CN101808325B - Method and device for allocating frequency spectrum - Google Patents
Method and device for allocating frequency spectrum Download PDFInfo
- Publication number
- CN101808325B CN101808325B CN2010101073444A CN201010107344A CN101808325B CN 101808325 B CN101808325 B CN 101808325B CN 2010101073444 A CN2010101073444 A CN 2010101073444A CN 201010107344 A CN201010107344 A CN 201010107344A CN 101808325 B CN101808325 B CN 101808325B
- Authority
- CN
- China
- Prior art keywords
- spectrum
- frequency spectrum
- nodes
- node
- information
- 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
- 238000001228 spectrum Methods 0.000 title claims abstract description 356
- 238000000034 method Methods 0.000 title claims abstract description 37
- 239000011159 matrix material Substances 0.000 claims description 57
- 230000008901 benefit Effects 0.000 claims description 47
- 241000854291 Dianthus carthusianorum Species 0.000 claims description 46
- 230000011218 segmentation Effects 0.000 claims description 37
- 239000013598 vector Substances 0.000 claims description 19
- 229920006395 saturated elastomer Polymers 0.000 claims description 14
- 238000003860 storage Methods 0.000 claims description 4
- 238000007726 management method Methods 0.000 description 10
- 230000001149 cognitive effect Effects 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000010276 construction Methods 0.000 description 3
- 238000009827 uniform distribution Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000013468 resource allocation Methods 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 230000008685 targeting Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a method and a device for allocating frequency spectrum. The method comprises the following steps of: S1, acquiring available frequency spectrum information and constructing a spectrogram model according to network topology; S2, performing spectrogram segmenting on the spectrogram model and acquiring different clusters; S3, managing the clusters obtained by the spectrogram segmenting; and S4, calculating a frequency spectrum efficiency value of nodes of the clusters, and allocating the frequency spectrum to the nodes of the clusters according to the frequency spectrum efficiency value. According to the method and the device for allocating the frequency spectrum, the use ratio of the frequency spectrum resource reaches the maximum, the allocation of the frequency spectrum resource can be balanced, and the justification of the frequency spectrum resource use is improved.
Description
[ technical field ] A method for producing a semiconductor device
The present invention relates to a method and an apparatus for allocating frequency spectrums, and more particularly, to a method and an apparatus for allocating frequency spectrums.
[ background of the invention ]
With the rapid development of society, economy and science and technology, the application field of radio technology is wider and wider, and the demand of radio services on frequency resources in urban emergency, subway, port and other departments is extremely urgent. Meanwhile, the research and development of various new technologies and the demand of new services for frequency resources are rapidly increasing, and the problem of shortage of radio frequency has become a bottleneck restricting the development of radio communication.
Conventionally, for the management of radio spectrum, a fixed spectrum allocation is adopted to authorize users to use the spectrum. However, not all authorized users (authorized users) occupy the licensed spectrum all the day long, and some frequency bands are not used by users in part of the time, that is, spectrum sharing cannot be achieved, which causes a great waste of spectrum resources.
It is desirable to address the current needs and trends. Cognitive radio is an intelligent communication device which changes self transmission parameters according to surrounding communication environments to meet communication requirements, so that an unauthorized user (cognitive user) can automatically sense the condition that an authorized user uses a frequency spectrum, and when the authorized user does not use the frequency spectrum, namely the frequency spectrum is idle, the frequency spectrum is subjected to opportunistic selection access.
However, when spectrum is effectively allocated in a spectrum management device of a cognitive radio network, the spectrum availability changes dynamically due to the presence of authorized users, the degradation of the quality of the spectrum, and other factors, and thus, the spectrum resource allocation is unbalanced. Meanwhile, due to the increase of the cognitive users, the spectrum allocation complexity is exponentially increased.
[ summary of the invention ]
In view of this, it is necessary to provide a spectrum allocation method based on a spectrogram segmentation technique for the technical defects of unbalanced spectrum resource allocation and high complexity of spectrum allocation.
In addition, it is necessary to provide a spectrum allocation apparatus based on the spectrogram segmentation technique.
The invention provides a frequency spectrum allocation method, which comprises the following steps: step S1, acquiring available spectrum information to construct a spectrum graph model according to network topology; step S2, performing spectrogram segmentation on the spectrogram model, and obtaining different clusters; step S3, managing the clusters obtained by spectrogram segmentation; step S4, calculating the frequency spectrum benefit value of the clustered node, and distributing the frequency spectrum to the clustered node according to the frequency spectrum benefit value; the step S1 includes the following steps in sequence: s11, acquiring available spectrum information and setting an attribute matrix; s12, establishing a frequency spectrum graph model according to the attribute matrix; the frequency spectrum graph model is G ═ V, (E) and Q, (V) is a node set, E is an edge set, and Q is a storage available frequency spectrum set; step S2 includes the following steps in order: s21, constructing a diagonal matrix of a spectrogram model according to the spectrogram model; s22, constructing a Laplace matrix according to the diagonal matrix; s23, calculating a characteristic vector according to the Laplace matrix and obtaining a corresponding element value; s24, segmenting the spectrogram model according to the element values of the feature vectors to obtain different clusters; step S4 includes the following steps in order: s41, the clustered nodes calculate spectrum benefit values according to benefit matrixes in the attribute matrixes; s42, comparing the clustered nodes through the spectrum benefit values; s43, if the benefit value obtained by the clustered nodes on the available frequency spectrum is the maximum, obtaining the frequency spectrum; s44, if the benefit values obtained by the clustered nodes on the available frequency spectrum are equal, the clustered nodes with less bandwidth obtain the frequency spectrum; and S45, when the spectrum allocation is finished or/and the node spectrum demand is saturated, the spectrum allocation is terminated.
