CN106658524B - Multi-target spectrum allocation method based on quantum pollination search mechanism in cognitive heterogeneous network - Google Patents

Multi-target spectrum allocation method based on quantum pollination search mechanism in cognitive heterogeneous network Download PDF

Info

Publication number
CN106658524B
CN106658524B CN201610859145.6A CN201610859145A CN106658524B CN 106658524 B CN106658524 B CN 106658524B CN 201610859145 A CN201610859145 A CN 201610859145A CN 106658524 B CN106658524 B CN 106658524B
Authority
CN
China
Prior art keywords
pollen
quantum
channel
spectrum
pollination
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
Application number
CN201610859145.6A
Other languages
Chinese (zh)
Other versions
CN106658524A (en
Inventor
高洪元
梁炎松
杜亚男
刁鸣
李佳
张世铂
苏雪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Harbin Engineering University
Original Assignee
Harbin Engineering University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Harbin Engineering University filed Critical Harbin Engineering University
Priority to CN201610859145.6A priority Critical patent/CN106658524B/en
Publication of CN106658524A publication Critical patent/CN106658524A/en
Application granted granted Critical
Publication of CN106658524B publication Critical patent/CN106658524B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The invention relates to a multi-target spectrum allocation method based on a quantum pollination search mechanism in a cognitive heterogeneous network, which is realized based on a multi-target quantum pollination search mechanism. The invention comprises the following steps: (1) a wireless access network sensing module in a base station senses network information; (2) the network reconfiguration management module divides the frequency spectrum resources into multi-granularity channels; (3) initializing a quantum pollen collection containing P quantum pollen; (4) mapping each pollen individual in the pollen set into a frequency spectrum distribution matrix for correction; (5) setting the conversion probability of cross pollination and self pollination in a quantum pollination search mechanism; (6) mixing the new generation pollen set with the previous generation pollen set; (7) resolving and selecting proper pollen from a Pareto front end and mapping the pollen into a frequency spectrum distribution matrix; (8) and the network reconfiguration management module divides the optimal distribution matrix into blocks. The invention solves the problem of multi-target spectrum allocation and improves the spectrum utilization rate.

Description

Multi-target spectrum allocation method based on quantum pollination search mechanism in cognitive heterogeneous network
Technical Field
The invention relates to a multi-target spectrum allocation method based on a quantum pollination search mechanism in a cognitive heterogeneous network, which is realized based on a multi-target quantum pollination search mechanism.
Background
In order to adapt to different communication environments and increasing user service demands, mobile wireless communication has been subject to successive developments from the first generation to the fourth generation, however, each generation of updates has no obvious limit in time, a long time is required from the putting into use of a new generation communication system to the full replacement of the old generation, and coexistence of users of new and old networks applying different technical standards is inevitable. As can be seen, a heterogeneous wireless network in which multiple access networks coexist has become a trend of future wireless network development. Radio communication inevitably occupies spectrum resources, and radio management departments of various countries adopt a spectrum fixed allocation management mechanism in terms of radio spectrum resource management. With the development of wireless communication technology, the frequency bands of new systems and new technologies are rarely reserved by fixed spectrum allocation policies, and the situation of shortage of spectrum resources is increasingly serious. On the other hand, the utilization of part of the allocated frequency bands is highly unbalanced, some frequency bands are crowded due to busy traffic, and the utilization rate of some frequency bands is very low, which causes the waste of frequency spectrum resources.
The heterogeneous wireless network can fully utilize heterogeneous dynamic management resources of the network so as to improve the spectrum use efficiency, and meanwhile, the cognitive technology enables the network to have a cognitive reconfiguration function, can flexibly change transmission parameters according to a wireless communication environment, transmits and receives information on different frequency bands, and provides possibility for dynamically managing the resources of the cognitive heterogeneous network. In order to support management and optimized utilization of wireless resources, the IEEE1099.4 standard explicitly defines a functional framework and an information interaction flow of a heterogeneous wireless network. As various wireless communication networks are basically and mature arranged, the multiple networks form the cognitive heterogeneous wireless network by sharing the core network, thereby being convenient for dynamically managing wireless resources without changing the original wireless communication technology of each network. The core network and the access network are respectively added with module entities: the network reconfiguration management system comprises a Network Reconfiguration Management (NRM) module, an access network reconfiguration management (RNRM) module, a radio access network perception (RMC) module and an access network reconfiguration control (RRC) module, wherein the RMC module perceives network information and submits the network information to the RNRM, the RNRM module performs information and decision interaction with the NRM to help the NRM manage access networks of different types of networks, the NRM module realizes dynamic resource allocation, and the RRC module realizes access network reconfiguration according to a reconfiguration request of the RNRM by informing allocation results of the RNRM interaction.
Through the search of the prior art documents, P.S. and et al published "Near-optical Dynamic Spectrum Allocation in Cellular Networks" on 2008 IEEE Symposium on New frontiers in Dynamic Spectrum Access Networks ", solved the problem of Dynamic Spectrum Allocation with a greedy algorithm, represented the interference between Access Networks with a collision graph and a physical interference model, and considered channels of different granularities. However, the greedy algorithm makes the best selection at each step, so that the local optimal solution is easy to obtain, and the problem of allocation of rare spectrum resources cannot be effectively solved. "dynamic spectrum allocation based on clone selection algorithm in cognitive heterogeneous network" published in "communication science report" (2012, vol.33, No.7) by shiwa et al proposes a dynamic spectrum allocation method based on clone selection algorithm, and compares it with greedy algorithm, but the convergence of clone selection algorithm is not high. And only the single-target condition is considered, the multi-target spectrum allocation problem cannot be solved, and the application range is not wide.
The cognitive heterogeneous network spectrum allocation needs to consider the access network spectrum demand, interference between access networks and interference constraint conditions between multi-granularity overlapping channels of different access technologies of the heterogeneous network, and is a multi-objective nonlinear constraint 0-1 integer programming problem, so that an optimal solution is difficult to find in a limited time. Therefore, a multi-target quantum flower pollination search mechanism is designed, and a novel method for multi-target spectrum allocation is obtained.
Disclosure of Invention
The invention aims to provide a multi-target spectrum allocation method based on a quantum pollination search mechanism in a cognitive heterogeneous network, which effectively solves the problem of discrete multi-target spectrum allocation in the cognitive heterogeneous network, aiming at overcoming the defects of the spectrum allocation method in the existing cognitive heterogeneous network.
The purpose of the invention is realized as follows:
(1) a wireless access network sensing module in the base station senses network information, predicts a service request, puts forward spectrum requirements and network resource benefits, and sends the information to an access network reconfiguration management module through an information interaction interface of the access network reconfiguration management module; the access network reconfiguration management module collects the frequency spectrum demand information of all base stations managed by the access network reconfiguration management module and provides the information to the network reconfiguration management module through an information interaction interface of the access network reconfiguration management module;
(2) the network reconfiguration management module carries out multi-granularity channel division on the frequency spectrum resources to obtain a channel set, and obtains a base station set, a frequency spectrum demand set and a frequency spectrum resource benefit set according to received information arrangement, and obtains a base station interference matrix and a channel interference matrix through calculation; finally, establishing a cognitive heterogeneous network multi-target spectrum allocation model;
(3) initializing a quantum pollen set containing P quantum pollens, and measuring the quantum pollen set to obtain a pollen set, wherein each pollen in the pollen set represents a possible frequency spectrum allocation result; setting the initial iteration time t to be 0;
(4) mapping each pollen individual in the pollen set to be a frequency spectrum distribution matrix for correction, mapping the corrected frequency spectrum distribution matrix to be pollen, and evaluating the fitness of each corrected pollen; setting 4 fitness functions such as frequency spectrum benefit, frequency spectrum demand satisfaction rate, frequency spectrum use efficiency and frequency spectrum occupation according to different network requirements;
(5) setting the conversion probability u of cross pollination and self pollination in a quantum pollination search mechanism, and updating by adopting a cross pollination mode if rand is less than u and rand is an element [0,1] which meets the uniformly distributed random number for each dimension of quantum pollen; if rand is more than or equal to u, updating by adopting a self-pollination mode; measuring the updated quantum pollen set to obtain a new-generation pollen set, mapping each pollen individual in the new-generation pollen set to be a frequency spectrum distribution matrix for correction, mapping the corrected frequency spectrum distribution matrix to be pollen, and evaluating the fitness of each corrected pollen;
(6) mixing a new-generation pollen set and a previous-generation pollen set, performing non-dominant solution level sorting and congestion degree calculation, performing descending order arrangement on the pollen with the same non-dominant solution level according to the congestion degree value, selecting the pollen with the non-dominant level of 1 and the higher congestion degree, and adding the pollen into an elite pollen set G; when the number of the pollen in the elite pollen set G is more than 2P, carrying out non-dominant solution level sorting and congestion degree calculation on the pollen in the elite pollen set G, carrying out descending order arrangement on the pollen with the same non-dominant solution level according to the congestion degree value, and selecting the top 2P excellent pollen as a new elite pollen set G;
(7) if the maximum iteration number is not reached, t is t +1, and the step 5 is returned to continue; otherwise, stopping iteration, and sequencing the pollen in the obtained elite pollen set by a non-dominant solution; selecting the pollen with the non-dominant solution grade of 1 as a final Pareto front-end solution, selecting proper pollen from the Pareto front-end solution and mapping the pollen to a frequency spectrum distribution matrix A *；
(8) The network reconfiguration management module allocates the optimal distribution matrix A *Partitioning is carried out, and a spectrum resource allocation result is notified to the access network reconfiguration management module through a decision interactive interface of the access network reconfiguration management module; and an access network reconfiguration control module in the base station realizes reconfiguration according to the spectrum allocation result, namely adaptively modifies working parameters through software.
The cognitive heterogeneous network multi-target spectrum allocation model comprises a channel set, a base station set, a spectrum demand set, a spectrum resource benefit set, a base station interference matrix and a channel interference matrix;
the width of the frequency spectrum resource to be distributed is F, and the channel bandwidths supported by the access technologies correspondingly adopted by the K wireless access networks are respectively W kK is 1,2, …, K, the spectrum resources are divided into multi-granularity channels before allocation; corresponding to the kth access network, the number of available channels is M k＝[F/W k]Therein, []Indicating rounding down, the total number of these available channels is
Figure BDA0001122392880000031
Numbering each channel by using an integer M, wherein M is more than or equal to 1 and less than or equal to M, and forming a channel set { phi (phi) } by using the channels corresponding to each access network 1,Φ 2,…,Φ KWhere phi is 1＝{1,2,…,M 1}，
Figure BDA0001122392880000032
Forming base station set by cell base stations participating in spectrum allocation Wherein the content of the first and second substances,
Figure BDA0001122392880000034
indicating the nth base station and the access network belonging to the kth type; obtaining a frequency spectrum demand set B ═ B by a corresponding base station set nN ═ 1,2, …, N } and the set of spectral benefits R ═ { R ═ R nmN is 1,2, …, N, M is 1,2, …, M }, wherein b is nNumber of channels, r, representing the requirement of the nth base station nmIndicating the spectrum resource benefit of the nth base station using the mth channel;
the base station interference matrix is represented as a block matrix
Figure BDA0001122392880000035
Wherein the sub-block matrix Indicating a base station
Figure BDA0001122392880000037
And
Figure BDA0001122392880000038
the interference relationship between the two or more of the two, indicating a base station
Figure BDA00011223928800000310
And
Figure BDA00011223928800000311
simultaneous use of the same channel or overlapping channels may interfere with each other
Figure BDA00011223928800000312
Represent non-interfering; when k is 1＝k 2When the temperature of the water is higher than the set temperature,
Figure BDA00011223928800000313
according to the spectral multiplexing coefficient of the access technical specification, and when j 1＝j 2When the temperature of the water is higher than the set temperature,
Figure BDA00011223928800000314
indicating that each BS cannot be allocated an overlapping channel; when k is 1≠k 2When, if
Figure BDA00011223928800000315
And
Figure BDA00011223928800000316
the distance between them is less than the sum of their radius of coverage, they are considered to interfere with each other,
Figure BDA0001122392880000041
otherwise
Figure BDA0001122392880000042
The channel interference matrix is represented as
Figure BDA0001122392880000043
When m is 1≠m 2Then, if channel m 1And channel m 2With an overlapping portion, then
Figure BDA0001122392880000044
Indicating that the base station uses channel m 1Will be to channel m 2Cause interference, otherwise
Figure BDA0001122392880000045
In the same way, when m 1＝m 2When the temperature of the water is higher than the set temperature,
Figure BDA0001122392880000046
the spectrum allocation matrix is denoted as A N×M＝(a nm) N×MIf, if Get the channel m, then a nm1, otherwise a nm＝0；
Figure BDA0001122392880000048
The obtained channel number is
Figure BDA0001122392880000049
Base stations for which a feasible allocation matrix is required to satisfy mutual interference cannot allocate overlapping or identical channels, i.e.
Figure BDA00011223928800000410
Access network
Figure BDA00011223928800000411
Dividable channel set phi kOf (3) a channel, i.e.
Figure BDA00011223928800000412
And its resulting channel number e n≤min(b n,M k) (ii) a Wherein the content of the first and second substances,
Figure BDA00011223928800000413
an integer in between;
Figure BDA00011223928800000414
an integer in between; indicating whether channel m belongs to set phi for indicating function kDefinition of
Figure BDA00011223928800000416
The fitness evaluation process of each pollen in the quantum flower pollination set is as follows:
measuring the quantum pollen set to obtain an initial pollen set
Figure BDA00011223928800000417
Binary string after measurement of ith quantum pollen in quantum pollen collection Is a pollen, itIndicating one possible spectrum allocation result, wherein,
Figure BDA00011223928800000419
Figure BDA00011223928800000420
is a random number satisfying uniform distribution; setting the initial iteration time t to be 0;
mapping each pollen individual in the pollen set to be a frequency spectrum distribution matrix for correction, and then mapping the corrected frequency spectrum distribution matrix to be pollen, the specific steps are as follows: setting a zero matrix A N×M(ii) a From pollen
Figure BDA00011223928800000421
Intermediate corresponding channel
Figure BDA00011223928800000422
Distributing the state segments to obtain the nth row element of the matrix A; a is subjected to a feasibility check if
Figure BDA00011223928800000423
And
Figure BDA00011223928800000424
in the presence of interference, i.e. I BS(n 1,n 2) 1, then look for the nth 1Line non-zero elements
Figure BDA00011223928800000425
And n is 2Line non-zero elements
Figure BDA00011223928800000426
Judging channel m 1And m 2Whether it is the same channel or an overlapping channel, if I Channel(m 1,m 2) 1, then modify randomly Or Is 0; computing
Figure BDA00011223928800000429
The obtained channel number is
Figure BDA00011223928800000430
If e nIs greater than
Figure BDA00011223928800000431
Required number of channels b nThen choose randomly (e) n-b n) Setting the allocated channels to zero, and finally mapping the corrected allocation matrix into pollen; repeating the steps to correct all the pollen in the pollen set;
evaluating the fitness of each corrected pollen, and performing fitness function according to different network requirements
Figure BDA00011223928800000433
Selected from the following functions:
(4.1) network benefits:
Figure BDA0001122392880000051
wherein A is the ith pollen in the pollen collection
Figure BDA0001122392880000052
A spectrum allocation matrix corresponding to mapping, wherein r is a spectrum benefit matrix, M is the total number of channels, and N is the total number of access networks;
(4.2) RAN Spectrum requirement satisfaction rate: wherein e is nNumber of spectrum finally allocated to the nth access network, b nThe spectrum requirement actually provided for the nth access network, wherein N is the total number of the access networks;
(4.3) spectrum use efficiency:
Figure BDA0001122392880000054
representing the spectrum bandwidth actually utilized by the access network; wherein e is nThe number of frequency spectrums W finally allocated to the nth access network kThe channel granularity of the kth heterogeneous network is obtained, and N is the total number of access networks;
(4.4) Spectrum Occupancy (SO):
Figure BDA0001122392880000055
representing an actual utilization ratio of the spectrum resources; wherein, SU is spectrum utilization efficiency, and F is spectrum resource width to be allocated.
The updating process of each pollen in the quantum flower pollination set is as follows:
setting the conversion probability u of cross pollination and self pollination in quantum pollination search mechanism, and setting the jth dimension of ith quantum pollen of the tth generation
Figure BDA0001122392880000056
If rand is less than u, updating by adopting a cross pollination mode, wherein the quantum rotation angle updating formula is
Figure BDA0001122392880000057
Wherein rand is from [0,1]]Is a random number that satisfies a uniform distribution,
Figure BDA0001122392880000058
is the ith pollen in the t generation pollen collection
Figure BDA0001122392880000059
The j-th dimension value of (1) is more than or equal to i and less than or equal to P, and 1 is more than or equal to j and less than or equal to Q; globally optimal pollen Randomly selecting the first 2% pollen in the elite pollen set,
Figure BDA00011223928800000511
is the value of its j-th dimension; c. C 1Is constant and represents the evolution of the global optimal pollen to the quantum pollen setThe degree of influence; if rand is more than or equal to u, updating by adopting a self-pollination mode, wherein the quantum rotation angle updating formula is
Figure BDA00011223928800000512
Among them, the preferred pollen
Figure BDA00011223928800000513
Randomly selecting pollen in the elite solution set,
Figure BDA00011223928800000514
is the value of its j-th dimension; random pollen
Figure BDA00011223928800000515
The selection is made at random from a collection of pollen,
Figure BDA00011223928800000516
is the value of its j-th dimension; c. C 2And c 3Is a constant which respectively represents the influence degree of the better pollen in the elite pollen set and the random pollen in the pollen set on the evolution of the quantum pollen set;
dimension j for ith quantum pollen of t generation
Figure BDA00011223928800000517
Using quantum rotary gate for updating, the formula is Wherein the content of the first and second substances,
Figure BDA00011223928800000519
is a random number satisfying uniform distribution, abs (. cndot.) represents an absolute value, and c is a variation probability at a quantum rotation angle of 0, whose value is [0,1/Q ]]A constant value of (1);
Figure BDA00011223928800000520
is a quantum not-gate (QNOT-gate), is rotated by quantumA gate, t +1 and t being the number of iterations; and measuring the updated quantum pollen set to obtain a new-generation pollen set, mapping each pollen individual in the new-generation pollen set to be a frequency spectrum distribution matrix for correction, mapping the corrected frequency spectrum distribution matrix to be pollen, and evaluating the fitness of each corrected pollen.
The non-dominated solution sorting and congestion degree calculation process is as follows:
for the maximum multi-objective optimization problem, f is set z(x i) (z is 1,2, …, w) is pollen x iW is the number of selected targets; for pollen x iAnd x gIf the fitness value f z(x i)≥f z(x g) If all the targets are satisfied and at least one strict inequality is satisfied, then the pollen x is called iDominating pollen x g，x iNon-dominant pollen; if the fitness value f z(x i)≤f z(x g) If all the targets are satisfied and at least one strict inequality is satisfied, then the pollen x is called gDominating pollen x i，x gNon-dominant pollen; if the above two conditions are not satisfied, pollen x iAnd x gNo dominance relationship;
the process of non-dominant pollen ranking is as follows: for each pollen x in the pollen collection iCalculating dominant pollen x iNumber of pollen of
Figure BDA0001122392880000061
And pollen x iSet of dominated pollen
Figure BDA0001122392880000062
Dominating pollen x iNumber of (2)
Figure BDA0001122392880000063
Pollen x iHas a non-dominated ranking of 1; pollen x with rank 1 for each non-dominant ranking iTraverse the collection of pollen it dominates
Figure BDA0001122392880000064
Each pollen x in (1) gCalculating dominant pollen x gNumber of If it is
Figure BDA0001122392880000066
Then the pollen x gPutting the pollen into a set H, and sequencing the non-dominant grade of the pollen in the set to be 2; repeating the above process for each pollen in set H, thereby obtaining a pollen set with a non-dominant ranking of 3, and repeating the process until all pollen non-dominant rankings are obtained;
the congestion degree calculation is carried out on the pollen with the same non-dominant grade, the pollen with the same non-dominant grade is sorted from small to large according to the fitness value of a certain target, the congestion degree value of the pollen with the maximum and minimum fitness values is infinity, and the congestion degree values of other pollen are the difference of the fitness values of two adjacent pollen divided by the difference of the maximum and minimum fitness values; the above calculation is performed for all targets, and the final congestion value is the sum of the congestion values calculated for all targets.
Compared with the prior art, the invention has the following advantages:
(1) the invention considers the spectrum demand of the access network in the cognitive heterogeneous network, the interference between the access networks and the interference constraint condition between multi-granularity overlapping channels of different access technologies of the heterogeneous network; meanwhile, a base station interference matrix, a channel interference matrix and a spectrum allocation matrix are established, and a multi-target spectrum allocation model of the cognitive heterogeneous network is established;
(2) the method solves the problem of multi-target spectrum allocation, a new method based on a quantum pollination search mechanism is used as an evolution strategy, non-dominated solution sequencing and congestion degree calculation are combined, so that Pareto front-end solutions which are uniformly distributed are obtained, and a proper spectrum allocation result can be selected according to different network requirements; the designed method can increase the network spectrum benefit and improve the spectrum utilization rate;
(3) compared with the existing spectrum allocation method, the method can solve the problem of single-target optimization, namely only one target of spectrum benefit, spectrum demand satisfaction rate, spectrum use efficiency, spectrum occupation and the like is considered; multi-objective optimization can be solved, and the frequency spectrum benefit and the frequency spectrum demand satisfaction rate are considered at the same time, so that the method is wider in applicability;
(4) simulation results show that the multi-target spectrum allocation method based on the quantum pollination search mechanism can obtain more excellent performance than a single-target spectrum allocation technology based on a quantum genetic algorithm, and the effectiveness of the method is demonstrated.
Drawings
FIG. 1 is a flow chart of a specific implementation of multi-target spectrum allocation based on a quantum pollination search mechanism in a cognitive heterogeneous network;
FIG. 2 is a flow chart of a multi-target spectrum allocation method based on a quantum pollination search mechanism in a cognitive heterogeneous network;
FIG. 3 is a schematic diagram of a cognitive heterogeneous network dynamic resource management system model;
FIG. 4 is a diagram of multi-granularity overlapping channel partitioning of spectral resources;
fig. 5 is a simulation result of the target network benefit and RAN spectrum demand satisfaction rate when the allocated spectrum resource bandwidth (F) is 10 MHz;
fig. 6 is a simulation result of target network efficiency and spectrum use efficiency when the spectrum resource bandwidth (F) to be allocated is 10 MHz;
fig. 7 is a simulation result of the target RAN spectrum demand satisfaction frequency and spectrum usage efficiency when the allocated spectrum resource bandwidth (F) is 10 MHz.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
The method is realized by a quantitative pollination search mechanism, non-dominated solution sorting and congestion degree calculation are combined, so that a Pareto front-end solution set which is uniformly distributed is obtained, and then a system selects a proper spectrum allocation scheme from the Pareto optimal solution set according to weights of different targets to complete spectrum allocation.
The invention is realized by the following technical scheme, which mainly comprises the following steps:
step 1: considering that K different wireless access networks participate in spectrum allocation, the wireless access networks are respectively allocated by RNRMs kAnd (K-1, 2, …, K). The number of cell base stations which provide frequency spectrum requirements in the kth access network is N kThen, then
Figure BDA0001122392880000071
Representing the total number of base stations that make up the spectrum demand. The RMC module in the system senses network information, predicts service requests, puts forward spectrum requirements and network resource benefits, and transmits the information
Figure BDA0001122392880000073
By RNRM kSends the information interaction interface to RNRM k；RNRM k(K-1, 2, …, K) collects the spectrum demand information of all the base stations it manages and passes this information on
Figure BDA0001122392880000074
The information is provided to the NRM through an information interaction interface with the NRM; wherein the content of the first and second substances,
Figure BDA0001122392880000075
indicating the nth base station and its belonging to the kth access network,
Figure BDA0001122392880000076
indicates the number of channels required by the nth base station belonging to the kth access network,
Figure BDA0001122392880000077
indicating the spectrum resource benefit of using the mth channel by the nth base station belonging to the kth access network.
Step 2: NRM carries out multi-granularity channel division on frequency spectrum resources to obtain a channel set, and obtains a base station set, a frequency spectrum demand set and a frequency spectrum resource benefit set according to received information arrangement, and obtains a base station interference matrix and a channel interference matrix through calculation; and finally, establishing a cognitive heterogeneous network multi-target spectrum allocation model.
The width of the frequency spectrum resource to be distributed is F, and the channel bandwidths supported by the access technologies correspondingly adopted by the K wireless access networks are respectively W k(K ═ 1,2, …, K), the spectrum resources are divided into multi-granularity channels before allocation. Corresponding to the kth access network, the number of available channels is M k＝[F/W k]Therein, []Indicating rounding down, the total number of these available channels is
Figure BDA0001122392880000081
Numbering each channel by an integer M (M is more than or equal to 1 and less than or equal to M), and forming a channel set { phi (phi) } by the channel corresponding to each access network 1,Φ 2,…,Φ KWhere phi is 1＝{1,2,…,M 1}，
Figure BDA0001122392880000082
Forming base station set by cell base stations participating in spectrum allocation
Figure BDA0001122392880000083
Wherein the content of the first and second substances,
Figure BDA0001122392880000084
indicating the nth base station and its belonging to the kth access network. Obtaining a frequency spectrum demand set B ═ B by a corresponding base station set nN ═ 1,2, …, N } and the set of spectral benefits R ═ { R ═ R nmN is 1,2, …, N, M is 1,2, …, M }, wherein b is nNumber of channels, r, representing the requirement of the nth base station nmIndicating the spectrum resource benefit of using the mth channel by the nth base station.
The base station interference matrix is represented as a block matrix Wherein the sub-block matrix Indicating a base station And
Figure BDA0001122392880000088
the interference relationship between the two or more of the two,
Figure BDA0001122392880000089
indicating a base station
Figure BDA00011223928800000810
And
Figure BDA00011223928800000811
simultaneous use of the same channel or overlapping channels may interfere with each other
Figure BDA00011223928800000812
Represent non-interfering; when k is 1＝k 2When the temperature of the water is higher than the set temperature,
Figure BDA00011223928800000813
according to the spectral multiplexing coefficient of the access technical specification, and when j 1＝j 2When the temperature of the water is higher than the set temperature,
Figure BDA00011223928800000814
indicating that each BS cannot be allocated an overlapping channel; when k is 1≠k 2When, if
Figure BDA00011223928800000815
And
Figure BDA00011223928800000816
the distance between them is less than the sum of their radius of coverage, they are considered to interfere with each other,
Figure BDA00011223928800000817
otherwise
Figure BDA00011223928800000818
The channel interference matrix is represented as When m is 1≠m 2Then, if channel m 1And channel m 2With an overlapping portion, then
Figure BDA00011223928800000820
Indicating that the base station uses channel m 1Will be to channel m 2Cause interference, otherwise
Figure BDA00011223928800000821
In the same way, when m 1＝m 2When the temperature of the water is higher than the set temperature,
the spectrum allocation matrix is denoted as A N×M＝(a nm) N×MIf, if
Figure BDA00011223928800000823
Get the channel m, then a nm1, otherwise a nm＝0； The obtained channel number is
Figure BDA00011223928800000825
Base stations for which a feasible allocation matrix is required to satisfy mutual interference cannot allocate overlapping or identical channels, i.e.
Figure BDA00011223928800000826
Access network
Figure BDA00011223928800000827
Dividable channel set phi kOf (3) a channel, i.e.
Figure BDA00011223928800000828
And its resulting channel number e n≤min(b n,M k). Wherein the content of the first and second substances,
Figure BDA00011223928800000829
an integer in between;
Figure BDA00011223928800000830
an integer in between; indicating whether channel m belongs to set phi for indicating function kDefinition of
Figure BDA0001122392880000091
And step 3: initializing a quantum pollen set, and measuring the quantum pollen set to obtain a pollen set. By D n＝M kTo represent
Figure BDA0001122392880000092
The maximum number of channels that can be obtained, then in each pollen, with a length D nBinary string representation of Whether a corresponding channel is obtained. The coding form of the ith quantum pollen in the quantum pollen collection is
Figure BDA0001122392880000094
Wherein the content of the first and second substances,
Figure BDA0001122392880000095
representation pair qubit
Figure BDA0001122392880000096
The probability of 0 being obtained at the time of measurement,
Figure BDA0001122392880000097
representation pair qubit
Figure BDA0001122392880000098
The probability of 1 is obtained during measurement, at the same time
Figure BDA0001122392880000099
And
Figure BDA00011223928800000910
need to satisfy normalization conditions
Figure BDA00011223928800000911
All qubits in Quantum pollen are initialized to
Figure BDA00011223928800000912
P is the number of quantum pollen contained in the quantum pollen collection,
Figure BDA00011223928800000913
is the number of qubits contained in each quantum pollen.
Measuring the quantum pollen set to obtain an initial pollen set Binary string after measurement of ith quantum pollen in quantum pollen collection
Figure BDA00011223928800000915
Is a pollen that represents one possible spectrum allocation result, wherein,
Figure BDA00011223928800000917
is a random number satisfying uniform distribution. And setting the initial iteration number t to be 0.
And 4, step 4: mapping each pollen individual in the pollen set to be a frequency spectrum distribution matrix for correction, and then mapping the corrected frequency spectrum distribution matrix to be pollen, the specific steps are as follows: setting a zero matrix A N×M(ii) a From pollen
Figure BDA00011223928800000918
Intermediate corresponding channel Distributing the state segments to obtain the nth row element of the matrix A; a is subjected to a feasibility check if And
Figure BDA00011223928800000921
in the presence of interference, i.e. I BS(n 1,n 2) 1, then look for the nth 1Line non-zero elements
Figure BDA00011223928800000922
And n is 2Line non-zero elements
Figure BDA00011223928800000923
Judging channel m 1And m 2Whether it is the same channel or an overlapping channel, if I Channel(m 1,m 2) 1, then modify randomly
Figure BDA00011223928800000924
Or
Figure BDA00011223928800000925
Is 0; computing
Figure BDA00011223928800000926
The obtained channel number is
Figure BDA00011223928800000927
If e nIs greater than
Figure BDA00011223928800000928
Required number of channels b nThen choose randomly (e) n-b n) Setting the allocated channels to zero, and finally mapping the corrected allocation matrix into pollen; repeating the above steps to correct all pollen in the pollen collection.
Evaluating the fitness of each modified pollen, and performing fitness function according to different network requirementsNumber of
Figure BDA00011223928800000929
Can be chosen among the following functions:
(1) network benefit (R, revenue):
Figure BDA00011223928800000930
wherein A is the ith pollen in the pollen collection
Figure BDA0001122392880000101
A spectrum allocation matrix corresponding to mapping, wherein r is a spectrum benefit matrix, M is the total number of channels, and N is the total number of access networks;
(2) RAN spectrum demand satisfaction rate (RS, RAN satisfactory):
Figure BDA0001122392880000102
wherein e is nNumber of spectrum finally allocated to the nth access network, b nThe spectrum requirement actually provided for the nth access network, wherein N is the total number of the access networks;
(3) spectrum utilization efficiency (SU):
Figure BDA0001122392880000103
representing the spectrum bandwidth actually utilized by the access network; wherein e is nThe number of frequency spectrums W finally allocated to the nth access network kThe channel granularity of the kth heterogeneous network is obtained, and N is the total number of access networks;
(4) spectrum Occupancy (SO):
Figure BDA0001122392880000104
representing an actual utilization ratio of the spectrum resources; wherein, SU is spectrum utilization efficiency, and F is spectrum resource width to be allocated.
And 5: setting the conversion probability u of cross pollination and self pollination in quantum pollination search mechanism, and setting the jth dimension of ith quantum pollen of the tth generation
Figure BDA0001122392880000105
If rand is less than u, updating by adopting a cross pollination mode, wherein the quantum rotation angle updating formula is
Figure BDA0001122392880000106
Wherein rand is from [0,1]]Is a random number that satisfies a uniform distribution,
Figure BDA0001122392880000107
is the ith pollen in the t generation pollen collection The j-th dimension value of (1) is more than or equal to i and less than or equal to P, and 1 is more than or equal to j and less than or equal to Q; globally optimal pollen
Figure BDA0001122392880000109
Randomly selecting the first 2% pollen in the elite pollen set, is the value of its j-th dimension; c. C 1The constant represents the influence degree of the globally optimal pollen on the evolution of the quantum pollen set; if rand is more than or equal to u, updating by adopting a self-pollination mode, wherein the quantum rotation angle updating formula is
Figure BDA00011223928800001011
Among them, the preferred pollen
Figure BDA00011223928800001012
Randomly selecting pollen in the elite solution set,
Figure BDA00011223928800001013
is the value of its j-th dimension; random pollen The selection is made at random from a collection of pollen, is the value of its j-th dimension; c. C 2And c 3Is a constant, which respectively represents the superior pollen in the elite pollen set andinfluence degree of random pollen in the pollen collection on quantum pollen collection evolution.
Dimension j for ith quantum pollen of t generation
Figure BDA00011223928800001016
Using quantum rotary gate for updating, the formula is
Figure BDA00011223928800001017
Wherein the content of the first and second substances, is a random number satisfying uniform distribution, abs (. cndot.) represents an absolute value, and c is a variation probability at a quantum rotation angle of 0, whose value is [0,1/Q ]]A constant value of (1);
Figure BDA00011223928800001019
is a quantum not-gate (QNOT-gate),
Figure BDA00011223928800001020
is a quantum revolving gate, and t +1 and t are iteration times; and measuring the updated quantum pollen set to obtain a new-generation pollen set, mapping each pollen individual in the new-generation pollen set to be a frequency spectrum distribution matrix for correction, mapping the corrected frequency spectrum distribution matrix to be pollen, and evaluating the fitness of each corrected pollen.
Step 6: mixing the new generation pollen set and the previous generation pollen set, performing non-dominant solution level sorting and congestion degree calculation, performing descending order arrangement on the pollen with the same non-dominant solution level according to the congestion degree value, selecting the pollen with the non-dominant level of 1 and the larger congestion degree, and adding the pollen into the elite pollen set G. And when the number of the pollen in the elite pollen set G is more than 2P, carrying out non-dominant solution level sorting and congestion degree calculation on the pollen in the elite pollen set G, carrying out descending order arrangement on the pollen with the same non-dominant solution level according to the congestion degree value, and selecting the top 2P excellent pollen as a new elite pollen set G.
For the maximum multi-objective optimization problem, f is set z(x i)(z＝1,2, …, w) is pollen x iW is the number of selected targets; for pollen x iAnd x gIf the fitness value f z(x i)≥f z(x g) If all the targets are satisfied and at least one strict inequality is satisfied, then the pollen x is called iDominating pollen x g，x iNon-dominant pollen; if the fitness value f z(x i)≤f z(x g) If all the targets are satisfied and at least one strict inequality is satisfied, then the pollen x is called gDominating pollen x i，x gNon-dominant pollen; if the above two conditions are not satisfied, pollen x iAnd x gWithout any dominating relationship.
The process of non-dominant pollen ranking is as follows: for each pollen x in the pollen collection iCalculating dominant pollen x iNumber of pollen of
Figure BDA0001122392880000111
And pollen x iSet of dominated pollen Dominating pollen x iNumber of (2)
Figure BDA0001122392880000113
Pollen x iHas a non-dominated ranking of 1; pollen x with rank 1 for each non-dominant ranking iTraverse the collection of pollen it dominates
Figure BDA0001122392880000114
Each pollen x in (1) gCalculating dominant pollen x gNumber of If it is
Figure BDA0001122392880000116
Then the pollen x gPut into the collection H and non-dominating pollen in the collectionThe rank ordering is 2; the above process is repeated for each pollen in set H, whereby a set of pollen with non-dominant ranking ranked 3 is obtained, and the process is repeated until non-dominant ranking of all pollen is obtained.
The congestion degree calculation is carried out on the pollen with the same non-dominant grade, the pollen with the same non-dominant grade is sorted from small to large according to the fitness value of a certain target, the congestion degree value of the pollen with the maximum and minimum fitness values is infinity, and the congestion degree values of other pollen are the difference of the fitness values of two adjacent pollen divided by the difference of the maximum and minimum fitness values; the above calculation is performed for all targets, and the final congestion value is the sum of the congestion values calculated for all targets. According to the calculation process, in order to ensure that a uniform Pareto front end solution set is obtained, the pollen with the non-dominant grade of 1 and the high crowding degree needs to be evolved. Sorting the pollen with the same non-dominant grade according to the congestion value from big to small, selecting the pollen with the non-dominant grade of 1, adding the pollen into the elite pollen set G as elite pollen, and setting the number of the elite pollen in the elite pollen set G to be 2P.
And 7: if the maximum iteration number is not reached, t is t +1, and the step 5 is returned to continue; otherwise, stopping iteration, and sequencing the pollen in the obtained elite pollen set by a non-dominant solution; selecting the pollen with the non-dominant solution grade of 1 as a final Pareto front-end solution, selecting proper pollen from the Pareto front-end solution and mapping the pollen to a frequency spectrum distribution matrix A *。
And 8: NRM module assigns optimal matrix A *The blocking is carried out as shown in the following formula,
Figure BDA0001122392880000121
by reaction with RNRM kDecision interactive interface of (2) allocating spectrum resources
Figure BDA0001122392880000122
Notification RNRM k(k＝1,2,…,K)；
Figure BDA0001122392880000123
The RRC module in (1) results from the spectrum allocationThe operating parameters are now reconstructed, i.e. adaptively modified by software.
By combining a heterogeneous wireless network system model, the specific implementation process of the multi-target spectrum allocation based on the quantum pollination search mechanism in the cognitive heterogeneous network comprises the following steps:
step 1: considering that K different wireless access networks participate in spectrum allocation, the wireless access networks are respectively allocated by RNRMs kAnd (K-1, 2, …, K). The number of cell base stations which provide frequency spectrum requirements in the kth access network is N kThen, then
Figure BDA0001122392880000124
Representing the total number of base stations that make up the spectrum demand. The RMC module in the system senses network information, predicts service requests, puts forward spectrum requirements and network resource benefits, and transmits the information
Figure BDA0001122392880000126
By RNRM kSends the information interaction interface to RNRM k；RNRM k(K-1, 2, …, K) collects the spectrum demand information of all the base stations it manages and passes this information on
Figure BDA0001122392880000127
The information is provided to the NRM through an information interaction interface with the NRM; wherein the content of the first and second substances,
Figure BDA0001122392880000128
indicating the nth base station and its belonging to the kth access network, indicates the number of channels required by the nth base station belonging to the kth access network, indicating the spectrum resource benefit of using the mth channel by the nth base station belonging to the kth access network.
Step 2: NRM carries out multi-granularity channel division on frequency spectrum resources to obtain a channel set, and obtains a base station set, a frequency spectrum demand set and a frequency spectrum resource benefit set according to received information arrangement, and obtains a base station interference matrix and a channel interference matrix through calculation; and finally, establishing a cognitive heterogeneous network multi-target spectrum allocation model.
The width of the frequency spectrum resource to be distributed is F, and the channel bandwidths supported by the access technologies correspondingly adopted by the K wireless access networks are respectively W k(K ═ 1,2, …, K), the spectrum resources are divided into multi-granularity channels before allocation. Corresponding to the kth access network, the number of available channels is M k＝[F/W k]Therein, []Indicating rounding down, the total number of these available channels is Numbering each channel by an integer M (M is more than or equal to 1 and less than or equal to M), and forming a channel set { phi (phi) } by the channel corresponding to each access network 1,Φ 2,…,Φ KWhere phi is 1＝{1,2,…,M 1}，
Forming base station set by cell base stations participating in spectrum allocation
Figure BDA00011223928800001213
Wherein the content of the first and second substances,
Figure BDA00011223928800001214
indicating the nth base station and its belonging to the kth access network. Obtaining a frequency spectrum demand set B ═ B by a corresponding base station set nN ═ 1,2, …, N } and the set of spectral benefits R ═ { R ═ R nmN is 1,2, …, N, M is 1,2, …, M }, wherein b is nNumber of channels, r, representing the requirement of the nth base station nmIndicating the spectrum resource benefit of using the mth channel by the nth base station.
The base station interference matrix is represented as a block matrix
Figure BDA0001122392880000131
WhereinMatrix of subblocks
Figure BDA0001122392880000132
Indicating a base station
Figure BDA0001122392880000133
And
Figure BDA0001122392880000134
the interference relationship between the two or more of the two,
Figure BDA0001122392880000135
indicating a base station
Figure BDA0001122392880000136
And simultaneous use of the same channel or overlapping channels may interfere with each other
Figure BDA0001122392880000138
Represent non-interfering; when k is 1＝k 2When the temperature of the water is higher than the set temperature,
Figure BDA0001122392880000139
according to the spectral multiplexing coefficient of the access technical specification, and when j 1＝j 2When the temperature of the water is higher than the set temperature,
Figure BDA00011223928800001310
indicating that each BS cannot be allocated an overlapping channel; when k is 1≠k 2When, if
Figure BDA00011223928800001311
And
Figure BDA00011223928800001312
the distance between them is less than the sum of their radius of coverage, they are considered to interfere with each other,
Figure BDA00011223928800001313
otherwise
Figure BDA00011223928800001314
The channel interference matrix is represented as
Figure BDA00011223928800001315
When m is 1≠m 2Then, if channel m 1And channel m 2With an overlapping portion, then
Figure BDA00011223928800001316
Indicating that the base station uses channel m 1Will be to channel m 2Cause interference, otherwise
Figure BDA00011223928800001317
In the same way, when m 1＝m 2When the temperature of the water is higher than the set temperature,
Figure BDA00011223928800001318
the spectrum allocation matrix is denoted as A N×M＝(a nm) N×MIf, if Get the channel m, then a nm1, otherwise a nm＝0；
Figure BDA00011223928800001320
The obtained channel number is
Figure BDA00011223928800001321
Base stations for which a feasible allocation matrix is required to satisfy mutual interference cannot allocate overlapping or identical channels, i.e.
Figure BDA00011223928800001322
Access network
Figure BDA00011223928800001323
Dividable channel set phi kOf (3) a channel, i.e. And its resulting channel number e n≤min(b n,M k). Wherein the content of the first and second substances,
Figure BDA00011223928800001325
an integer in between;
Figure BDA00011223928800001326
an integer in between;
Figure BDA00011223928800001327
indicating whether channel m belongs to set phi for indicating function kDefinition of
Figure BDA00011223928800001328
And step 3: initializing a quantum pollen set, and measuring the quantum pollen set to obtain a pollen set. By D n＝M kTo represent
Figure BDA00011223928800001329
The maximum number of channels that can be obtained, then in each pollen, with a length D nBinary string representation of Whether a corresponding channel is obtained. The coding form of the ith quantum pollen in the quantum pollen collection is Wherein the content of the first and second substances,
Figure BDA00011223928800001332
representation pair qubit
Figure BDA00011223928800001333
The probability of 0 being obtained at the time of measurement,
Figure BDA00011223928800001334
representation pair qubit
Figure BDA00011223928800001335
The probability of 1 is obtained during measurement, at the same time And
Figure BDA00011223928800001337
need to satisfy normalization conditions
Figure BDA00011223928800001338
All qubits in Quantum pollen are initialized to
Figure BDA0001122392880000141
P is the number of quantum pollen contained in the quantum pollen collection,
Figure BDA0001122392880000142
is the number of qubits contained in each quantum pollen.
Measuring the quantum pollen set to obtain an initial pollen set
Figure BDA0001122392880000143
Binary string after measurement of ith quantum pollen in quantum pollen collection
Figure BDA0001122392880000144
Is a pollen that represents one possible spectrum allocation result, wherein, is a random number satisfying uniform distribution. And setting the initial iteration number t to be 0.
And 4, step 4: mapping each pollen individual in the pollen set to be a frequency spectrum distribution matrix for correction, and then mapping the corrected frequency spectrum distribution matrix to be pollen, the specific steps are as follows: setting a zero matrix A N×M(ii) a From pollen
Figure BDA0001122392880000147
Intermediate corresponding channel
Figure BDA0001122392880000148
Distributing the state segments to obtain the nth row element of the matrix A; a is subjected to a feasibility check if
Figure BDA0001122392880000149
And
Figure BDA00011223928800001410
in the presence of interference, i.e. I BS(n 1,n 2) 1, then look for the nth 1Line non-zero elements
Figure BDA00011223928800001411
And n is 2Line non-zero elements
Figure BDA00011223928800001412
Judging channel m 1And m 2Whether it is the same channel or an overlapping channel, if I Channel(m 1,m 2) 1, then modify randomly
Figure BDA00011223928800001413
Or
Figure BDA00011223928800001414
Is 0; computing
Figure BDA00011223928800001415
The obtained channel number is
Figure BDA00011223928800001416
If e nIs greater than
Figure BDA00011223928800001417
Required number of channels b nThen choose randomly (e) n-b n) Setting the allocated channels to zero, and finally mapping the corrected allocation matrix into pollen; repeat the aboveThe method comprises the following steps of correcting all pollen in the pollen set.
Evaluating the fitness of each corrected pollen, and performing fitness function according to different network requirements Can be chosen among the following functions:
(1) network benefit (R, revenue):
Figure BDA00011223928800001419
wherein A is the ith pollen in the pollen collection
Figure BDA00011223928800001420
A spectrum allocation matrix corresponding to mapping, wherein r is a spectrum benefit matrix, M is the total number of channels, and N is the total number of access networks;
(2) RAN spectrum demand satisfaction rate (RS, RAN satisfactory): wherein e is nNumber of spectrum finally allocated to the nth access network, b nThe spectrum requirement actually provided for the nth access network, wherein N is the total number of the access networks;
(3) spectrum utilization efficiency (SU): representing the spectrum bandwidth actually utilized by the access network; wherein e is nThe number of frequency spectrums W finally allocated to the nth access network kThe channel granularity of the kth heterogeneous network is obtained, and N is the total number of access networks;
(4) spectrum Occupancy (SO):
Figure BDA0001122392880000151
representing an actual utilization ratio of the spectrum resources; wherein, SU is spectrum utilization efficiency, and F is spectrum resource width to be allocated.
And 5: setting the transformation outline of cross pollination and self pollination in quantum pollination search mechanismRate u, j dimension for ith quantum pollen of t generation
Figure BDA0001122392880000152
If rand is less than u, updating by adopting a cross pollination mode, wherein the quantum rotation angle updating formula is
Figure BDA0001122392880000153
Wherein rand is from [0,1]]Is a random number that satisfies a uniform distribution, is the ith pollen in the t generation pollen collection
Figure BDA0001122392880000155
The j-th dimension value of (1) is more than or equal to i and less than or equal to P, and 1 is more than or equal to j and less than or equal to Q; globally optimal pollen
Figure BDA0001122392880000156
Randomly selecting the first 2% pollen in the elite pollen set,
Figure BDA0001122392880000157
is the value of its j-th dimension; c. C 1The constant represents the influence degree of the globally optimal pollen on the evolution of the quantum pollen set; if rand is more than or equal to u, updating by adopting a self-pollination mode, wherein the quantum rotation angle updating formula is
Figure BDA0001122392880000158
Among them, the preferred pollen
Figure BDA0001122392880000159
Randomly selecting pollen in the elite solution set,
Figure BDA00011223928800001510
is the value of its j-th dimension; random pollen
Figure BDA00011223928800001511
The selection is made at random from a collection of pollen,
Figure BDA00011223928800001512
is the value of its j-th dimension; c. C 2And c 3Is a constant which respectively represents the influence degree of the superior pollen in the elite pollen set and the random pollen in the pollen set on the evolution of the quantum pollen set.
Dimension j for ith quantum pollen of t generation
Figure BDA00011223928800001513
Using quantum rotary gate for updating, the formula is
Figure BDA00011223928800001514
Wherein the content of the first and second substances,
Figure BDA00011223928800001515
is a random number satisfying uniform distribution, abs (. cndot.) represents an absolute value, and c is a variation probability at a quantum rotation angle of 0, whose value is [0,1/Q ]]A constant value of (1);
Figure BDA00011223928800001516
is a quantum not-gate (QNOT-gate),
Figure BDA00011223928800001517
is a quantum revolving gate, and t +1 and t are iteration times; and measuring the updated quantum pollen set to obtain a new-generation pollen set, mapping each pollen individual in the new-generation pollen set to be a frequency spectrum distribution matrix for correction, mapping the corrected frequency spectrum distribution matrix to be pollen, and evaluating the fitness of each corrected pollen.
Step 6: mixing the new generation pollen set and the previous generation pollen set, performing non-dominant solution level sorting and congestion degree calculation, performing descending order arrangement on the pollen with the same non-dominant solution level according to the congestion degree value, selecting the pollen with the non-dominant level of 1 and the larger congestion degree, and adding the pollen into the elite pollen set G. And when the number of the pollen in the elite pollen set G is more than 2P, carrying out non-dominant solution level sorting and congestion degree calculation on the pollen in the elite pollen set G, carrying out descending order arrangement on the pollen with the same non-dominant solution level according to the congestion degree value, and selecting the top 2P excellent pollen as a new elite pollen set G.
For the maximum multi-objective optimization problem, f is set z(x i) (z is 1,2, …, w) is pollen x iW is the number of selected targets; for pollen x iAnd x gIf the fitness value f z(x i)≥f z(x g) If all the targets are satisfied and at least one strict inequality is satisfied, then the pollen x is called iDominating pollen x g，x iNon-dominant pollen; if the fitness value f z(x i)≤f z(x g) If all the targets are satisfied and at least one strict inequality is satisfied, then the pollen x is called gDominating pollen x i，x gNon-dominant pollen; if the above two conditions are not satisfied, pollen x iAnd x gWithout any dominating relationship.
The process of non-dominant pollen ranking is as follows: for each pollen x in the pollen collection iCalculating dominant pollen x iNumber of pollen of
Figure BDA0001122392880000161
And pollen x iSet of dominated pollen Dominating pollen x iNumber of (2)
Figure BDA0001122392880000163
Pollen x iHas a non-dominated ranking of 1; pollen x with rank 1 for each non-dominant ranking iTraverse the collection of pollen it dominates Each pollen x in (1) gCalculating dominant pollen x gNumber of If it is Then the pollen x gPutting the pollen into a set H, and sequencing the non-dominant grade of the pollen in the set to be 2; the above process is repeated for each pollen in set H, whereby a set of pollen with non-dominant ranking ranked 3 is obtained, and the process is repeated until non-dominant ranking of all pollen is obtained.
The congestion degree calculation is carried out on the pollen with the same non-dominant grade, the pollen with the same non-dominant grade is sorted from small to large according to the fitness value of a certain target, the congestion degree value of the pollen with the maximum and minimum fitness values is infinity, and the congestion degree values of other pollen are the difference of the fitness values of two adjacent pollen divided by the difference of the maximum and minimum fitness values; the above calculation is performed for all targets, and the final congestion value is the sum of the congestion values calculated for all targets. According to the calculation process, in order to ensure that a uniform Pareto front end solution set is obtained, the pollen with the non-dominant grade of 1 and the high crowding degree needs to be evolved. Sorting the pollen with the same non-dominant grade according to the congestion value from big to small, selecting the pollen with the non-dominant grade of 1, adding the pollen into the elite pollen set G as elite pollen, and setting the number of the elite pollen in the elite pollen set G to be 2P.
And 7: if the maximum iteration number is not reached, t is t +1, and the step 5 is returned to continue; otherwise, stopping iteration, and sequencing the pollen in the obtained elite pollen set by a non-dominant solution; selecting the pollen with the non-dominant solution grade of 1 as a final Pareto front-end solution, selecting proper pollen from the Pareto front-end solution and mapping the pollen to a frequency spectrum distribution matrix A *。
And 8: NRM module assigns optimal matrix A *The blocking is carried out as shown in the following formula,
Figure BDA0001122392880000167
by reaction with RNRM kDecision interactive interface of (2) allocating spectrum resources
Figure BDA0001122392880000168
Notification RNRM k(k＝1,2,…,K)；
Figure BDA0001122392880000169
The RRC module in (1) performs reconfiguration according to the spectrum allocation result, i.e. adaptively modifies the operating parameters through software.
In fig. 7, the cognitive heterogeneous wireless network simulation scenario adopts 3 networks such as GSM, CDMA2000, WCDMA, etc., corresponding channel bandwidths are 0.5MHz, 1MHz, and 2MHz, respectively, there are 5 GSM cells proposing spectrum requirements, 4 CDMA2000 cells, and 6 WCDMA cells, whose spectrum requirements obey the uniform distribution of [0,100], [0,16], and [0,4], respectively, and spectrum benefits obey the uniform distribution of [0.2,2], [1.25,12.5], and [5,50], respectively. All cells are uniformly distributed in a 100 x 100 area, the coverage radius of all cells is 10, and the cells generating the spectrum requirements are fixed.
The multi-target dynamic spectrum allocation method of the quantum flower pollination search mechanism is marked as MQFPA-DSA, the quantum pollen set scale P is 20, the iteration termination time is 800, c 1＝0.08，c 2＝0.01，c 30.06, 0.001/Q for c, and 0.5 for u. Compared with a single-target dynamic spectrum allocation method (QGA-DSA) using a quantum genetic algorithm, the population scale of the quantum genetic algorithm is set to be P20, the number of times of iteration termination is 800, and other parameters are consistent with those of cognitive radio spectrum allocation based on the quantum genetic algorithm, which is published in physical science and newspaper by Zhao Zhijin and the like.

Claims (5)

1. A multi-target spectrum allocation method based on a quantum pollination search mechanism in a cognitive heterogeneous network is characterized by comprising the following steps:
(1) a wireless access network sensing module in the base station senses network information, predicts a service request, puts forward spectrum requirements and network resource benefits, and sends the information to an access network reconfiguration management module through an information interaction interface of the access network reconfiguration management module; the access network reconfiguration management module collects the frequency spectrum demand information of all base stations managed by the access network reconfiguration management module and provides the information to the network reconfiguration management module through an information interaction interface of the access network reconfiguration management module;
(2) the network reconfiguration management module carries out multi-granularity channel division on the frequency spectrum resources to obtain a channel set, and obtains a base station set, a frequency spectrum demand set and a frequency spectrum resource benefit set according to received information arrangement, and obtains a base station interference matrix and a channel interference matrix through calculation; finally, establishing a cognitive heterogeneous network multi-target spectrum allocation model;
(3) initializing a quantum pollen set containing P quantum pollens, and measuring the quantum pollen set to obtain a pollen set, wherein each pollen in the pollen set represents a possible frequency spectrum allocation result; setting the initial iteration time t to be 0;
(4) mapping each pollen individual in the pollen set to be a frequency spectrum distribution matrix for correction, mapping the corrected frequency spectrum distribution matrix to be pollen, and evaluating the fitness of each corrected pollen; setting 4 fitness functions such as frequency spectrum benefit, frequency spectrum demand satisfaction rate, frequency spectrum use efficiency and frequency spectrum occupation according to different network requirements;
(5) setting the conversion probability u of cross pollination and self pollination in a quantum pollination search mechanism, and updating by adopting a cross pollination mode if rand is less than u and rand is an element [0,1] which meets the uniformly distributed random number for each dimension of quantum pollen; if rand is more than or equal to u, updating by adopting a self-pollination mode; measuring the updated quantum pollen set to obtain a new-generation pollen set, mapping each pollen individual in the new-generation pollen set to be a frequency spectrum distribution matrix for correction, mapping the corrected frequency spectrum distribution matrix to be pollen, and evaluating the fitness of each corrected pollen;
(6) mixing a new-generation pollen set and a previous-generation pollen set, performing non-dominant solution level sorting and congestion degree calculation, performing descending order arrangement on the pollen with the same non-dominant solution level according to the congestion degree value, selecting the pollen with the non-dominant level of 1 and the higher congestion degree, and adding the pollen into an elite pollen set G; namely, when the number of the pollen in the elite pollen set G is more than 2P, the pollen in the elite pollen set G is subjected to non-dominant solution level sorting and congestion degree calculation, the pollen with the same non-dominant solution level is subjected to descending order sorting according to the congestion degree value, and the top 2P excellent pollen is selected as a new elite pollen set G;
(7) if the maximum iteration number is not reached, t is t +1, and the step 5 is returned to continue; otherwise, stopping iteration, and sequencing the pollen in the obtained elite pollen set by a non-dominant solution; selecting the pollen with the non-dominant solution grade of 1 as a final Pareto front-end solution, selecting proper pollen from the Pareto front-end solution and mapping the pollen to a frequency spectrum distribution matrix A *；
(8) The network reconfiguration management module allocates the optimal distribution matrix A *Partitioning is carried out, and a spectrum resource allocation result is notified to the access network reconfiguration management module through a decision interactive interface of the access network reconfiguration management module; and an access network reconfiguration control module in the base station realizes reconfiguration according to the spectrum allocation result, namely adaptively modifies working parameters through software.
2. The multi-target spectrum allocation method based on the quantum pollination search mechanism in the cognitive heterogeneous network as claimed in claim 1, wherein the multi-target spectrum allocation method comprises the following steps: the cognitive heterogeneous network multi-target spectrum allocation model comprises a channel set, a base station set, a spectrum demand set, a spectrum resource benefit set, a base station interference matrix and a channel interference matrix;
the width of the frequency spectrum resource to be distributed is F, and the channel bandwidths supported by the access technologies correspondingly adopted by the K wireless access networks are respectively W kK is 1,2, …, K, the spectrum resources are divided into multi-granularity channels before allocation; corresponding to the kth access network, the number of available channels is M k＝[F/W k]Therein, []Indicating rounding down, the total number of these available channels is
Figure FDA0002236936910000021
Numbering each channel by using an integer M, wherein M is more than or equal to 1 and less than or equal to M, and forming a channel set { phi (phi) } by using the channels corresponding to each access network 1,Φ 2,…,Φ KWhere phi is 1＝{1,2,…,M 1}，
Figure FDA0002236936910000022
Forming base station set by cell base stations participating in spectrum allocation
Figure FDA0002236936910000023
Wherein the content of the first and second substances,
Figure FDA0002236936910000024
indicating the nth base station and the access network belonging to the kth type; obtaining a frequency spectrum demand set B ═ B by a corresponding base station set nN ═ 1,2, …, N } and the set of spectral benefits R ═ { R ═ R nmN is 1,2, …, N, M is 1,2, …, M }, wherein b is nNumber of channels, r, representing the requirement of the nth base station nmIndicating the spectrum resource benefit of the nth base station using the mth channel;
the base station interference matrix is represented as a block matrix
Figure FDA0002236936910000025
Wherein the sub-block matrix
Figure FDA0002236936910000026
Indicating a base station
Figure FDA0002236936910000027
And
Figure FDA0002236936910000028
the interference relationship between the two or more of the two,
Figure FDA0002236936910000029
indicating a base station
Figure FDA00022369369100000210
And
Figure FDA00022369369100000211
simultaneous use of the same channel or overlapping channels may interfere with each other
Figure FDA00022369369100000212
Represent non-interfering; when k is 1＝k 2When the temperature of the water is higher than the set temperature,
Figure FDA00022369369100000213
according to the spectral multiplexing coefficient of the access technical specification, and when j 1＝j 2When the temperature of the water is higher than the set temperature,
Figure FDA00022369369100000214
indicating that each BS cannot be allocated an overlapping channel; when k is 1≠k 2When, if
Figure FDA00022369369100000215
And
Figure FDA00022369369100000216
the distance between them is less than the sum of their radius of coverage, they are considered to interfere with each other,
Figure FDA00022369369100000217
otherwise
The channel interference matrix is represented as
Figure FDA00022369369100000219
When m is 1≠m 2Then, if channel m 1And channel m 2With an overlapping portion, then
Figure FDA00022369369100000220
Indicating that the base station uses channel m 1Will be to channel m 2Cause interference, otherwise In the same way, when m 1＝m 2When the temperature of the water is higher than the set temperature,
the spectrum allocation matrix is denoted as A N×M＝(a nm) N×MIf, if
Figure FDA00022369369100000223
Get the channel m, then a nm1, otherwise a nm＝0；
Figure FDA00022369369100000224
The obtained channel number is
Figure FDA00022369369100000225
Base stations for which a feasible allocation matrix is required to satisfy mutual interference cannot allocate overlapping or identical channels, i.e.
Figure FDA00022369369100000226
Access network
Figure FDA00022369369100000227
Dividable channel set phi kOf (3) a channel, i.e.
Figure FDA00022369369100000228
And its resulting channel number e n≤min(b n,M k) (ii) a Wherein the content of the first and second substances, n 1,n 2∈[1,N]an integer in between;
Figure FDA00022369369100000230
m 1,m 2∈[1,M]an integer in between;
Figure FDA00022369369100000231
indicating whether channel m belongs to set phi for indicating function kDefinition of
Figure FDA0002236936910000031
3. The multi-target spectrum allocation method based on the quantum pollination search mechanism in the cognitive heterogeneous network as claimed in claim 1, wherein the fitness evaluation process for each pollen in the quantum pollination set is as follows:
measuring the quantum pollen set to obtain an initial pollen set
Figure FDA0002236936910000032
Binary string after measurement of ith quantum pollen in quantum pollen collection
Figure FDA0002236936910000033
Is a pollen that represents one possible spectrum allocation result, wherein,
Figure FDA0002236936910000034
Figure FDA0002236936910000035
is a random number satisfying uniform distribution; mapping each pollen individual in the pollen set to be a frequency spectrum distribution matrix for correction, and then mapping the corrected frequency spectrum distribution matrix to be pollen, the specific steps are as follows: setting a zero matrix A N×M(ii) a From pollen
Figure FDA0002236936910000036
Intermediate corresponding channel
Figure FDA0002236936910000037
Distributing the state segments to obtain the nth row element of the matrix A; a is subjected to a feasibility check if
Figure FDA0002236936910000038
And in the presence of interference, i.e. I BS(n 1,n 2) 1, then look for the nth 1Line non-zero elements
Figure FDA00022369369100000310
And n is 2Line non-zero elements
Figure FDA00022369369100000311
Judging channel m 1And m 2Whether it is the same channel or an overlapping channel, if I Channel(m 1,m 2) 1, then modify randomly Or
Figure FDA00022369369100000313
Is 0; computing
Figure FDA00022369369100000314
The obtained channel number is
Figure FDA00022369369100000315
If e nIs greater than Required number of channels b nThen choose randomly (e) n-b n) Setting the allocated channels to zero, and finally mapping the corrected allocation matrix into pollen; repeating the steps to correct all the pollen in the pollen set;
evaluating the fitness of each corrected pollen, and performing fitness function according to different network requirements
Figure FDA00022369369100000317
z is chosen from the following functions:
(4.1) network EffectBenefit: wherein A is the ith pollen in the pollen collection
Figure FDA00022369369100000319
A spectrum allocation matrix corresponding to mapping, wherein r is a spectrum benefit matrix, M is the total number of channels, and N is the total number of access networks;
(4.2) RAN Spectrum requirement satisfaction rate: wherein e is nNumber of spectrum finally allocated to the nth access network, b nThe spectrum requirement actually provided for the nth access network, wherein N is the total number of the access networks;
(4.3) spectrum use efficiency:
Figure FDA00022369369100000321
representing the spectrum bandwidth actually utilized by the access network; wherein e is nThe number of frequency spectrums W finally allocated to the nth access network kThe channel granularity of the kth heterogeneous network is obtained, and N is the total number of access networks;
(4.4) Spectrum Occupancy (SO):
Figure FDA00022369369100000322
representing an actual utilization ratio of the spectrum resources; wherein, SU is spectrum utilization efficiency, and F is spectrum resource width to be allocated.
4. The multi-target spectrum allocation method based on the quantum pollination search mechanism in the cognitive heterogeneous network as claimed in claim 1, wherein the updating process of each pollen in the quantum pollination set is as follows:
setting the conversion probability u of cross pollination and self pollination in quantum pollination search mechanism, and setting the jth dimension of ith quantum pollen of the tth generation
Figure FDA0002236936910000041
If rand is less than u, updating by adopting a cross pollination mode, wherein the quantum rotation angle updating formula is
Figure FDA0002236936910000042
Wherein rand is from [0,1]]Is a random number that satisfies a uniform distribution,
Figure FDA0002236936910000043
is the ith pollen in the t generation pollen collection
Figure FDA0002236936910000044
The j-th dimension value of (1) is more than or equal to i and less than or equal to P, and 1 is more than or equal to j and less than or equal to Q; globally optimal pollen
Figure FDA0002236936910000045
Randomly selecting the first 2% pollen in the elite pollen set,
Figure FDA0002236936910000046
is the value of its j-th dimension; c. C 1The constant represents the influence degree of the globally optimal pollen on the evolution of the quantum pollen set; if rand is more than or equal to u, updating by adopting a self-pollination mode, wherein the quantum rotation angle updating formula is
Figure FDA0002236936910000047
Among them, the preferred pollen
Figure FDA0002236936910000048
Randomly selecting pollen in the elite solution set,
Figure FDA0002236936910000049
is the value of its j-th dimension; random pollen
Figure FDA00022369369100000410
The selection is made at random from a collection of pollen,
Figure FDA00022369369100000411
is the value of its j-th dimension; c. C 2And c 3Is a constant which respectively represents the influence degree of the better pollen in the elite pollen set and the random pollen in the pollen set on the evolution of the quantum pollen set;
dimension j for ith quantum pollen of t generation
Figure FDA00022369369100000412
Using quantum rotary gate for updating, the formula is
Figure FDA00022369369100000413
Wherein the content of the first and second substances,
Figure FDA00022369369100000414
is a random number satisfying uniform distribution, abs (. cndot.) represents an absolute value, and c is a variation probability at a quantum rotation angle of 0, whose value is [0,1/Q ]]A constant value of (1);
Figure FDA00022369369100000415
is a quantum not-gate (QNOT-gate),
Figure FDA00022369369100000416
is a quantum revolving gate, and t +1 and t are iteration times; and measuring the updated quantum pollen set to obtain a new-generation pollen set, mapping each pollen individual in the new-generation pollen set to be a frequency spectrum distribution matrix for correction, mapping the corrected frequency spectrum distribution matrix to be pollen, and evaluating the fitness of each corrected pollen.
5. The method for multi-target spectrum allocation based on the quantum pollination search mechanism in the cognitive heterogeneous network as claimed in claim 1, wherein the non-dominated solution sorting and crowding degree calculation process is as follows:
for the maximum multi-objective optimization problem, f is set z(x i) (z is 1,2, …, w) is flowerPowder x iW is the number of selected targets; for pollen x iAnd x gIf the fitness value f z(x i)≥f z(x g) If all the targets are satisfied and at least one strict inequality is satisfied, then the pollen x is called iDominating pollen x g，x iNon-dominant pollen; if the fitness value f z(x i)≤f z(x g) If all the targets are satisfied and at least one strict inequality is satisfied, then the pollen x is called gDominating pollen x i，x gNon-dominant pollen; if the above two conditions are not satisfied, pollen x iAnd x gNo dominance relationship;
the process of non-dominant pollen ranking is as follows: for each pollen x in the pollen collection iCalculating dominant pollen x iNumber of pollen of
Figure FDA00022369369100000417
And pollen x iSet of dominated pollen Dominating pollen x iNumber of (2)
Figure FDA00022369369100000419
Pollen x iHas a non-dominated ranking of 1; pollen x with rank 1 for each non-dominant ranking iTraverse the collection of pollen it dominates
Figure FDA0002236936910000051
Each pollen x in (1) gCalculating dominant pollen x gNumber of
Figure FDA0002236936910000052
If it is
Figure FDA0002236936910000053
Then the pollen x gPutting the pollen into a set H, and sequencing the non-dominant grade of the pollen in the set to be 2; repeating the above process for each pollen in set H, thereby obtaining a pollen set with a non-dominant ranking of 3, and repeating the process until all pollen non-dominant rankings are obtained;
the congestion degree calculation is carried out on the pollen with the same non-dominant grade, the pollen with the same non-dominant grade is sorted from small to large according to the fitness value of a certain target, the congestion degree value of the pollen with the maximum and minimum fitness values is infinity, and the congestion degree values of other pollen are the difference of the fitness values of two adjacent pollen divided by the difference of the maximum and minimum fitness values; the above calculation is performed for all targets, and the final congestion value is the sum of the congestion values calculated for all targets.
CN201610859145.6A 2016-09-28 2016-09-28 Multi-target spectrum allocation method based on quantum pollination search mechanism in cognitive heterogeneous network Active CN106658524B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610859145.6A CN106658524B (en) 2016-09-28 2016-09-28 Multi-target spectrum allocation method based on quantum pollination search mechanism in cognitive heterogeneous network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610859145.6A CN106658524B (en) 2016-09-28 2016-09-28 Multi-target spectrum allocation method based on quantum pollination search mechanism in cognitive heterogeneous network

Publications (2)

Publication Number Publication Date
CN106658524A CN106658524A (en) 2017-05-10
CN106658524B true CN106658524B (en) 2020-02-11

Family

ID=58854716

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610859145.6A Active CN106658524B (en) 2016-09-28 2016-09-28 Multi-target spectrum allocation method based on quantum pollination search mechanism in cognitive heterogeneous network

Country Status (1)

Country Link
CN (1) CN106658524B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107656239B (en) * 2017-08-22 2020-12-04 哈尔滨工程大学 Coherent information source direction finding method based on polarization sensitive array
CN108055665B (en) * 2017-10-27 2020-04-28 杭州电子科技大学 Target channel access method based on improved multi-target PSO optimization
CN110110753B (en) * 2019-04-03 2023-08-25 河南大学 Effective mixed characteristic selection method based on elite flower pollination algorithm and ReliefF
CN110309887A (en) * 2019-07-09 2019-10-08 哈尔滨理工大学 Based on the Fuzzy C-Means Clustering method for detecting abnormality for improving flower pollination
CN113630886B (en) * 2021-08-27 2023-07-18 黑龙江八一农垦大学 Spectrum allocation method based on particle swarm algorithm in heterogeneous Internet of things

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104168569A (en) * 2014-07-15 2014-11-26 哈尔滨工程大学 Dynamic frequency spectrum distribution method of cognitive heterogeneous network

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20120063983A (en) * 2010-12-08 2012-06-18 한국전자통신연구원 Method of channel allocation in multi-channel cognitive radio network

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104168569A (en) * 2014-07-15 2014-11-26 哈尔滨工程大学 Dynamic frequency spectrum distribution method of cognitive heterogeneous network

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
Quantum-behaved Flower Pollination Algorithm;Kezhong Lu等;《2015 14th International Symposium on Distributed Computing and Applications for Business Engineering and Science》;20151231;第66-69页 *

Also Published As

Publication number Publication date
CN106658524A (en) 2017-05-10

Similar Documents

Publication Publication Date Title
CN106658524B (en) Multi-target spectrum allocation method based on quantum pollination search mechanism in cognitive heterogeneous network
CN107172682B (en) Ultra-dense network wireless resource allocation method based on dynamic clustering
US8023459B2 (en) Method and apparatus for staggered zone resource allocation in orthogonal frequency division multiple access system
CN1620177A (en) Apparatus and method for allocating channel in a wireless communication system
Ismail et al. Decentralized radio resource allocation for single-network and multi-homing services in cooperative heterogeneous wireless access medium
JP6879737B2 (en) Methods and devices for mitigating interference in heterogeneous networks using cell-to-cell interference control
US20130196676A1 (en) Method and apparatus of communications
CN104619029A (en) Method and device for allocating baseband pool resources under integrated cellular network architecture
CN111328146B (en) Service scheduling method for optimizing transmission rate weight based on genetic algorithm
Khanbary et al. A GA-based effective fault-tolerant model for channel allocation in mobile computing
CN113630886A (en) Spectrum allocation method based on particle swarm algorithm in heterogeneous Internet of things
CN108848535A (en) A kind of mist calculating environmental resource distribution method towards shared model
CN104168569A (en) Dynamic frequency spectrum distribution method of cognitive heterogeneous network
US9253781B2 (en) Scheduling in consideration of terminal groups in a mobile communication system
CN1541004A (en) Method of and device for resource allocation in digital celluar communication systems
CN106575983B (en) Method and apparatus for switching control region in wireless communication system
Veeramakali et al. Intelligent dynamic spectrum allocation with bandwidth flexibility in cognitive radio network
Duan et al. Research on spectrum allocation of HF access network based on intelligent frequency hopping
Singh et al. A model for cognitive channel allocation using GA
WO2021004287A1 (en) Resource allocation method, apparatus and device for intelligent space-divided cell cluster, and storage medium
CN107182057B (en) Inter-cell cooperation method and device in LTE system
KR20080096978A (en) System and method for allocating resource in a communication system
Chen et al. An interference-mitigation channel allocation algorithm for energy-efficient femtocell networks
CN116367332B (en) Hierarchical control-based D2D resource allocation method under 5G system
Wang et al. Spectrum aggregation and allocation based on shuffled frog leaping algorithm for cognitive radio networks

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