 5G mass access scene by performing beam forming and power control on the user equipment for interference suppression, and the interference suppression scheme at the device level in the traditional method is difficult to guarantee the service quality of the clustered user equipment in the research scene; in addition, most of the current research on the user clustering method is a prototype clustering scheme based on information such as the physical location of the equipment, and the interference information between the equipment is ignored. The above problem can be improved by using the excellent characteristics in the 5G mist wireless access network, so that the interference information among users is considered in the clustering process.
In summary, a new user clustering method and system combining interference suppression in a 5G wlan is needed.
Disclosure of Invention
The invention aims to provide a user clustering method and a user clustering system combining interference suppression in a fog wireless access network, which utilize the characteristics of an F-RAN to perform interference suppression in the user clustering process, have important effects on solving the problems in the prior art and have innovative significance.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention discloses a user clustering method combining interference suppression in a fog wireless access network, which comprises the following steps:
s1, obtaining the user position coordinate and the service type information;
s2, carrying out normalization and weighted combination processing on the user position coordinates and the service types to obtain a processed data space;
s3, performing Laplace feature mapping on the processed data space, and mapping the data space to a feature data space with inter-device interference information;
s4, based on the characteristic data space, randomly initializing a centroid coordinate corresponding to the cluster number k being 2; calculating the distance between the user and each clustering center, and performing clustering matching according to a minimum distance principle to obtain a user cluster; calculating the average value of each dimension of the data of each user cluster and updating the original clustering center as the calculated centroid coordinate;
s5, judging whether the distance between the devices in each user cluster exceeds the D2D communication threshold, if so, returning to execute the step S4 when the cluster number k is k +1, otherwise, jumping to execute the step S6;
and S6, judging whether the iterated clustering center is not changed any more based on the preset convergence condition, finishing user clustering and outputting clustering results if the iterated clustering center is not changed any more, and returning to the step S4 if the iterated clustering center is not changed any more.
A further improvement of the present invention is that step S1 specifically includes:
and the fog wireless access network controller acquires the user position coordinates and the service type information through the fog access node.
A further improvement of the present invention is that step S2 specifically includes:
carrying out MinMax normalization and weighted combination processing on the position coordinates of the user and the service types; wherein the weighted combination processing comprises: multiplying the user position coordinate and the service type by a position weight omega respectively1Type weight omega2Thereafter, clustering is started by changing ω1,ω2The value of (d) changes the degree of influence of the traffic class on the distance.
A further development of the invention consists in that, in step S2, by changing ω1,ω2The step of altering the degree of influence of the traffic class on the distance comprises:
ω1=1,ω2when the ratio is 0: based only on user location coordinates;
ω1=ω20.5: equally considering the position coordinates of the user and the service type;
ω1=0.2,ω20.8: emphasis is placed on the traffic class.
A further improvement of the present invention is that step S3 specifically includes:
in the feature mapping process, the inter-cluster interference received by the user is represented as:
in the formula, Cl、MnRespectively corresponding column vectors of a user-clustering indication array and a user-physical resource block indication array; w is an adjacency matrix constructed according to a kernel function, and the weight W of the corresponding positionijRepresenting co-channel interference between users at corresponding positions; l denotes a user cluster index and p denotesTotal number of clusters, if user i belongs to cluster 1, c il1 is ═ 1; n denotes the index of the PRB to which the user is attached, k denotes the total number of PRBs, if the user i is attached to the PRBnThen m isin1 is ═ 1; all user equipment which multiplexes the same PRB and belongs to different clusters are indicated in brackets;
the kernel function is represented as:
κ(xi,xj)=K(||xi-xj||)-α
where K ═ P ξ > 0, is a parameter determined by the F-UE transmit power and the path loss; α is a constant coefficient parameter based on antenna characteristics and average path loss; x is the number ofi,xjAre two input variables of the kernel function;
calculating the degree of each node and forming a degree matrix D ═ diag (D)1,…,dn) The computational expression of the degree of a node is,
wherein j represents the node index, N represents the total number of nodes, diRepresenting the sum of the weights corresponding to the node i and all the connected nodes;
laplace matrix L ═ D-W, according to the formula L ═ D-1/2LD-1/2And carrying out standardization processing on the Laplace matrix to complete the feature mapping of the user.
The invention relates to a user clustering system combining interference suppression in a fog wireless access network, which comprises the following steps:
the information acquisition module is used for acquiring the position coordinates of the user and the service type information;
the preprocessing module is used for carrying out normalization and weighted combination processing on the position coordinates of the user and the service types to obtain a processed data space;
the mapping module is used for carrying out Laplace feature mapping on the processed data space and mapping the data space to a feature data space with inter-device interference information;
the iteration updating module is used for initializing the centroid coordinate corresponding to the clustering number k which is 2 at random according to the characteristic data space; calculating the distance between the user and each clustering center, and performing clustering matching according to a minimum distance principle to obtain a user cluster; calculating the average value of each dimension of the data of each user cluster and updating the original clustering center as the calculated centroid coordinate;
the distance judging module is used for judging whether the distance between the devices in each user cluster exceeds a D2D communication threshold, if so, the cluster number k is k +1 and returns to the iteration updating execution module, and otherwise, the distance judging module jumps to the execution judgment output module;
and the judgment output module is used for judging whether the clustering center after iteration is not changed any more according to the preset convergence condition, finishing user clustering and outputting a clustering result if the clustering center after iteration is not changed any more, and returning to the iteration updating execution module if the clustering center after iteration is not changed any more.