Preferably, step S3 includes the following steps in sequence: s31, the clustering selects a cluster head; s32, the cluster head stores the member table information of the cluster node; s33, the clustering node senses frequency spectrum information and sends the frequency spectrum information to a cluster head of a cluster to which the clustering node belongs; s34, the cluster head sends the received spectrum information to the central controller for managing the cluster; s35, the central controller obtains clustered spectrum use information according to the spectrum information and feeds the spectrum use information back to a clustered cluster head; and S36, the node receives the frequency spectrum using information from the cluster head of the cluster to which the node belongs at regular time, and a new node in the node selects the cluster according to the frequency spectrum using information.
Preferably, the method further comprises step S5: and updating the frequency spectrum information between the nodes according to the frequency spectrum allocation result.
Preferably, step S5 includes the following steps in sequence: s51, the node for obtaining the frequency spectrum deletes the allocated available frequency spectrum information of the node; s52, the connected nodes of the frequency spectrum obtaining node delete the distributed available frequency spectrum information; s53, deleting nodes without available frequency spectrum; s54, deleting the nodes with saturated spectrum allocation; s55 allocates spectrum to the individual nodes without connected nodes until saturation and deletes the individual nodes.
The invention also provides a spectrum allocation device based on spectrogram segmentation, which at least comprises a spectrogram model module, a spectrogram segmentation module, a clustering management module and a spectrum allocation module which are connected in sequence: the spectrum map model module is used for constructing a spectrum map model according to available spectrum information and network topology according to the available spectrum information; the frequency spectrum graph model is G ═ V, (E) and Q, (V) is a node set, E is an edge set, and Q is a storage available frequency spectrum set; the spectrogram segmentation module is used for carrying out spectrogram segmentation on the spectrogram model and obtaining different clusters; the clustering management module is used for managing clusters obtained by spectrogram segmentation; the frequency spectrum allocation module is used for calculating the frequency spectrum benefit value of the clustered nodes and allocating frequency spectrums to the clustered nodes according to the frequency spectrum benefit value; the frequency spectrum map model module is also used for acquiring available frequency spectrum information, setting an attribute matrix and establishing a frequency spectrum map model according to the attribute matrix; the spectrogram segmentation module is further used for constructing a diagonal matrix according to the spectrum model, then constructing a Laplace matrix, and calculating a feature vector and obtaining a corresponding element value according to the Laplace matrix; the spectrum graph model is further divided according to the element values of the feature vectors to obtain different clusters; the spectrum allocation module is further configured to calculate a spectrum benefit value of the clustered nodes, acquire a clustered node with a maximum benefit value, allocate the spectrum to the clustered node with the maximum benefit value or acquire clustered nodes with equal benefit values, and allocate the spectrum to clustered nodes with equal benefit values and less bandwidths.
Preferably, the cluster management module includes a central controller, and the cluster management module is further configured to enable the cluster to select a cluster head, where the cluster head stores member table information of the cluster node, and is further configured to enable the cluster node to sense surrounding spectrum information and send the spectrum information to the cluster head of the cluster to which the cluster node belongs, and the cluster head sends received spectrum information to the central controller managing the cluster at regular time; the central controller is used for acquiring clustered spectrum use information according to the received spectrum information sent by the cluster head; the central controller feeds back frequency spectrum use information to the clustered cluster head; and selecting the cluster by a new node in the nodes according to the frequency spectrum use information.
Preferably, the apparatus further includes an updating module connected to the spectrum allocation module, and the updating module is configured to update spectrum information between the nodes according to a spectrum allocation result.
Preferably, the updating module is further configured to delete available spectrum information already allocated to the node that obtains a spectrum, available spectrum information already allocated to nodes connected to the node that obtains a spectrum, a node without an available spectrum, and a node whose spectrum allocation is saturated; and the method is also used for allocating the frequency spectrum to the independent nodes without the connected nodes until the independent nodes are saturated and deleting the independent nodes.
The invention has the following beneficial effects:
the spectrum allocation method comprises the steps of firstly constructing a spectrum graph model, then dividing the spectrum graph model by adopting a spectrum division technology to obtain different clusters and managing and allocating the divided clusters, so that the utilization rate of spectrum resources is maximized, the allocation of the spectrum resources can be balanced, and the fairness of the utilization of the spectrum resources is improved;
when the spectrum allocation is finished or/and the node spectrum demand is saturated, the spectrum allocation is terminated; according to the node requirements, the calculation method of the convergence spectrum is fast, and spectrum allocation is carried out concurrently, so that the allocation and load of spectrum resources are more balanced, and the fairness and the utilization rate of the spectrum resources are further improved;
the method comprises the steps of selecting the most clusters of available frequency spectrums for nodes (cognitive users) which move dynamically and enter or leave and adding the clusters to achieve uniform distribution of frequency spectrum resources, further improving the fairness of frequency spectrum resource utilization and further improving the utilization rate of the frequency spectrum resources;
the network topology structure is updated, the uniform distribution of the frequency spectrum resources can be carried out on the dynamically mobile transformed nodes again, the fairness of the utilization of the frequency spectrum resources is further improved, and the utilization rate of the frequency spectrum resources is further improved.
[ description of the drawings ]
For a better understanding of the nature and technical aspects of the present invention, reference should be made to the following drawings which are provided for purposes of illustration and description, and are not intended to limit the invention.
FIG. 1 is a flow chart of a method for spectrum allocation according to the present invention;
FIG. 2 is a diagram of a spectrum allocation method according to an embodiment of the present invention;
FIG. 3 is another diagram illustrating a spectrum allocation method according to an embodiment of the present invention;
fig. 4 is a block diagram of a spectrum allocation apparatus according to the present invention.
[ detailed description ] embodiments
The technical solution and other advantages of the present invention will become apparent from the following detailed description of specific embodiments of the present invention, which is to be read in connection with the accompanying drawings.
The invention divides a spectrum graph model into a plurality of different clusters through spectrogram division, autonomously selects a cluster head through each cluster, manages the clusters, and then distributes spectrum to nodes (namely cognitive users) in the clusters.
Referring to fig. 1, the spectrum allocation method of the present invention includes:
firstly, step S1, constructing a spectrogram model, acquiring available spectrum information, and constructing the spectrogram model according to network topology;
the method comprises the following specific steps:
s11, setting the following attribute matrix for the available spectrum information:
spectrum availability: a ═ ai，k|ai，k∈{0，1}}N×MWhich represents the availability of spectrum, wherein
Interference limitation matrix: c ═ Ci，j，k|ci，j，k∈{0，1}}N×N×MAnd representing that the nodes interfere with the limitation on the same frequency spectrum. Wherein,
spectrum allocation matrix: s ═ Si，k|si，k∈{0，1}}N×MRepresents an efficient spectrum allocation scheme in which
In addition, the matrix S needs to satisfy the constraint defined by the interference constraint matrix C, that is: si，k·sj，k＝0，if ci，j，k＝1， <math>
<mrow>
<mo>∀</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo><</mo>
<mi>N</mi>
<mo>,</mo>
<mi>k</mi>
<mo><</mo>
<mi>M</mi>
<mo>;</mo>
</mrow>
</math>
Neighbor set: h ═ Hi，k}N×MRepresenting the number of neighbors (neighbors between cluster heads or neighbors between nodes in a cluster) of each node i on a frequency spectrum k;
bandwidth matrix: b ═ Bi，k}N×MA node i is described to successfully obtain the bandwidth of the available spectrum k;
a benefit matrix: r ═ Ri，k}N×MRepresenting the benefit value obtained by node i using spectrum k, i.e. ri，kIs defined as: <math>
<mrow>
<msub>
<mi>r</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
<mo>=</mo>
<mi>max</mi>
<mfrac>
<msub>
<mi>b</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
<mrow>
<mrow>
<mo>(</mo>
<msub>
<mi>h</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
<mo>+</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>·</mo>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>0</mn>
</mrow>
<mi>M</mi>
</munderover>
<msub>
<mi>a</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
<mo>·</mo>
<msub>
<mi>b</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
</mrow>
</mfrac>
<mo>,</mo>
</mrow>
</math>
wherein, in order to avoid the condition that the denominator is 0, the number h of neighbors of the node i on the k frequency spectrumi，kIs changed into hi，k+1；
Based on the attribute matrix set by the available spectrum information, a spectrogram model can be built, i.e., step S12: setting a frequency spectrum graph model according to the attribute matrix;
that is, a undirected spectrum graph model G ═ V, E, Q is defined, where V is a node set, E is an edge set, Q stores an available spectrum set, ViRepresenting cognitive nodes, eijRepresents viAnd vjThe edge therebetween;
the adjacent matrix corresponding to the undirected spectrogram model G is W ═ Wij]，wijIs an edge eijThe weight of (2). w is aijReference to viTo vjThe number of available spectrum in between.