The further improvement of the present invention is that, in the information obtaining module, the step of obtaining the user position coordinate and the service type information specifically includes:
and the fog wireless access network controller acquires the user position coordinates and the service type information through the fog access node.
The further improvement of the present invention is that, in the preprocessing module, the step of performing normalization and weighted combination processing on the user position coordinates and the service types to obtain the processed data space specifically includes:
carrying out MinMax normalization and weighted combination processing on the position coordinates of the user and the service types; wherein the weighted combination processing comprises: multiplying the user position coordinate and the service type by a position weight omega respectively1Type weight omega2Thereafter, clustering is started by changing ω1,ω2The value of (d) changes the degree of influence of the traffic class on the distance.
In a further development of the invention, the preprocessing module is designed to process the data by changing ω1,ω2The step of altering the degree of influence of the traffic class on the distance comprises:
ω1=1,ω2when the ratio is 0: based only on user location coordinates;
ω1=ω20.5: equally considering the position coordinates of the user and the service type;
ω1=0.2,ω20.8: emphasis is placed on the traffic class.
A further improvement of the present invention is that, in the mapping module, the step of performing laplacian eigenmapping on the processed data space and mapping to an eigen data space having inter-device interference information specifically includes:
in the feature mapping process, the inter-cluster interference received by the user is represented as:
in the formula, Cl、MnRespectively corresponding column vectors of a user-clustering indication array and a user-physical resource block indication array; w is an adjacency matrix constructed from a kernel function, and the weight W of the corresponding positionijRepresenting co-channel interference between users at corresponding positions; l represents the index of the user cluster, p represents the total number of the user clusters, and c if the user i belongs to the cluster l il1 is ═ 1; n denotes the index of the PRB to which the user is attached, k denotes the total number of PRBs, if the user i is attached to the PRBnThen m isin1 is ═ 1; all user equipment which multiplexes the same PRB and belongs to different clusters are indicated in brackets;
the kernel function is represented as:
κ(xi,xj)=K(||xi-xj||)-α
where K ═ P ξ > 0, is a parameter determined by the F-UE transmit power and the path loss; α is a constant coefficient parameter based on antenna characteristics and average path loss; x is the number ofi,xjAre two input variables of the kernel function;
calculating the degree of each node and forming a degree matrix D ═ diag (D)1,…,dn) The computational expression of the degree of a node is,
wherein j represents the node index, N represents the total number of nodes, diRepresenting the sum of the weights corresponding to the node i and all the connected nodes;
laplace matrix L ═ D-W, according to the formula L ═ D-1/2LD-1/2And carrying out standardization processing on the Laplace matrix to complete the feature mapping of the user.
Compared with the prior art, the invention has the following beneficial effects:
aiming at the same frequency interference among devices in an F-RAN and the scarcity of frequency spectrum resources, the invention provides a self-adaptive user clustering method in a fog wireless access network architecture, which has the following three advantages:
(1) the data preprocessing method of normalization and weight combination enables the service types of the devices in the cluster to be as close as possible, reduces the caching requirement on the cluster head devices, reduces the network cost and improves the overall performance of the network;
(2) compared with the traditional k-means and spectral clustering based on RBF, the method can effectively reduce the interference between the D2D communication user clusters.
(3) The method of the invention does not need to set the hyper-parameters, can self-adaptively complete the clustering task and is convenient for automatic deployment at the access node.
In the invention, the self-adaptive adjustment of the cluster type is realized based on the limitation of the communication transmitting power between the devices, and the cluster is updated when the cluster radius is larger than the limitation of the intra-cluster communication distance, thereby avoiding the operations such as manual configuration of the cluster type and the like, and having important significance for the 5G access network sensitive to time delay.
In the invention, the user service type information is considered in the clustering process through a data preprocessing method, and the service types of the same cluster equipment are made to be as close as possible through carrying out weighted combination operation after information such as physical positions, service types and the like is normalized, so that the aim of reducing the cluster head caching requirement is fulfilled.
The invention provides a kernel function based on path loss between user equipment, and the validity of the kernel function is verified by proving the limited semipositive nature of a corresponding kernel matrix. The kernel function provided by the invention maps the sample points to the feature space with interference information, and the inter-cluster interference can be reduced by clustering in the feature space.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art are briefly introduced below; it is obvious that the drawings in the following description are some embodiments of the invention, and that for a person skilled in the art, other drawings can be derived from them without inventive effort.