Step S2, performing spectrogram segmentation on the spectrogram model, and obtaining different clusters;
the spectrogram segmentation technology uses an adjacency matrix and a Laplacian (Laplacian) matrix to represent a graph model, and then characteristic values of the Laplacian (Laplacian) matrix are solved; for the second smallest of the eigenvalues λ2Corresponding feature vector x2And using the second small feature vector x2To divide the spectrogram. Feature vector x2The elements of (a) represent the weights of the corresponding nodes in the spectrogram model, so that the difference between the elements reflects the difference between the nodes;
according to this rule, the spectrogram model is divided into two parts, the second smallest eigenvector x2If the value of the characteristic element corresponding to the node serial number is less than 0, the node representing the serial number is classified into one frequency spectrum subgraph model, and the other nodes larger than 0 are classified into the other frequency spectrum subgraph model;
specifically, the spectrogram segmentation technology is a cyclic iteration process, wherein in each iteration process, a spectrogram model (or referred to as a spectrum parent graph model) is divided into two spectrum subgraph models (namely, clustering, and referred to as a spectrum subgraph model in the step for visually understanding the relationship of the spectrogram models); a spectrogram segmentation algorithm is run according to the spectrogram model G ═ (V, E, Q) as input in step S12, and the detailed steps of the process are as follows:
s21: constructing a diagonal matrix D ═ D of the spectrogram model Gij]Wherein <math>
<mrow>
<msub>
<mi>d</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>=</mo>
<mfenced open='{' close=''>
<mtable>
<mtr>
<mtd>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msub>
<mi>w</mi>
<mi>ik</mi>
</msub>
<mo>,</mo>
</mtd>
<mtd>
<mi>if i</mi>
<mo>=</mo>
<mi>j</mi>
</mtd>
</mtr>
<mtr>
<mtd>
<mn>0</mn>
<mo>,</mo>
</mtd>
<mtd>
<mi>otherwise</mi>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>;</mo>
</mrow>
</math>
S22: constructing a Laplacian (Laplacian) matrix L according to the diagonal matrix, wherein L is D-W, and W is the adjacent matrix in the step S1;
s23: calculating the second smallest eigenvalue λ according to Laplacian matrix2Corresponding feature vector x2；
S24: according to the second small feature vector x2The value of the element(s) of (c) is used to segment the spectrogram model.
Simultaneously, respectively defining the number of internal edges and the number of external edges for the frequency spectrum graph model, wherein the number of the internal edges refers to the number of the connecting edges between nodes in the frequency spectrum sub-graph model (cluster), and the number of the external edges refers to the number of the connecting edges between the frequency spectrum sub-graph models (cluster);
controlling whether the iterative algorithm is terminated or not through the internal edge number and the external edge number; if the number of internal edges of at least one spectrum subgraph model (cluster) exceeds the number of external edges, the spectrum subgraph model (cluster) is not converged in a spectrum parent graph model, the division of the spectrum subgraph model is stopped, and finally the generated collection of spectrum subgraph models (clusters) is obtained.
Step S3, clustering management, managing the clusters obtained by spectrogram segmentation; the further steps are as follows:
s31, clustering and selecting a cluster head, wherein the node with the most edges is used as the cluster head in all the nodes (cognitive users) in the cluster;
s32, the cluster head stores the member table information of the node, and the cluster head stores the member table information of the node in the cluster to which the cluster head belongs;
s33, sending the frequency spectrum information sensed by the clustering nodes to the cluster heads of the clusters to which the nodes belong;
s34, the cluster head sends the received spectrum information to the central controller for managing the cluster; the cluster head timing (the time can be set by self) forwards the received spectrum information of the cluster to the central controller;
s35, the central controller analyzes and obtains the information of the use of the clustering frequency spectrum according to the frequency spectrum information received from the cluster head of each cluster, and the frequency spectrum use information is processed and fed back to the cluster head of the cluster;
s36, the node receives the spectrum use information from the cluster head of the cluster at regular time (the time can be set by itself); the node selects the cluster according to the spectrum use information, and the node comprises: moving, entering or leaving the new nodes of the cluster, selecting the cluster with the most frequency spectrum and adding the cluster;
according to the movement, entering or leaving of the nodes, the nodes select and join the clusters with the most available frequency spectrum according to the frequency spectrum use information, so that the frequency spectrum utilization rate of the nodes is improved, and the fairness of frequency spectrum resource utilization is improved.
Step S4, calculating the frequency spectrum benefit value of the clustered node, and distributing the frequency spectrum to the clustered node according to the calculated value; the further steps are as follows:
s41, the clustered nodes calculate the spectrum benefit value according to the benefit matrix in the attribute matrix, and each perceived available spectrum is according to the benefit matrix in the step 2
Calculating different benefit values;
s42, distributing frequency spectrums of each frequency spectrum in the available frequency spectrums one by one to nodes in a cluster, and comparing the benefit values of each available frequency spectrum with the benefit values of the frequency spectrums of the neighbors one by the nodes;
s43, if the node obtains the maximum benefit value return on the available frequency spectrum in the neighborhood, the node obtains the frequency spectrum;
s44, if the node and its neighbor have the same benefit value return, then the node with less bandwidth obtains frequency spectrum;
s45, after the current spectrum allocation is completed, each node broadcasts the spectrum allocation information and the updated available spectrum information to its neighbors and prepares to enter the next iteration cycle; and terminating the spectrum allocation until all the spectrums are allocated or/and the spectrum demands of all the nodes are saturated.
The spectrum allocation method based on the spectrogram segmentation technology can minimize the weight of connection between nodes, maximize the weight of internal connection of the nodes, quickly converge the calculation method of the spectrum, and concurrently perform spectrum allocation, so that the utilization rate of the spectrum resources is maximized, the allocation of the spectrum resources can be balanced, and the fairness of the utilization of the spectrum resources is further improved.
The method for spectrum allocation based on spectrogram segmentation further includes step S5: after the frequency spectrum is distributed, updating the structure (namely a network topological structure) between the nodes; the specific updating method comprises the following steps:
s51 the node obtaining the frequency spectrum deletes the allocated available frequency spectrum information of the node;
the connected nodes of the spectrum obtaining node delete the allocated available spectrum information S52;
s53 deleting nodes without available frequency spectrum;
s54 deleting nodes whose spectrum allocation has been saturated;
s55 allocates spectrum to an individual node that has no connected nodes until saturation and deletes the individual node.
The network topology structure is updated, so that the uniform distribution of the spectrum resources can be carried out on the dynamically mobile transformed nodes again, the fairness of the utilization of the spectrum resources is further improved, and the utilization rate of the spectrum resources is further improved.
An embodiment is provided based on the above method in conjunction with fig. 2 and 3, specifically as follows:
firstly, constructing a frequency spectrum graph model; setting up a maximum of 5 available frequency spectrums (CH1, CH2, CH3, CH4, CH5), 10 nodes (SU1, SU2, … SU10), the topology structure formed by them is shown in fig. 2, and the adjacency matrix of the diagram can be expressed as:
the rows and columns of the adjacency matrix represent the cognitive nodes SU1-10, and the values within the adjacency matrix represent the amount of spectrum available between the nodes.
Then, the spectrogram model is divided into different clusters by the frequency spectrum, which comprises the following steps:
constructing a diagonal matrix D ═ D of the spectrogram model Gij]，
Constructing a Laplace matrix L, and enabling L to be D-W
The eigenvector x corresponding to the second smallest eigenvalue λ 2 is calculated for the Laplacian (Laplacian) matrix2，x2＝-0.0029、-0.0026、-0.0031、0.2982、0.2679、0.3355、0.3834、-0.2864、-0.4331、-0.5569
The feature vector x2The nodes corresponding to the elements smaller than 0 are divided into a spectrum subgraph model (cluster) G1([v1，...v3，v8，...v10]) And the nodes corresponding to the elements larger than 0 are classified into another spectrum subgraph model (cluster) G2([v4，...v7])；G1The number of inner sides of (2) is 5 and the number of outer sides is 2; g2The number of inner sides of (2) is 3, and the number of outer sides is 2; the number of internal edges of both subgraphs is greater than the number of external edges, so the partitioning continues.
And carrying out a cycle iteration process of spectrogram segmentation.
Construction of a spectral subgraph model G1Adjacency matrix of model
Construction of a spectral subgraph model G1Is diagonal matrix of
Construction of subgraph G1Laplacian (Laplacian) matrix of
Finding the second smallest eigenvalue λ2Corresponding feature vector x2
x2＝-0.3813、-0.3219、-0.4348、0.1214、0.3854、0.6313
The second smallest eigenvalue λ2Corresponding feature vector x2Nodes corresponding to elements smaller than 0 in the feature vector are divided into a spectrum subgraph model (cluster) G3([v1，...v3]) And the nodes corresponding to the elements larger than 0 are classified into another spectrum subgraph model (cluster) G4([v8，...v10])；G3OfThe number of sides is 2, and the number of external sides is 1; g4The number of internal sides of (1) is 2 and the number of external sides is 1; due to G2、G3、G4The segmentation is not converged, so that the final result of the graph segmentation can be obtained as G2([v4，...v7])，G3([v1，...v3])，G4([v8，...v10])。
Then managing the clusters, including: cluster head selection, in order to better adapt to such changes, proposes a spectrum allocation algorithm targeting fast convergence. Allocating available frequency spectrums to nodes as much as possible in one iteration process of frequency spectrum allocation, so that one complete frequency spectrum allocation process is completed by using as few iteration times as possible; the algorithm is executed on a plurality of clusters obtained in the step of spectrogram segmentation, so that the convergence rate of the frequency spectrum allocation can be greatly increased;
and the cluster head of each cluster autonomously manages the nodes, and when a new node appears or the node moves, the node periodically senses the frequency spectrum use information sent by the central controller, selects the cluster with the most available frequency spectrum, joins the cluster, and sends the cluster information by the cluster head.