Fig. 1 is a schematic flow chart of a user clustering method with interference suppression in a 5G mist wireless access network according to an embodiment of the present invention;
FIG. 2 is a block diagram of an architecture and interference diagram according to an embodiment of the present invention; wherein, fig. 2 (a) is a schematic diagram of an architecture of an adopted misty radio access network, and fig. 2 (b) is a schematic diagram of intra-cluster interference and inter-cluster interference in a network;
FIG. 3 is a schematic diagram illustrating a user clustering comparison between a method according to an embodiment of the present invention and a conventional method; wherein, fig. 3 (a) is a schematic diagram of user equipment distribution in a simulated cell, fig. 3 (b) is a schematic diagram of traditional k-means user clustering, fig. 3 (c) is a schematic diagram of user clustering after RBF kernel feature mapping, and fig. 3 (d) is a schematic diagram of user clustering after feature mapping by using the kernel function in the present invention;
FIG. 4 is a diagram illustrating a comparison between the performance of a clustering algorithm according to an embodiment of the present invention and that of a conventional algorithm; the algorithm l is a traditional k-means algorithm, the algorithm 2 is an AP-k-means algorithm, the algorithm 3 is an AP-ISC algorithm, and the algorithm 4 is a spectral clustering algorithm adopting RBF kernel; fig. 4 (a) is a comparison diagram of average channel capacity, and fig. 4 (b) is a diagram of average co-channel interference.
Detailed Description
In order to make the purpose, technical effect and technical solution of the embodiments of the present invention clearer, the following clearly and completely describes the technical solution of the embodiments of the present invention with reference to the drawings in the embodiments of the present invention; it is to be understood that the described embodiments are only some of the embodiments of the present invention. Other embodiments, which can be derived by one of ordinary skill in the art from the disclosed embodiments without inventive faculty, are intended to be within the scope of the invention.
Referring to fig. 2, the user cluster in D2D communication mode includes: in a cell with a macro base station (i.e., a High Power Node, HPN, High Power Node) as a center and R as a radius, defining an access point set AP (FAP) formed by m fog access nodes F-APs in a current state1,FAP2,…FAPm) N fog computing user equipments forming a user set F-UEs ═ x1,x2,…xn) And p idle physical resource blocks constitute a resource pool RP ═ (PRB)1,PRB2,…PRBp). All the F-UE sharing RP in the communication process stipulates that the same physical resource block PRB is used for communication in the same user cluster.
Therefore, the communication between the F-UEs will be subject to co-channel interference from the rest of the devices in the cell, as shown in fig. 2. The content of the research of the embodiment of the invention is how to divide the F-UEs into k user clustersAnd reducing inter-cluster interference between users so that the F-UEs can obtain a larger downlink communication rate.
When accessing the network, the equipment selects the access point according to the shortest distance principle. Defining a binary indicator variable aijFor indicating a user-access point matching situation, as shown in equation (1). From a to aijForming a user-access point matching matrix An×m。
After clustering, the F-UEs are partitioned into k disjoint user clustersSatisfy between each cluster And isDefining a binary indicator variable bijFor indicating the user clustering situation, as shown in equation (2). B is formed byijForming a user-cluster matching matrix Bn×k。
And after clustering, replacing other devices in the cluster to access the F-AP according to the cluster head of each user cluster. Each F-AP randomly selects one PRB in an available resource pool RP for all user clusters in the node for intra-cluster communication. Defining a binary variable cijThe method is used for indicating the matching condition of the user cluster and the PRB, and is shown in a formula (3). From cijForming user cluster-PRB matching matrix Ck×p。
The PRB multiplexing information matrix M can be obtained according to the definitionn×p=Bn×k×Ck×pEach column M of MjThe middle value of 1 represents that the corresponding user equipment multiplexes PRBjSuch a setting facilitates calculation of co-channel interference between user equipments.
In the embodiment of the invention, the used channel model is a path loss model combining Rayleigh fading and based on distance, and the user xjAnd user xiGain g of the path therebetweeni,iAs shown in formula (4):
in the formula: xi is a Rayleigh fading parameter; dijRepresenting a user xjAnd user xiThe distance between them; α is a constant coefficient based on the antenna characteristics and the average path loss. According to the Shannon formula, user xjAnd xiThe channel capacity at the time of communication can be expressed by equation (5):
Ci,j=B×log2(1+γij) (5)
wherein B is PRB corresponding bandwidth, gammaijRepresenting a user xjAnd xiThe signal to interference plus noise ratio (snr) therebetween can be obtained by equation (6):
in the formula: piRepresenting a user xjTransmission power of gijRepresenting a user xjAnd xiChannel gain between, the numerator represents user x as a wholejThe received useful signal power; first term representation of denominator and user xiThe power of the co-channel interference experienced, and the second term represents the noise power.
Referring to fig. 1, a user clustering method (AP-ISC algorithm) for combining interference suppression in a 5G misty wireless access network according to an embodiment of the present invention specifically includes the following steps:
1. the fog wireless access network controller acquires information such as user positions and service types in the cell through each fog access node;
2. carrying out MinMax normalization processing on the user coordinates and the service types, and weighting and combining the user coordinates and the service types into a new data space;
3. according to the kernel function provided by the embodiment of the invention, the normalized data is mapped to a characteristic data space with the interference information between the devices;
4. randomly initializing a centroid coordinate corresponding to the clustering number k being 2;
5. calculating the distance between the user and each clustering center, and performing clustering matching according to the minimum distance principle;
6. calculating the average value of each dimension of the data of each user cluster as a new centroid coordinate, and updating the original clustering center;
7. judging whether the distance between the devices in each cluster exceeds a D2D communication threshold, if so, judging that the cluster number k is k +1 and returning to the step 4, otherwise, executing the next step;
8. and judging whether the iterative clustering center is not changed any more, finishing user clustering and outputting a clustering result if the iterative clustering center is not changed any more, and returning to the step 4 if the iterative clustering center is not changed.
In the embodiment of the invention, in order to compare the effectiveness of the interference suppression strategy in the invention, the process is decomposed into two algorithms for simulation comparison. The adaptive user clustering part is defined as an AP-k-means algorithm, the specific flow is shown in table 1, the clustering algorithm combined with interference suppression is defined as an AP-ISC algorithm, and the specific flow is shown in table 2. The comparative results are shown in FIG. 4.
TABLE 1 AP-k-means Algorithm
TABLE 2 AP-ISC Algorithm
In the embodiment of the invention, the user position and the service type are respectively multiplied by the weight omega in the step 2 data preprocessing process1,ω2Thereafter, the clustering process is started by changing omega1,ω2The value of (c) may alter the degree of influence of the traffic class on the distance, for example:
ω1=1,ω2when the ratio is 0: based only on the UE location;
ω1=ω20.5: equally considering the UE location and the traffic type;
ω1=0.2,ω20.8: emphasis is placed on UE traffic classes.
In the embodiment of the invention, the user coordinates and the service type set are combined into a new space coordinate, and MinMax normalization processing is carried out on all dimensions according to the formula (7).
Randomly initializing K clustering center coordinates;
and (3) calculating the distances between the users and the K clustering centers, and defining the distance between the users as an equation (8).
The user is divided into clusters with the smallest distance, and the cluster center is updated.
In the embodiment of the present invention, in the interference suppression process in step 3, in order to consider interference information between users in the clustering process, Laplacian Eigenmaps (LE) are performed on data in the present invention. The basic idea of LE is to map the raw data set into the feature space and preserve the structural relationship of the raw data, i.e. if two data points close in the raw space should also be close in the feature space, while additional information features between the data points can also be taken into account in the mapping process. Defining weight value ω in the present inventionijRepresenting a node niAnd njThe weight value of the side between, and has wij=wji. All the weight values constitute a weight matrix W in the graph. In the feature mapping process, W reflects the similarity between data, the similarity can be represented by the distance between two points in the prototype clustering, and the weighted value of the two points with the closer distance is larger. A commonly used method for obtaining a fully-connected adjacency matrix W is to adopt a Gaussian Kernel function (Gaussian Kernel) to obtain a weight between two points, and a user xiAnd xjWeight w betweenijCan be represented by the formula (9).
In the formula, σ2Is a parameter of a gaussian kernel function. In order to consider interference information between user equipment in a feature space, a kernel function is designed in the research of the invention as shown in formula (10):
κ(xi,xj)=K(||xi-xj||)-α (10)
where K ═ pxi > 0, is a parameter determined by the F-UE transmit power and path loss; α is a constant coefficient parameter based on antenna characteristics and average path loss; x is the number ofi,xjTwo input variables (nodes) of the kernel function, namely two preprocessed user data points, are obtained through a formula (10) to obtain weight values omega corresponding to the two user data pointsij。
In the embodiment of the present invention, the effective kernel function in the sample space is proved as follows.
And (3) proving that: by definition, any two points x in the D-dimensional sample space(i)And x(j)With κ (x)(i),x(j))=K(||xi-xj||)-α
Note the bookThen the Kernel Matrix (Kernel Matrix) corresponding to equation 8 can be expressed as follows:
at any point y in the sample space, there is:
according to the equation (11), the kernel matrix corresponding to the kernel function has semi-positive characteristics in the sample space, so that it is an effective kernel function.
According to the aforementioned configuration, the inter-cluster interference received by the user can be expressed as:
wherein C isl、MnColumn vectors respectively corresponding to the user-clustering indicator array and the user-PRB indicator array, wherein W is an adjacent matrix formed by the kernel function shown in the formula 8, and the weight W of the corresponding positionijIndicating co-channel interference between users at corresponding locations. l represents the index of the user cluster, p represents the total number of user clusters in the cell, and c if the user i belongs to the cluster 1il1 is ═ 1; n represents the index of PRB to which the user is attached, k represents the total number of PRBs that can be allocated in the cell, and if the user i is attached to the PRBnThen m isin1. And all user equipment multiplexing the same PRB and belonging to different clusters are indicated in brackets, and W indicates the co-channel interference value among the equipment.
The degree of each node is then calculated according to equation (15) and the degree matrix D ═ diag (D) is constructed1,…,dn)。
In the formula (15), j represents the node index, N represents the total number of nodes, and diRepresenting the sum of the weights corresponding to node i and all its connected nodes.