Finally, allocating frequency spectrums;
the spectrum allocation algorithm is executed concurrently on all clusters obtained after the spectrogram model is subjected to spectrogram segmentation for different clusters in the step of spectrum segmentation;
therefore, it is sufficient to enumerate a clustered spectrum allocation, which is denoted as G below4([v8，...v10]) For example, spectrum allocation is performed. Assuming a bandwidth of 0.5M per spectrum, the available spectrum list of SU8 is: (CH1, CH3, CH5), bandwidth requirement of 0.8M; the available spectrum list for SU9 is: (CH1, CH3, CH4, CH5), bandwidth requirement 0.9M; the available spectrum list for SU10 is: (CH1, CH2, CH3, CH4) with a bandwidth requirement of 1.0M. Then the spectrum allocated to SU8 is: (CH1, CH3), SU9 is allocated the frequency spectrum: (CH4, CH5) since there is no space between SU9 and SU8Spectrum collisions, and therefore the same spectrum can be used simultaneously, then the spectrum allocated to SU9 is: (CH1, CH 2). So far the three nodes are all saturated, the allocation is terminated.
The device for spectrum allocation based on spectrogram segmentation at least comprises a spectrogram model module, a spectrogram segmentation module, a clustering management module and a spectrum allocation module which are connected in sequence;
the system comprises a spectrogram model module, a network topology establishing module and a spectrum attribute matrix establishing module, wherein the spectrogram model module is used for establishing a spectrogram model according to network topology aiming at available spectrum information, setting a spectrum attribute matrix for the spectrogram model and establishing an undirected spectrogram model G as (V, E, Q);
the spectrogram segmentation module is used for carrying out spectrogram segmentation on the spectrogram model and obtaining different clusters; the spectrogram segmentation is carried out on the spectrogram model, and different clusters are obtained; the spectrogram segmentation module constructs a diagonal matrix according to the spectrum model, then constructs a Laplace matrix, and calculates a characteristic vector and obtains a corresponding element value according to the Laplace matrix; and the method is also used for segmenting the frequency spectrum graph model according to the element values of the characteristic vectors to obtain different clusters.
The clustering management module is used for managing clusters obtained by spectrogram segmentation; the clustering management module is used for enabling the clustering to select a cluster head, the cluster head stores member table information of the clustering nodes, and is also used for enabling the clustering nodes to sense surrounding frequency spectrum information and send the frequency spectrum information to the cluster head of the clustering to which the clustering nodes belong, and the cluster head sends the received frequency spectrum information to a central controller for managing the clustering at regular time; the central controller is used for receiving the frequency spectrum information sent by the cluster head and obtaining clustered frequency spectrum use information according to the frequency spectrum information; the central controller feeds back the frequency spectrum use information to the cluster head; and selecting the cluster by a new node in the nodes according to the frequency spectrum use information.
The frequency spectrum allocation module is used for calculating the frequency spectrum benefit value of the clustered nodes and allocating frequency spectrums to the clustered nodes according to the frequency spectrum benefit value; the frequency spectrum allocation module is used for calculating the frequency spectrum benefit value of the clustered nodes, acquiring the clustered nodes with the maximum benefit value, allocating the frequency spectrum to the clustered nodes with the maximum benefit value or allocating the frequency spectrum to the clustered nodes with the equal benefit values and less bandwidths.
The device for spectrum allocation based on spectrogram segmentation further comprises an updating module connected with the spectrum allocation module, wherein the updating module is used for updating spectrum information (namely a network topology) between nodes according to a spectrum allocation result; the updating module is further used for deleting available spectrum information which is allocated to the node for obtaining the spectrum, available spectrum information which is allocated to the nodes connected with the node for obtaining the spectrum, nodes without available spectrum and nodes with saturated spectrum allocation; and allocating spectrum to an independent node without connected nodes until saturation, and deleting the independent node without available spectrum.
The above-mentioned embodiments only express the embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (8)
1. A method of spectrum allocation comprising the steps of:
step S1, acquiring available spectrum information to construct a spectrum graph model according to network topology;
step S2, performing spectrogram segmentation on the spectrogram model, and obtaining different clusters;
step S3, managing the clusters obtained by spectrogram segmentation;
step S4, calculating the frequency spectrum benefit value of the clustered node, and distributing the frequency spectrum to the clustered node according to the frequency spectrum benefit value;
the step S1 includes the following steps in sequence:
s11, acquiring available spectrum information and setting an attribute matrix;
s12, establishing a frequency spectrum graph model according to the attribute matrix;
the frequency spectrum graph model is G ═ V, (E) and Q, (V) is a node set, E is an edge set, and Q is a storage available frequency spectrum set;
step S2 includes the following steps in order:
s21, constructing a diagonal matrix of the spectrogram model according to the spectrogram model;
s22, constructing a Laplace matrix according to the diagonal matrix;
s23, calculating a characteristic vector according to the Laplace matrix and obtaining a corresponding element value;
s24, segmenting the spectrogram model according to the element values of the feature vectors to obtain different clusters;
step S4 includes the following steps in order:
s41, the clustered nodes calculate spectrum benefit values according to benefit matrixes in the attribute matrixes;
s42, comparing the clustered nodes through the spectrum benefit values;
s43, if the benefit value obtained by the clustered nodes on the available frequency spectrum is the maximum, obtaining the frequency spectrum;
s44, if the benefit values obtained by the clustered nodes on the available frequency spectrum are equal, the clustered nodes with less bandwidth obtain the frequency spectrum;
and S45, when the spectrum allocation is finished or/and the node spectrum demand is saturated, the spectrum allocation is terminated.