Subtracting the degree matrix from the adjacency matrix to obtain a Laplace matrix L ═ D-W, and calculating the L ═ D-1/2LD-1/2And carrying out standardization processing on the Laplace matrix, thereby completing the feature mapping of the user. The sample space after feature mapping takes the degree of interference between users into consideration, and the larger the interference between devices is, the larger the corresponding weight is. Usually, the spectral clustering algorithm performs dimensionality reduction after feature mapping, namely, the spectral clustering algorithm performs dimensionality reduction on the featuresThe characteristic space is subjected to characteristic value decomposition, and m maximum characteristic values { lambda ] are selected from the characteristic space1,…,λmThe corresponding feature vector { xi }1,…,ξmCompose a new sample space and perform prototype clustering on it to obtain a result classified into clusters. In the invention, the sample space only has 3 dimensions of abscissa, ordinate and slice type, and the dimension reduction is not carried out on the data because some information in the data is lost in consideration of the dimension reduction.
In the embodiment of the invention, the average channel capacity and the inter-cluster interference of the clustering algorithm are simulated by a computer, specific parameters are shown in a table 3, and a simulation result is shown in a figure 4, wherein the algorithm 1 is a traditional k-means algorithm, the algorithm 2 is an AP-k-means algorithm, the algorithm 3 is an AP-ISC algorithm, and the algorithm 4 is a spectral clustering algorithm adopting RBF kernel. It can be obtained from the simulation result of fig. 4 that the algorithms 1, 2, 4 are all clustering in the sample space, the performance performances are basically the same, the average inter-cluster interference and the average channel capacity increase approximately linearly with the increase of the available resource blocks, and the improvement degree is lower; algorithm 3 is a clustering scheme in the sample space, and the service quality of the users increases approximately logarithmically with the increase of the number of available PRBs, which is a great improvement over other algorithms. Because the resources of the network are limited, the number of idle resource blocks allocated to D2D communication is usually small, and the AP-ISC algorithm proposed in this section can effectively reduce inter-cluster interference of users when the number of PRBs is small, thereby improving the service quality of users.
TABLE 3 simulation parameter settings
To sum up, the embodiment of the invention relates to a user clustering algorithm combined with interference suppression under a 5G mobile edge computing architecture. Aiming at the problems of radio resource scarcity and same frequency interference superposition under a 5G massive access scene, the invention firstly provides a self-adaptive clustering algorithm combined with user service classes, then provides a kernel function based on inter-equipment interference information and proves the effectiveness of the kernel function, and on the basis, the clustering algorithm is used for carrying out feature mapping on access users in a cell. And finally, applying the clustering algorithm provided by the invention to a user access mode selection process in the F-RAN. Computer simulation shows that the algorithm can reduce inter-cluster interference, improve the average channel capacity of cell users and save the burden of a forward network on the premise of ensuring the service quality of users. The invention relates to mobile communication network access user clustering, interference suppression and wireless communication.
In the invention, the user service type information is considered in the clustering process by designing a data preprocessing scheme, and the service types of the same cluster equipment are made to be as close as possible by carrying out weighted combination operation after the information such as the physical position, the service type and the like is normalized so as to realize the purpose of reducing the cluster head caching requirement. Then, a kernel function based on path loss between user equipment is designed, and the validity of the kernel function is verified by proving the limited semi-positive nature of a corresponding kernel matrix. The kernel function provided by the invention maps the sample points to the feature space with interference information, and the inter-cluster interference can be reduced by clustering in the feature space. And finally, realizing self-adaptive adjustment of the cluster type based on the limitation of the communication transmitting power between the devices, updating the cluster when the cluster radius is larger than the limitation of the intra-cluster communication distance, avoiding the operations of manually configuring the cluster type and the like, and having important significance for the 5G access network sensitive to time delay. In summary, aiming at the co-frequency interference among the devices in the F-RAN and the scarcity of frequency spectrum resources, the invention designs a self-adaptive user clustering algorithm in a fog wireless access network architecture, and the algorithm has the following three advantages: firstly, the service types of the devices in the cluster are made to be as close as possible by designing a data preprocessing method, so that the cache requirement on the cluster head device is reduced, the network cost is reduced, and the overall performance of the network is improved; second, compared with the traditional k-means and RBF-based spectral clustering, the algorithm can effectively reduce the interference between the D2D communication user clusters. Thirdly, the algorithm does not need to set the hyper-parameters, and can self-adaptively complete the clustering task, thereby facilitating the automatic deployment at the access node.
As will be appreciated by one skilled in the art, 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, disk storage, CD-ROM, 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 flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program 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 computer program instructions may also be stored in a computer-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 computer-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 computer program 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.
Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art can make modifications and equivalents to the embodiments of the present invention without departing from the spirit and scope of the present invention, which is set forth in the claims of the present application.
Claims (10)
1.A user clustering method combining interference suppression in a fog wireless access network is characterized by comprising the following steps:
s1, obtaining the user position coordinate and the service type information;
s2, carrying out normalization and weighted combination processing on the user position coordinates and the service types to obtain a processed data space;
s3, performing Laplace feature mapping on the processed data space, and mapping the data space to a feature data space with inter-device interference information;
s4, based on the characteristic data space, randomly initializing a centroid coordinate corresponding to the cluster number k being 2; calculating the distance between the user and each clustering center, and performing clustering matching according to a minimum distance principle to obtain a user cluster; calculating the average value of each dimension of the data of each user cluster and updating the original clustering center as the calculated centroid coordinate;
s5, judging whether the distance between the devices in each user cluster exceeds the D2D communication threshold, if so, returning to execute the step S4 when the cluster number k is k +1, otherwise, jumping to execute the step S6;
and S6, judging whether the iterated clustering center is not changed any more based on the preset convergence condition, finishing user clustering and outputting clustering results if the iterated clustering center is not changed any more, and returning to the step S4 if the iterated clustering center is not changed any more.
2. The method for clustering users by combining interference suppression in a mist wireless access network according to claim 1, wherein the step S1 specifically comprises:
and the fog wireless access network controller acquires the user position coordinates and the service type information through the fog access node.
3. The method for clustering users by combining interference suppression in a mist wireless access network according to claim 1, wherein the step S2 specifically comprises:
carrying out MinMax normalization and weighted combination processing on the position coordinates of the user and the service types; wherein the weighted combination processing comprises: multiplying the user position coordinate and the service type by a position weight omega respectively1Type weight omega2After which clustering is started.
4. The method for clustering users by combining interference suppression in a mist wireless access network according to claim 3, wherein in step S2, ω is a user position coordinate1=1,ω20; when the user position coordinate and the service type are considered equally, omega1=ω20.5; with emphasis on traffic class, ω1=0.2,ω2=0.8。
5. The method for clustering users by combining interference suppression in a mist wireless access network according to claim 1, wherein the step S3 specifically comprises:
in the feature mapping process, the inter-cluster interference received by the user is represented as:
in the formula, Cl、MnRespectively corresponding column vectors of a user-clustering indication array and a user-physical resource block indication array; w is an adjacency matrix constructed from a kernel function, and the weight W of the corresponding positionijRepresenting co-channel interference between users at corresponding positions; l represents the index of the user cluster, p represents the total number of the user clusters, and c if the user i belongs to the cluster lil1 is ═ 1; n denotes the index of the PRB to which the user is attached, k denotes the total number of PRBs, if the user i is attached to the PRBnThen m isin1 is ═ 1; in brackets, all user equipments multiplexing the same PRB and belonging to different clusters are indicated;
The kernel function is represented as:
κ(xi,xj)=K(||xi-xj||)-α
where K ═ P ξ > 0, is a parameter determined by the F-UE transmit power and the path loss; α is a constant coefficient parameter based on antenna characteristics and average path loss; x is the number ofi,xjAre two input variables of the kernel function;
calculating the degree of each node and forming a degree matrix D ═ diag (D)1,…,dn) The computational expression of the degree of a node is,
wherein j represents the node index, N represents the total number of nodes, diRepresenting the sum of the weights corresponding to the node i and all the connected nodes;
laplace matrix L ═ D-W, according to the formula L ═ D-1/2LD-1/2And carrying out standardization processing on the Laplace matrix to complete the feature mapping of the user.
6. A user clustering system incorporating interference mitigation in a fog wireless access network, comprising:
the information acquisition module is used for acquiring the position coordinates of the user and the service type information;
the preprocessing module is used for carrying out normalization and weighted combination processing on the position coordinates of the user and the service types to obtain a processed data space;
the mapping module is used for carrying out Laplace feature mapping on the processed data space and mapping the data space to a feature data space with inter-device interference information;
the iteration updating module is used for initializing the centroid coordinate corresponding to the clustering number k which is 2 at random according to the characteristic data space; calculating the distance between the user and each clustering center, and performing clustering matching according to a minimum distance principle to obtain a user cluster; calculating the average value of each dimension of the data of each user cluster and updating the original clustering center as the calculated centroid coordinate;
the distance judging module is used for judging whether the distance between the devices in each user cluster exceeds a D2D communication threshold, if so, the cluster number k is k +1 and returns to the iteration updating execution module, and otherwise, the distance judging module jumps to the execution judgment output module;
and the judgment output module is used for judging whether the clustering center after iteration is not changed any more according to the preset convergence condition, finishing user clustering and outputting a clustering result if the clustering center after iteration is not changed any more, and returning to the iteration updating execution module if the clustering center after iteration is not changed any more.
7. The system according to claim 6, wherein the step of acquiring the user location coordinates and the service type information in the information acquisition module specifically comprises:
and the fog wireless access network controller acquires the user position coordinates and the service type information through the fog access node.
8. The user clustering system combining interference suppression in a fog wireless access network as claimed in claim 6, wherein in the preprocessing module, the step of performing normalization and weighted combination processing on the user position coordinates and the service types to obtain the processed data space specifically comprises:
carrying out MinMax normalization and weighted combination processing on the position coordinates of the user and the service types; wherein the weighted combination processing comprises: multiplying the user position coordinate and the service type by a position weight omega respectively1Type weight omega2Thereafter, clustering is started by changing ω1,ω2The value of (d) changes the degree of influence of the traffic class on the distance.
9. The system of claim 8, wherein the preprocessing module is configured to perform the operations of changing ω and combining interference suppression with the mist radio access network1,ω2Change the course of influence of the traffic class on the distanceThe method comprises the following steps:
ω1=1,ω2when the ratio is 0: based only on user location coordinates;
ω1=ω20.5: equally considering the position coordinates of the user and the service type;
ω1=0.2,ω20.8: emphasis is placed on the traffic class.
10. The system according to claim 6, wherein the mapping module performs laplacian eigenmapping on the processed data space, and the step of mapping to an eigen data space with inter-device interference information specifically includes:
in the feature mapping process, the inter-cluster interference received by the user is represented as:
in the formula, Cl、MnRespectively corresponding column vectors of a user-clustering indication array and a user-physical resource block indication array; w is an adjacency matrix constructed from a kernel function, and the weight W of the corresponding positionijRepresenting co-channel interference between users at corresponding positions; l represents the index of the user cluster, p represents the total number of the user clusters, and c if the user i belongs to the cluster lil1 is ═ 1; n denotes the index of the PRB to which the user is attached, k denotes the total number of PRBs, if the user i is attached to the PRBnThen m isin1 is ═ 1; all user equipment which multiplexes the same PRB and belongs to different clusters are indicated in brackets;
the kernel function is represented as:
κ(xi,xj)=K(||xi-xj||)-α
where K ═ P ξ > 0, is a parameter determined by the F-UE transmit power and the path loss; α is a constant coefficient parameter based on antenna characteristics and average path loss; x is the number ofi,xjAre two input variables of the kernel function;
calculating the degree of each node and forming a degree matrix D ═ diag (D)1,…,dn) The computational expression of the degree of a node is,
wherein j represents the node index, N represents the total number of nodes, diRepresenting the sum of the weights corresponding to the node i and all the connected nodes;
laplace matrix L ═ D-W, according to the formula L ═ D-1/2LD-1/2And carrying out standardization processing on the Laplace matrix to complete the feature mapping of the user.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110444218.6A CN113194031B (en) | 2021-04-23 | 2021-04-23 | User clustering method and system combining interference suppression in fog wireless access network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110444218.6A CN113194031B (en) | 2021-04-23 | 2021-04-23 | User clustering method and system combining interference suppression in fog wireless access network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113194031A true CN113194031A (en) | 2021-07-30 |
CN113194031B CN113194031B (en) | 2023-03-31 |
Family
ID=76978302
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110444218.6A Active CN113194031B (en) | 2021-04-23 | 2021-04-23 | User clustering method and system combining interference suppression in fog wireless access network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113194031B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110968426A (en) * | 2019-11-29 | 2020-04-07 | 西安交通大学 | Edge cloud collaborative k-means clustering model optimization method based on online learning |
CN114363988A (en) * | 2021-12-10 | 2022-04-15 | 北京佰才邦技术股份有限公司 | Clustering method and device and electronic equipment |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120059777A1 (en) * | 2010-09-03 | 2012-03-08 | The University Of Utah Research Foundation | Characterizing datasets using sampling, weighting, and approximation of an eigendecomposition |
CN106991430A (en) * | 2017-02-28 | 2017-07-28 | 浙江工业大学 | A kind of cluster number based on point of proximity method automatically determines Spectral Clustering |
CN107169500A (en) * | 2017-03-09 | 2017-09-15 | 中国矿业大学 | A kind of Spectral Clustering about subtracted based on neighborhood rough set and system |
US20180131575A1 (en) * | 2016-11-07 | 2018-05-10 | Beijing Jiaotong University | Method for clustering wireless channel mpcs based on a kpd doctrine |
CN109033415A (en) * | 2018-08-06 | 2018-12-18 | 浙江大学 | A kind of dimensionality reduction and method for visualizing of the multidimensional nerve signal based on laplacian eigenmaps |
CN109190661A (en) * | 2018-07-26 | 2019-01-11 | 安徽师范大学 | A kind of Spectral Clustering based on difference secret protection |
CN109327838A (en) * | 2018-11-20 | 2019-02-12 | 吉林大学 | Frequency spectrum resource allocation method based on spectral clustering in a kind of isomery cellular network |
CN109409407A (en) * | 2018-09-19 | 2019-03-01 | 西安理工大学 | A kind of industry monitoring data clustering method based on LE algorithm |
CN110753065A (en) * | 2019-10-28 | 2020-02-04 | 国网河南省电力公司信息通信公司 | Network behavior detection method, device, equipment and storage medium |
US10685008B1 (en) * | 2016-08-02 | 2020-06-16 | Pindrop Security, Inc. | Feature embeddings with relative locality for fast profiling of users on streaming data |
-
2021
- 2021-04-23 CN CN202110444218.6A patent/CN113194031B/en active Active
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120059777A1 (en) * | 2010-09-03 | 2012-03-08 | The University Of Utah Research Foundation | Characterizing datasets using sampling, weighting, and approximation of an eigendecomposition |
US10685008B1 (en) * | 2016-08-02 | 2020-06-16 | Pindrop Security, Inc. | Feature embeddings with relative locality for fast profiling of users on streaming data |
US20180131575A1 (en) * | 2016-11-07 | 2018-05-10 | Beijing Jiaotong University | Method for clustering wireless channel mpcs based on a kpd doctrine |
CN106991430A (en) * | 2017-02-28 | 2017-07-28 | 浙江工业大学 | A kind of cluster number based on point of proximity method automatically determines Spectral Clustering |
CN107169500A (en) * | 2017-03-09 | 2017-09-15 | 中国矿业大学 | A kind of Spectral Clustering about subtracted based on neighborhood rough set and system |
CN109190661A (en) * | 2018-07-26 | 2019-01-11 | 安徽师范大学 | A kind of Spectral Clustering based on difference secret protection |
CN109033415A (en) * | 2018-08-06 | 2018-12-18 | 浙江大学 | A kind of dimensionality reduction and method for visualizing of the multidimensional nerve signal based on laplacian eigenmaps |
CN109409407A (en) * | 2018-09-19 | 2019-03-01 | 西安理工大学 | A kind of industry monitoring data clustering method based on LE algorithm |
CN109327838A (en) * | 2018-11-20 | 2019-02-12 | 吉林大学 | Frequency spectrum resource allocation method based on spectral clustering in a kind of isomery cellular network |
CN110753065A (en) * | 2019-10-28 | 2020-02-04 | 国网河南省电力公司信息通信公司 | Network behavior detection method, device, equipment and storage medium |
Non-Patent Citations (5)
Title |
---|
L. HUANG: "A Multi-Dimensional Features-based Clustering Algorithm for Massive MIMO System", 《2019 14TH INTERNATIONAL CONFERENCE ON COMPUTER SCIENCE & EDUCATION (ICCSE)》, 23 September 2019 (2019-09-23) * |
R. TANG: "User-Centric Joint Admission Control and Resource Allocation for 5G D2D Extreme Mobile Broadband: A Sequential Convex Programming Approach", 《 IEEE COMMUNICATIONS LETTERS》, 13 March 2017 (2017-03-13) * |
秦红星等: "基于拉普拉斯特征映射的传递函数设计方法", 《计算机工程与设计》, no. 01, 16 January 2015 (2015-01-16) * |
郑建炜等: "块对角拉普拉斯约束的平滑聚类算法", 《计算机辅助设计与图形学学报》 * |
郑建炜等: "块对角拉普拉斯约束的平滑聚类算法", 《计算机辅助设计与图形学学报》, no. 01, 15 January 2018 (2018-01-15) * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110968426A (en) * | 2019-11-29 | 2020-04-07 | 西安交通大学 | Edge cloud collaborative k-means clustering model optimization method based on online learning |
CN110968426B (en) * | 2019-11-29 | 2022-02-22 | 西安交通大学 | Edge cloud collaborative k-means clustering model optimization method based on online learning |
CN114363988A (en) * | 2021-12-10 | 2022-04-15 | 北京佰才邦技术股份有限公司 | Clustering method and device and electronic equipment |
CN114363988B (en) * | 2021-12-10 | 2023-07-07 | 北京佰才邦技术股份有限公司 | Clustering method and device and electronic equipment |
Also Published As
Publication number | Publication date |
---|---|
CN113194031B (en) | 2023-03-31 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112181666B (en) | Equipment assessment and federal learning importance aggregation method based on edge intelligence | |
CN111586720B (en) | Task unloading and resource allocation combined optimization method in multi-cell scene | |
CN110798849A (en) | Computing resource allocation and task unloading method for ultra-dense network edge computing | |
CN111800828B (en) | Mobile edge computing resource allocation method for ultra-dense network | |
CN112105062B (en) | Mobile edge computing network energy consumption minimization strategy method under time-sensitive condition | |
CN113194031B (en) | User clustering method and system combining interference suppression in fog wireless access network | |
CN110233755B (en) | Computing resource and frequency spectrum resource allocation method for fog computing in Internet of things | |
CN114356545B (en) | Task unloading method oriented to privacy protection and energy consumption optimization | |
CN113687875B (en) | Method and device for unloading vehicle tasks in Internet of vehicles | |
CN114585006B (en) | Edge computing task unloading and resource allocation method based on deep learning | |
CN113590279A (en) | Task scheduling and resource allocation method for multi-core edge computing server | |
CN114389652A (en) | Low-power-consumption large-connection method for large-scale cellular MIMO network | |
CN113038612B (en) | Cognitive radio power control method based on deep learning | |
CN108282888B (en) | D2D resource allocation method based on improved fuzzy clustering | |
KR20230032754A (en) | Apparatus and Method for Task Offloading of MEC-Based Wireless Network | |
CN109412661A (en) | A kind of user cluster-dividing method under extensive mimo system | |
CN111930435B (en) | Task unloading decision method based on PD-BPSO technology | |
CN113342504A (en) | Intelligent manufacturing edge calculation task scheduling method and system based on cache | |
CN112988275A (en) | Task perception-based mobile edge computing multi-user computing unloading method | |
CN117156450A (en) | Mobile edge computing task unloading method based on improved artificial buzzing algorithm | |
CN113709817B (en) | Task unloading and resource scheduling method and device under multi-base-station multi-server scene | |
CN111935824B (en) | Wireless resource allocation strategy updating method, device, equipment and storage medium | |
CN113873525A (en) | Task unloading method and terminal for ultra-dense edge computing network | |
CN110753374B (en) | Task-unloading-oriented cooperative network relay node selection method and device | |
CN114513814A (en) | Edge network computing resource dynamic optimization method based on unmanned aerial vehicle auxiliary node |
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 |