2. The spectrum allocation method according to claim 1, wherein step S3 comprises the following steps in sequence:
s31, the clustering selects a cluster head;
s32, the cluster head stores the member table information of the cluster node;
s33, the clustering node senses frequency spectrum information and sends the frequency spectrum information to a cluster head of a cluster to which the clustering node belongs;
s34, the cluster head sends the received spectrum information to the central controller for managing the cluster;
s35, the central controller obtains clustered spectrum use information according to the spectrum information and feeds the spectrum use information back to a clustered cluster head;
and S36, the node receives the frequency spectrum using information from the cluster head of the cluster to which the node belongs at regular time, and a new node in the node selects the cluster according to the frequency spectrum using information.
3. Method for spectrum allocation according to any of claims 1 or 2, wherein said method further comprises a step S5: and updating the frequency spectrum information between the nodes according to the frequency spectrum allocation result.
4. The spectrum allocation method according to claim 3, wherein step S5 comprises the following steps in sequence:
s51, the node for obtaining the frequency spectrum deletes the allocated available frequency spectrum information of the node;
s52, the connected nodes of the frequency spectrum obtaining node delete the distributed available frequency spectrum information;
s53, deleting nodes without available frequency spectrum;
s54, deleting the nodes with saturated spectrum allocation;
and S55, allocating the frequency spectrum to the independent nodes without the connected nodes until the independent nodes are saturated, and deleting the independent nodes.
5. The utility model provides a spectrum distribution device based on spectrogram is cut apart which characterized in that includes spectrogram model module, spectrogram cut apart module, clustering management module and spectrum distribution module that connect gradually at least:
the spectrum map model module is used for constructing a spectrum map model according to available spectrum information and network topology according to the available spectrum information;
the spectrogram segmentation module is used for carrying out spectrogram segmentation on the spectrogram model and obtaining different clusters;
the clustering management module is used for managing clusters obtained by spectrogram segmentation;
the frequency spectrum allocation module is used for calculating the frequency spectrum benefit value of the clustered nodes and allocating frequency spectrums to the clustered nodes according to the frequency spectrum benefit value;
the frequency spectrum map model module is also used for acquiring available frequency spectrum information, setting an attribute matrix and establishing a frequency spectrum map model according to the attribute matrix; the frequency spectrum graph model is G ═ V, (E) and Q, (V) is a node set, E is an edge set, and Q is a storage available frequency spectrum set;
the spectrogram segmentation module is further used for constructing a diagonal matrix according to the spectrogram model, then constructing a Laplace matrix, and calculating a feature vector and obtaining a corresponding element value according to the Laplace matrix; the spectrum graph model is further divided according to the element values of the feature vectors to obtain different clusters;
the frequency spectrum allocation module is further configured to calculate a frequency spectrum benefit value of the clustered nodes, obtain a clustered node with a maximum benefit value, and allocate the frequency spectrum to the clustered node with the maximum benefit value, or obtain a clustered node with an equal benefit value if the benefit values obtained by the clustered nodes on an available frequency spectrum are equal, and allocate the frequency spectrum to a clustered node with an equal benefit value and a small bandwidth.
6. The spectrum allocation apparatus according to claim 5, wherein the cluster management module includes a central controller, and the cluster management module is further configured to enable the cluster to select a cluster head, where the cluster head stores member table information of the cluster nodes, and further configured to enable the cluster nodes to sense surrounding spectrum information and send the spectrum information to the cluster head of the cluster to which the cluster nodes belong, and the cluster head sends the received spectrum information to the central controller that manages the cluster at regular time; the central controller is used for acquiring clustered spectrum use information according to the received spectrum information sent by the cluster head; the central controller feeds back frequency spectrum use information to the clustered cluster head; and selecting the cluster by a new node in the nodes according to the frequency spectrum use information.
7. The spectrum allocation apparatus according to claim 5, further comprising an update module connected to the spectrum allocation module, wherein the update module is configured to update spectrum information between nodes according to the spectrum allocation result.
8. The spectrum allocation apparatus according to claim 7, wherein the updating module is further configured to delete available spectrum information already allocated to the obtained spectrum node, available spectrum information already allocated to nodes connected to the obtained spectrum node, a node without available spectrum, and a node whose spectrum allocation is saturated; and the method is also used for allocating the frequency spectrum to the independent nodes without the connected nodes until the independent nodes are saturated and deleting the independent nodes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010101073444A CN101808325B (en) | 2010-02-02 | 2010-02-02 | Method and device for allocating frequency spectrum |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010101073444A CN101808325B (en) | 2010-02-02 | 2010-02-02 | Method and device for allocating frequency spectrum |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101808325A CN101808325A (en) | 2010-08-18 |
| CN101808325B true CN101808325B (en) | 2012-11-14 |
Family
ID=42609897
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010101073444A Active CN101808325B (en) | 2010-02-02 | 2010-02-02 | Method and device for allocating frequency spectrum |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101808325B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102595419B (en) * | 2011-01-04 | 2015-07-01 | 中国移动通信集团公司 | Method and system for spectrum allocation in cognitive radio network |
| CN102355730A (en) * | 2011-06-30 | 2012-02-15 | 哈尔滨工业大学 | Frequency spectrum distribution method based on system revenue in cognitive radio |
| CN102685761B (en) * | 2012-05-11 | 2015-05-20 | 中国联合网络通信集团有限公司 | Cellular network architecture implementation method and device |
| CN106330361B (en) * | 2016-11-07 | 2018-11-20 | 桂林电子科技大学 | Weak Signal Detection Method based on connectivity of graph analysis |
| CN109831792B (en) * | 2019-03-11 | 2021-09-24 | 中国科学院上海微系统与信息技术研究所 | Wireless sensor network topology control method based on multi-objective optimization |
| CN110365589B (en) * | 2019-07-30 | 2021-09-28 | 国网福建省电力有限公司 | Electric power optical transmission route and frequency spectrum allocation method based on elastic optical network |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2009119463A2 (en) * | 2008-03-27 | 2009-10-01 | Mitsubishi Electric Corporation | Method for allocating channel resources in orthogonal frequency-division multiple access network |
-
2010
- 2010-02-02 CN CN2010101073444A patent/CN101808325B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2009119463A2 (en) * | 2008-03-27 | 2009-10-01 | Mitsubishi Electric Corporation | Method for allocating channel resources in orthogonal frequency-division multiple access network |
Non-Patent Citations (1)
| Title |
|---|
| 薛钰.认知无线电网络中的频谱分配技术研究.《中国优秀硕士学位论文全文数据库 信息科技辑》.2008,第3.1-3.4、4.1-4.4节. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101808325A (en) | 2010-08-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107333267B (en) | A kind of edge calculations method for 5G super-intensive networking scene | |
| Zhou et al. | Resource sharing and task offloading in IoT fog computing: A contract-learning approach | |
| CN101808325B (en) | Method and device for allocating frequency spectrum | |
| CN107566194B (en) | Method for realizing cross-domain virtual network mapping | |
| CN111586720A (en) | Task unloading and resource allocation combined optimization method in multi-cell scene | |
| CN104619029B (en) | It is a kind of centralization cellular network architecture under baseband pool resource allocation methods and device | |
| CN101568124A (en) | Channel allocation method in wireless mesh network | |
| JP6939260B2 (en) | Wireless communication system, wireless communication method and centralized control station | |
| CN104684095B (en) | Resource allocation methods based on genetic operation in a kind of heterogeneous network converged scene | |
| CN107659973A (en) | Super-intensive network cluster dividing method based on density K means algorithms | |
| CN113596785A (en) | D2D-NOMA communication system resource allocation method based on deep Q network | |
| CN104080091A (en) | Family base station frequency spectrum allocation method based on load prediction grouping in layered heterogenous network | |
| CN115802389A (en) | Federal learning method for training by utilizing digital twin auxiliary model | |
| CN106658526B (en) | Simulated annealing algorithm-based frequency spectrum allocation method in ultra-dense small cellular network | |
| CN117676896A (en) | 6G supported mIoT resource allocation method and system based on reinforcement learning | |
| CN107333301B (en) | Cognitive-based joint resource reallocation method in multi-generic heterogeneous network | |
| CN107148053B (en) | Data configuration method and device | |
| CN103167578B (en) | By the method for Hopfield neural net to Wireless sensor network clustering | |
| Yu et al. | Connected dominating set construction in cognitive radio networks | |
| Nguyen et al. | Utility optimization for blockchain empowered edge computing with deep reinforcement learning | |
| CN107743302A (en) | Rate-allocation and route combined optimization algorithm in wireless sensor network based on compressed sensing | |
| JP2022523082A (en) | Spectrum management device, electronic device, wireless communication method and storage medium | |
| Cui et al. | A clustering approach to wireless scheduling | |
| Cheng et al. | An approximate bandwidth allocation algorithm for tradeoff between fairness and throughput in WSN | |
| Yu et al. | Proximal Policy Optimization-based Federated Client Selection for Internet of Vehicles |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |