CN110505681B - Non-orthogonal multiple access scene user pairing method based on genetic method - Google Patents

Non-orthogonal multiple access scene user pairing method based on genetic method Download PDF

Info

Publication number
CN110505681B
CN110505681B CN201910743280.8A CN201910743280A CN110505681B CN 110505681 B CN110505681 B CN 110505681B CN 201910743280 A CN201910743280 A CN 201910743280A CN 110505681 B CN110505681 B CN 110505681B
Authority
CN
China
Prior art keywords
population
user
users
scheme
pairing
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
CN201910743280.8A
Other languages
Chinese (zh)
Other versions
CN110505681A (en
Inventor
潘志文
由瀚良
刘楠
尤肖虎
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.)
Southeast University
Original Assignee
Southeast 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 Southeast University filed Critical Southeast University
Priority to CN201910743280.8A priority Critical patent/CN110505681B/en
Publication of CN110505681A publication Critical patent/CN110505681A/en
Application granted granted Critical
Publication of CN110505681B publication Critical patent/CN110505681B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/12—Computing arrangements based on biological models using genetic models
    • G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L5/00—Arrangements affording multiple use of the transmission path
    • H04L5/003—Arrangements for allocating sub-channels of the transmission path
    • H04L5/0037—Inter-user or inter-terminal allocation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L5/00—Arrangements affording multiple use of the transmission path
    • H04L5/003—Arrangements for allocating sub-channels of the transmission path
    • H04L5/0058—Allocation criteria
    • H04L5/0064—Rate requirement of the data, e.g. scalable bandwidth, data priority
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W52/00—Power management, e.g. TPC [Transmission Power Control], power saving or power classes
    • H04W52/04—TPC
    • H04W52/18—TPC being performed according to specific parameters
    • H04W52/26—TPC being performed according to specific parameters using transmission rate or quality of service QoS [Quality of Service]
    • H04W52/267—TPC being performed according to specific parameters using transmission rate or quality of service QoS [Quality of Service] taking into account the information rate
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W52/00—Power management, e.g. TPC [Transmission Power Control], power saving or power classes
    • H04W52/04—TPC
    • H04W52/30—TPC using constraints in the total amount of available transmission power
    • H04W52/34—TPC management, i.e. sharing limited amount of power among users or channels or data types, e.g. cell loading
    • H04W52/346—TPC management, i.e. sharing limited amount of power among users or channels or data types, e.g. cell loading distributing total power among users or channels

Abstract

The invention discloses a genetic method-based non-orthogonal multiple access scene user pairing method, which comprises the following steps: necessary parameters required by heredity are configured, and the parameters comprise a population scale P and an evolution algebra T; and (3) encoding: converting a specific user pairing scheme into a sequence; selecting a pairing scheme, and generating an initial population as an initial value of iterative computation; the pairing schemes are elements of the initial population; evaluation: substituting a sequence corresponding to the pairing scheme individual in the population of the current agent into a fitness function to obtain a corresponding fitness function value; genetic manipulation: when the current calculation iteration number does not reach T, performing genetic operation on the current population to generate a new generation of population; until evolution to T generation; and decoding the individual with the highest fitness function value in the P individuals of the current population to obtain an approximate optimal solution of the user pairing scheme in the NOMA scene. This method provides a considerable reduction in the time required, compared to other conventional solutions to this problem.

Description

Non-orthogonal multiple access scene user pairing method based on genetic method
Technical Field
The invention belongs to the technical field of non-orthogonal multiple access in wireless communication, and particularly relates to a user pairing method in a non-orthogonal multiple access scene based on a genetic method.
Background
Limited by the constraints of signal processing technology and system complexity, Orthogonal Multiple Access (OMA) schemes are adopted in the first to fourth generation wireless communication systems. To meet the requirements of new generation mobile communication systems for transmission rate and transmission reliability, a Non-Orthogonal Multiple Access (NOMA) scheme is considered as a better alternative to OMA. In OMA, the signals of different users are orthogonal to each other on the channel resources, i.e. each user occupies a respective channel resource; in NOMA, multiple users are allowed to multiplex channel resources for communication, and a required signal is obtained by a Serial Interference Cancellation (SIC) technique at a receiving end. For any given number of users and number of packets to be grouped, how to plan access user pairing in the NOMA scenario to maximize the information transmission rate of the whole system is a complex problem, and the traditional method cannot be used for solving in a short time. The NOMA scene user pairing method based on the genetic method can obtain the approximate optimal solution of the problem in a short time.
Disclosure of Invention
The technical problem to be solved by the invention is as follows:
in order to maximize the information transmission rate of the whole system when the access users in a mixed NOMA scene are paired, the invention provides a genetic method-based non-orthogonal multiple access scene user pairing method.
The invention adopts the following technical scheme for solving the technical problems:
the invention provides a genetic method-based non-orthogonal multiple access scene user pairing method, which comprises the following steps:
step one, configuring necessary parameters required by heredity, wherein the parameters comprise a population scale P and an evolution algebra T;
step two, coding: converting a specific user pairing scheme into a sequence;
selecting a user pairing scheme, and generating an initial population as an initial value of iterative computation by using the user pairing scheme as an element;
step four, evaluation: substituting a sequence corresponding to a user pairing scheme in a population of a current agent into a fitness function to obtain a fitness function value corresponding to the user pairing scheme;
step five, genetic manipulation: when the current calculation iteration number does not reach T, performing genetic operation on the current population to generate a new generation of population; the genetic operations comprise selection operations, crossover operations and mutation operations;
step six, when the evolution algebra counter T is equal to T, the genetic method is not continuously executed; and decoding the individual with the highest fitness function value in the current population P individuals, and considering the pairing scheme obtained by decoding as an approximate optimal solution of the user pairing scheme in the NOMA scene.
The method for pairing users in a non-orthogonal multiple access scenario based on a genetic method as described above, further, the specific step of encoding in step two includes:
step 2.1, grouping users, expressed as:
M＝m1+m2+…+mK
wherein, K is the number of the packets to be grouped, M is the number of users, MkRepresenting the number of users in the kth group;
2.2, selecting user pairing schemes with different permutation and coding strategies for coding; for a given user set containing M users, numbering the users according to the descending order of the channel gain to obtain a user number set N, wherein the channel gain of communication between each user and a base station can be obtained by channel estimation;
step 2.3, for the user pairing scheme corresponding to the kth individual in the current iteration, the sequence I is obtained after the codingkComprises the following steps:
Figure BDA0002164713210000021
Figure BDA0002164713210000022
wherein, the sequence IkMiddle front m1The value is the number assigned to the user in the first group, followed by m2The value is the number assigned to the user in the second group and so on.
As mentioned above, the method for pairing users in a non-orthogonal multiple access scenario based on a genetic method further includes the specific steps of generating an initial population in step three:
step 3.1, roughly randomly selecting P schemes in a user pairing scheme set to be selected to form an initial population;
and 3.2, setting the evolution algebra counter t to be 0 to record the number of the calculated iterations.
As mentioned above, in the method for pairing users in a non-orthogonal multiple access scenario based on a genetic method, further, the fitness function in step four is taken as an information transmission rate of the whole system, and is expressed as:
Figure BDA0002164713210000023
wherein K is the number of the users to be grouped, M is the number of the users, B is the total bandwidth of the base station, and n0Is additive white Gaussian noise power spectral density hiChannel gain for communication between the ith user and the base station is obtained by channel estimation; p is a radical ofi,jThe power value obtained when the ith user is distributed to the jth group can be obtained by the power distribution scheme in the group; x is the number ofi,j1 means that the ith user is assigned to the jth group, xi,jWhen the number is 0, the ith user is not allocated to the jth group;
Figure BDA0002164713210000024
is an additional noise part of the SIC.
As mentioned above, in the method for pairing users in a non-orthogonal multiple access scenario based on a genetic method, further, in the fifth step, the selection operation uses an elite tournament policy as a selection policy, and the specific steps in the fifth step include:
step 5.1.1, directly reserving the user pairing scheme with the maximum information transmission rate in the current iteration to the next iteration;
step 5.1.2, randomly selecting a certain number of individual user pairing schemes from the population, and selecting the optimal one of the schemes to enter the next generation of population;
step 5.1.3, judging whether the number of the individual user pairing schemes in the next generation population reaches n; if the number of the N is less than n, repeating the step 5.1.2; if n, finishing the genetic operation of the agent;
step 5.1.4, judging the iteration times of the current population; if the iteration times do not reach T, taking the next generation population obtained in the step 5.1.3 as the population of the current agent and returning to the step 5.1.1; and if the iteration number reaches T, ending the iteration.
The method for pairing the users in the non-orthogonal multiple access scenario based on the genetic method as described above, further, the interleaving operation in step five includes:
step 5.2.1, roughly randomly selecting two individuals I in the current generation populationmAnd InAs a parent individual;
step 5.2.2, selection of ImAnd InDesignated portions of the gene sequence as matching sequences:
Figure BDA0002164713210000031
wherein l is the matching section length;
step 5.2.3, exchanging the matching sequences to obtain filial generation individuals
Figure BDA0002164713210000032
And
Figure BDA0002164713210000033
Figure BDA0002164713210000034
step 5.2.4, contemporary individuals
Figure BDA0002164713210000035
And
Figure BDA0002164713210000036
if the gene duplication occurs, the matched part is kept unchanged after the exchange, and the duplicated genes outside the matched part are corrected according to the one-to-one correspondence of the matched part until no duplicated genes exist.
The method for pairing the users in the non-orthogonal multiple access scene based on the genetic method further comprises the step fiveThe different operations comprise: pairing scheme individuals I for mutated usersmIn other words, two genes with variations in the gene sequence are selected almost randomly
Figure BDA0002164713210000037
And
Figure BDA0002164713210000038
the position of the individual is exchanged to obtain the individual of the user pairing scheme after genotype variation
Figure BDA0002164713210000039
Expressed as:
Figure BDA00021647132100000310
wherein the content of the first and second substances,
Figure BDA00021647132100000311
and
Figure BDA00021647132100000312
namely two genes which are subjected to position exchange in the mutation operation,
Figure BDA00021647132100000313
and matching the scheme individuals for the users containing the new gene sequences obtained after the mutation.
Compared with the prior art, the invention adopting the technical scheme has the following technical effects:
the user pairing method in the NOMA scene based on the genetic method can provide an approximately optimal solution of a pairing scheme in a short time by utilizing the parallel computing capability and the random searching capability of the genetic method aiming at any given user number M and the number K to be grouped; and under the condition of reasonable parameter setting, the optimal solution of the pairing scheme can be found out. The method of the present invention provides a considerable reduction in the time required, compared to other conventional solutions to this problem.
In addition, for a mixed NOMA scenario in which only 2 users or 1 user (in this case, OMA) exists in each group, the optimal pairing scheme may be directly given without a genetic method, that is, users with high channel gain are allocated to a single user group as much as possible, and then the remaining users are paired according to the optimal pairing scheme in the NOMA scenario of 2 users. The pairing result obtained in the way is proved to be an optimal solution, and the conclusion has certain application value in consideration of the limitation of the NOMA technology on the computation complexity and the equipment complexity in the practical application.
Drawings
FIG. 1 is a block diagram of the steps of the method of the present invention.
Detailed Description
The technical scheme of the invention is further explained in detail by combining the attached drawings:
it will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
In the NOMA scenario, multiple users are allowed to occupy the same channel resource for communication, and limited by device complexity and computational complexity, all users participating in communication are generally grouped, users in the same group share the channel resource by using the NOMA scheme, and the channel resource between different groups is still orthogonal. Such a scheme is also known as the hybrid NOMA scheme. Therefore, in this scenario, the information transmission rate of the whole system will be affected by the change of the packet where any user is located. How to find a grouping scheme that maximizes the system information transmission rate is a valuable problem, namely the pairing problem of access users in a hybrid NOMA scenario.
In the invention, when solving the pairing problem of the access users in the mixed NOMA scene, the aim is to maximize the information transmission rate of the whole system:
Figure BDA0002164713210000051
wherein, K is the number of users to be grouped, M is the number of users, and B is the total bandwidth of the base station, and it is assumed here that the channel resources divided by each group of users are equal, so that B/K is the bandwidth divided by each group of NOMA access users. n is0Is additive white Gaussian noise power spectral density hiThe channel gains for the ith user in communication with the base station may all be obtained from channel estimates. p is a radical ofi,jThe power level assigned to the ith user when assigned to the jth group may be obtained from the intra-group power assignment scheme. x is the number ofi,jIs a variable from 0 to 1, xi,j1 means that the ith user is assigned to the jth group, xi,j0 means that the ith user is not assigned to the jth group, while it should be noted that each user can and can only belong to one group.
Figure BDA0002164713210000052
The noise part added in the SIC (SIC is the existing concept, and the content is only briefly described here) process, different users in the same group can be divided into different powers according to the requirement defined by NOMA, and the SIC is carried out at the receiving end according to the different powers of the users. The smaller the channel gain of the user in the same group is, the higher the power of the user is, the receiving end first solves the signal sent by the user with the minimum channel gain in the group, and because the power of the user with the minimum channel gain is the maximum (determined by NOMA definition), the signals transmitted by all other users can be regarded as noise processing, so as to obtain the signal sent by the user. And then directly subtracting the obtained signal transmitted by the user with the minimum channel gain from the signal received by the receiving end, and continuously repeating the process until the receiving end obtains the signals transmitted by all the users. The procedure of sequentially obtaining the NOMA scheme user sending signals is the SIC procedure.
The present invention solves the above problems by using a genetic method, which is a computational model that simulates the natural selection and genetic mechanisms of biological evolution, starting with an iteration from a set of potential solutions, also called populations, of the problem to be solved. The population consists of a certain number of individuals, each individual is a feasible user pairing scheme, a proper coding mode is selected for the individual to code the user pairing scheme accessed in a specific NOMA scene, and the selection of the coding strategy directly influences the selection of subsequent strategies. And secondly, selecting a proper fitness function to evaluate the performance of each individual, so that the individual with better performance in each generation of population can be selected. And finally, simulating concepts of chromosome crossing and gene variation in genetics, generating individuals with brand-new codes in each calculation iteration, wherein the new codes mean a brand-new pairing scheme, and searching the individuals with better performance by the method so as to enable the offspring population to be more suitable for the requirements of the problem to be solved than the previous generations. The optimal individual in the last generation population can be used as an approximate optimal solution of the problem after being decoded. The invention adopts a genetic method with an Elitist Preservation (EP) strategy, which can directly reserve the optimal individuals in the previous generation population into the next generation population and can accelerate the convergence speed of calculation. By adopting the method, the user pairing scheme in the NOMA scene can be obtained in a short time aiming at any user number M and the number K to be grouped.
The technical scheme of the invention is shown in figure 1, and comprises the following steps:
the first step is as follows: necessary parameters required by the genetic method are configured. The population scale P and the evolution algebra T are two necessary hyper-parameters in the genetic method, and need to be set before calculation execution, and the value of the hyper-parameters is directly related to the time required by calculation and the closeness degree of the result and the optimal solution. The method comprises the following steps:
(1) selecting the population scale P: the size of the population scale means the size of the coverage area of the parallel calculation of the genetic method, and the larger the population scale is, the larger the calculation coverage area is, but the more the calculation resources are occupied. The population size varying from several tens to several hundreds can be set according to the number M of users in the present invention.
(2) Selecting evolution algebra T: the evolution algebra means the number of times of loop iteration of the genetic method, and the algebra of the current population is recorded by an evolution algebra counter t in actual operation. When the initial seed group is generated, t is 0. The more evolutionary algebra, the closer the result is to the optimal solution of the problem. The specific numerical value of the evolution algebra can be set according to the number M of the users.
The second step is that: encoding a specific user pairing scheme. In the genetic method-based user pairing scheme, a specific user pairing scheme is converted into a process of finding a specific sequence required, each sequence representing a different pairing scheme, and such a process is called encoding. For any packet determined by the number K of packets to be grouped and the number M of users, there are:
M＝m1+m2+…+mK (2)
wherein m is1Representing the number of users in the first group, m2Representing the number of users in the second group, and so on. In the invention, a permutation coding strategy (the permutation coding strategy is an existing concept, and the content of the permutation coding strategy is only briefly described) is selected to code different user pairing methods in the NOMA scene. The permutation coding strategy is designed aiming at the permutation and combination problem, aiming at a given user set U containing M users, the users are numbered according to the descending order of the channel gain (the channel gain of the communication between each user and the base station can be obtained by channel estimation) to obtain a user number set N, namely, the user with the largest channel gain in the set U is numbered as 1, the user with the second largest channel gain is numbered as 2, and so on, the user with the smallest channel gain is numbered as M. For the user pairing scheme corresponding to the kth individual in the current iteration, a sequence I obtained after coding is obtainedkComprises the following steps:
Figure BDA0002164713210000061
sequence IkMiddle front m1The value is the number assigned to the user in the first group, followed by m2The value is the number assigned to the user in the second group and so on. Thus, each encoded sequence can be completely tabulatedTo a user pairing scheme. Following the concept in biology, in genetic methods IkAlso referred to as the genotype of the individual, sequence IkThe value at each position in (a) is also referred to as the gene at that position.
The third step: and generating an initial population. And the initial population of the genetic method is an initial value of calculation, and P schemes are selected randomly in a potential user pairing scheme set to form the initial population, wherein P is the size of the population scale set in the first step. And meanwhile, setting an evolution algebra counter t to be 0 to record the number of calculated iterations.
The fourth step: and evaluating the advantages and disadvantages of the pairing schemes corresponding to the individuals in the current population by means of the fitness function, wherein the pairing scheme with a larger fitness function value is considered to be a more excellent scheme, and the pairing scheme is more likely to be reserved in the subsequent selection operation until the next calculation iteration. The problem to be solved in the invention is a single-target optimization problem, so that the information transmission rate of the whole system can be directly used as a fitness function:
Figure BDA0002164713210000071
wherein K is the number of the users to be grouped, M is the number of the users, B is the total bandwidth of the base station, and n0Is additive white Gaussian noise power spectral density hiThe channel gains for the ith user in communication with the base station may all be obtained from channel estimates. p is a radical ofi,jThe power level assigned to the ith user when assigned to the jth group may be obtained from the intra-group power assignment scheme. x is the number ofi,j1 means that the ith user is assigned to the jth group, xi,jWhen 0 means that the ith user is not assigned to the jth group.
Figure BDA0002164713210000072
Is an additional noise part of the SIC.
The fifth step: and when the current calculation iteration number does not reach T, performing genetic operation on the current population, wherein T is the size of an evolution algebra set in the first step. Genetic manipulations can be divided into three categories, respectively:
(1) selecting operation: the selection operation follows the concept of natural selection in biology, and aims to establish a proper selection mechanism, so that a pairing scheme for obtaining a larger fitness function (the fitness function is the system information transmission rate in the invention is explained in the fourth step) can be reserved until the next calculation iteration. And selecting the Elite tournament strategy as a selection strategy according to the characteristics of the problem to be solved. The strategy is provided with an elite reservation mechanism, namely, a pairing scheme for obtaining the maximum information transmission rate in the current iteration is directly reserved to the next iteration, then a certain number of individuals are randomly selected from the population each time, and the optimal one of the individuals is selected to enter the next generation of population. And repeating the operation until the size of the new generation of population reaches the size of the original population.
(2) And (3) cross operation: the cross operation imitates the concept of chromosome crossing in genetics, is a core mechanism for improving the searching capability of a genetic method, and refers to randomly selecting two individuals I in a previous generation populationmAnd InPerforming a crossover operation, ImAnd InAlso called parent, willmAnd InPartial structure in gene sequence is replaced and recombined to generate individual with new gene sequence
Figure BDA0002164713210000073
And
Figure BDA0002164713210000074
Figure BDA0002164713210000075
and
Figure BDA0002164713210000076
also called as offspring individuals. The crossover operation can introduce a completely new pairing scheme for the calculation, thereby continuously exploring the pairing scheme capable of obtaining higher information transmission rate. Aiming at the characteristics of the permutation coding strategy, the invention selects a partial matching cross strategy. Selection of ImAnd InThe designated part in the gene sequence is used as a matching sequence, and the length l of the matching part can be set by self：
Figure BDA0002164713210000081
Wherein, the dotted line frame is a matching part, and the matching sequence is exchanged to obtain an offspring individual
Figure BDA0002164713210000082
And
Figure BDA0002164713210000083
Figure BDA0002164713210000084
if the gene duplication occurs in the generated offspring individuals, the matched parts are kept unchanged after the exchange, and the duplicated genes outside the matched parts are corrected according to the one-to-one correspondence relationship of the matched parts until no duplicated genes exist.
(3) Mutation operation: mutation operation follows the concept of gene mutation in genetics and is a necessary mechanism for genetic methods to improve random search capability. The mutation operation can also introduce a brand new pairing scheme for calculation, which can be regarded as effective supplement of cross operation, thereby continuously exploring the pairing scheme capable of obtaining higher information transmission rate and accelerating the convergence process of the result to the optimal solution. Each individual in the population has a certain mutation probability, and the mutation individual I takes account of the characteristic that the permutation code does not allow the repeated user numbers in the gene sequence of the same individualmIn other words, two genes with variations in the gene sequence are selected almost randomly
Figure BDA0002164713210000085
And
Figure BDA0002164713210000086
subjecting to position exchange to obtain genotypic variation individuals
Figure BDA0002164713210000087
Figure BDA0002164713210000088
Wherein the content of the first and second substances,
Figure BDA0002164713210000089
and
Figure BDA00021647132100000810
namely two genes which are subjected to position exchange in the mutation operation,
Figure BDA00021647132100000811
the individual containing the new gene sequence obtained after mutation.
And a sixth step: when the evolution algebra counter T is T, the genetic method does not continue to execute. And decoding the individuals with the optimal performance in the current population P individuals, and considering the pairing scheme obtained by decoding as an approximately optimal solution of the user pairing scheme in the NOMA scene.
Example one
In the NOMA scenario based on the genetic method, the user pairing method is described by taking the example that the number of users M is 8 and the number of groups to be grouped K is 3. There are many choices of grouping schemes, and in this example, only the grouping scheme of M ═ 2+3+3 ═ 8 is taken as an example to illustrate the steps of the genetic method, and other grouping schemes are the same.
The example comprises the following steps:
the first step is as follows: necessary parameters required by the genetic method are configured. The population scale P and the evolution algebra T are two necessary parameters in the genetic method and need to be set before calculation execution, and the value of the parameters is directly related to the time required by calculation and the proximity degree of the result and the optimal solution.
The method comprises the following steps:
(1) selecting the population scale P: the size of the population scale means the size of the coverage area of parallel calculation by the genetic method, and the larger the population scale is, the more resources are occupied by calculation. The population size can be set to 50 in this example according to the number of users M being 8.
(2) Selecting evolution algebra T: the evolution algebra means the number of times of loop iteration of the genetic method, and the algebra of the current population is recorded by an evolution algebra counter t in actual operation. When the initial seed group is generated, t is 0. The more evolutionary algebra, the closer the result is to the optimal solution of the problem. In this example, the evolution algebra can be set to 20 generations according to the number M of users being 8. If it is desired to find or approach the optimal solution as much as possible, the evolution generations can be increased to 50.
The second step is that: encoding a specific user pairing scheme. In the genetic method-based user matching scheme, a specific user matching scheme needs to be converted into a process of finding a required specific sequence, each sequence represents a different matching scheme, and the process is coding. For any packet determined by the number K of packets to be grouped and the number M of users, there are:
M＝m1+m2+…+mK (1)
wherein m is1Representing the number of users in the first group, m2Representing the number of users in the second group, and so on. In this example, M ═ M1+m2+m32+2+ 3. In the invention, a permutation coding strategy (the permutation coding strategy is an existing concept, and the content of the permutation coding strategy is only briefly described) is selected to code different user pairing methods in the NOMA scene. The permutation coding strategy is designed aiming at the permutation combination problem, aiming at a given user set U containing M users, the users are numbered according to the descending order of the channel gain (the channel gain of the communication between each user and the base station can be obtained by channel estimation) to obtain a user number set N, namely, the user with the largest channel gain in the set U is numbered as 1, the user with the second largest channel gain is numbered as 2, and so on, the user with the smallest channel gain is numbered as 8.
Figure BDA0002164713210000091
For the user pairing scheme corresponding to the kth individual in the current iteration, the sequence obtained after coding is obtainedColumn IkComprises the following steps:
Figure BDA0002164713210000092
sequence IkThe first 2 values are numbers assigned to 2 users in the first group, the last 3 values are numbers assigned to 3 users in the second group, and the last 3 values are numbers assigned to 3 users in the third group. Therefore, each encoded sequence can completely express a user pairing scheme. Following the concept in biology, in genetic methods IkAlso referred to as the genotype of the individual, sequence IkThe value at each position in (a) is also referred to as the gene at that position.
The third step: the initial population of the genetic method is an initial value of the calculation, P schemes are selected randomly in a potential user pairing scheme set to form the initial population, and the size of the population is set to be 50 in the first step of the example. And meanwhile, setting an evolution algebra counter t to be 0 to record the number of calculated iterations.
The fourth step: and evaluating the advantages and disadvantages of the pairing schemes corresponding to the individuals in the current population by means of the fitness function, wherein the pairing scheme with a larger fitness function value is considered to be a more excellent scheme, and the pairing scheme is more likely to be reserved in the subsequent selection operation until the next calculation iteration. The problem to be solved in the invention is a single-target optimization problem, so that the information transmission rate of the whole system can be directly used as a fitness function:
Figure BDA0002164713210000101
in this embodiment, the power spectral density n of additive white gaussian noise0Is 1W/Hz, the channel bandwidth B is 3Hz, eta1＝2,η2＝3,η3＝3。hiThe channel gain for the communication of the ith user with the base station may be obtained from channel estimation. p is a radical ofi,jThe power value obtained when the ith user is allocated to the jth group can be determined by the power value in the groupA rate allocation scheme is obtained. x is the number ofi,j1 means that the ith user is assigned to the jth group, xi,jWhen 0 means that the ith user is not assigned to the jth group.
Figure BDA0002164713210000102
Is an additional noise part of the SIC.
The fifth step: and when the current calculation iteration number does not reach T, performing genetic operation on the current population, wherein T is the size of an evolution algebra set in the first step. In this example, the genetic procedure included the following:
(1) selecting operation: the selection operation follows the concept of natural selection in biology, and aims to establish a proper selection mechanism, so that a pairing scheme for obtaining a larger fitness function (the fitness function is the system information transmission rate in the invention is explained in the fourth step) can be reserved until the next calculation iteration. And selecting the Elite tournament strategy as a selection strategy according to the characteristics of the problem to be solved. In this example, the individual with the maximum fitness function among 50 individuals in the current population is directly reserved to the next generation, then 5 individuals are randomly selected from the 50 individuals in the current population each time, and the one with the maximum fitness function among the 5 individuals is selected and reserved to the next generation. The above operations are repeated until the size of the new generation population reaches 50.
(2) And (3) cross operation: cross operation is a core mechanism for improving search capability by a genetic method, and refers to the random selection of two individuals I in a previous generation populationmAnd InPerforming a crossover operation, ImAnd InAlso called parent, willmAnd InPartial structure in gene sequence is replaced and recombined to generate individual with new gene sequence
Figure BDA0002164713210000103
And
Figure BDA0002164713210000104
Figure BDA0002164713210000105
and
Figure BDA0002164713210000106
also called as offspring individuals. The crossover operation can introduce a completely new pairing scheme for the calculation, thereby continuously exploring the pairing scheme capable of obtaining higher information transmission rate. Aiming at the characteristics of the permutation coding strategy, the invention selects a partial matching cross strategy. Selection of ImAnd InThe designated part in the gene sequence is used as a matching sequence, and the length l of the matching part in the example is 3:
Figure BDA0002164713210000107
in this embodiment, the matching part is located inside the dashed box, and the matching sequences are exchanged to obtain the offspring individuals
Figure BDA0002164713210000111
And
Figure BDA0002164713210000112
Figure BDA0002164713210000113
if the gene duplication occurs in the generated offspring individuals, the matched parts are kept unchanged after the exchange, and the duplicated genes outside the matched parts are corrected according to the one-to-one correspondence relationship of the matched parts until no duplicated genes exist. In this example
Figure BDA0002164713210000114
A third gene of (a) and
Figure BDA0002164713210000115
the 7 th gene is repeated with the matched partial gene, and the corrected result is:
Figure BDA0002164713210000116
(3) mutation operation: mutation operation follows the concept of gene mutation in genetics and is a necessary mechanism for genetic methods to improve random search capability. The mutation operation can also introduce a brand new pairing scheme for calculation, which can be regarded as effective supplement of cross operation, thereby continuously exploring the pairing scheme capable of obtaining higher information transmission rate and accelerating the convergence process of the result to the optimal solution. Each individual in the population has a certain mutation probability, in this example the mutation probability is 0.1, and for the individual with mutation Im：
Im＝[2 5 3 4 1 6 7 8] (8)
Isogenerally randomly selecting two genes on a Gene sequence
Figure BDA0002164713210000117
And
Figure BDA0002164713210000118
performing position exchange to obtain genotypic variation individuals
Figure BDA0002164713210000119
In this example, genes at the third and fifth positions were selected to obtain mutated individuals
Figure BDA00021647132100001110
Comprises the following steps:
Figure BDA00021647132100001111
wherein the content of the first and second substances,
Figure BDA00021647132100001112
and
Figure BDA00021647132100001113
namely two genes which are subjected to position exchange in the mutation operation,
Figure BDA00021647132100001114
the individual containing the new gene sequence obtained after mutation.
And a sixth step: when the evolution algebra counter t is 20, the genetic method does not continue to execute. Decoding the individuals with the optimal performance in 50 individuals in the current population, and considering the pairing scheme obtained by decoding as an approximately optimal solution of the user pairing scheme in the NOMA scene.
The foregoing is only a partial embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.

Claims (3)

1. A non-orthogonal multiple access scene user pairing method based on a genetic method is characterized by comprising the following steps:
step one, configuring necessary parameters required by heredity, wherein the parameters comprise a population scale P and an evolution algebra T;
step two, coding: converting a specific user pairing scheme into a sequence;
selecting a user pairing scheme, and generating an initial population as an initial value of iterative computation by using the user pairing scheme as an element;
step four, evaluation: substituting a sequence corresponding to a user pairing scheme in a population of a current agent into a fitness function to obtain a fitness function value corresponding to the user pairing scheme;
and step four, taking the fitness function as the information transmission rate of the whole system, and expressing the fitness function as follows:
Figure FDA0003292387680000011
wherein K is the number of the users to be grouped, M is the number of the users, B is the total bandwidth of the base station, and n0Is additive white Gaussian noise power spectral density hiChannel gain for communication between the ith user and the base station is obtained by channel estimation; p is a radical ofi，jIs allocated for the ith userThe power value obtained by the time of reaching the jth group can be obtained by the power distribution scheme in the group; x is the number ofi，j1 means that the ith user is assigned to the jth group, xi，jWhen the number is 0, the ith user is not allocated to the jth group;
Figure FDA0003292387680000012
is the noise part added to the SIC;
step five, genetic manipulation: when the current calculation iteration number does not reach T, performing genetic operation on the current population to generate a new generation of population; the genetic operations comprise selection operations, crossover operations and mutation operations;
step five the selection operation takes the elite tournament strategy as a selection strategy, and the concrete steps of the step five comprise:
step 5.1.1, directly reserving the user pairing scheme with the maximum information transmission rate in the current iteration to the next iteration;
step 5.1.2, randomly selecting a certain number of individual user pairing schemes from the population, and selecting the optimal one of the schemes to enter the next generation of population;
step 5.1.3, judging whether the number of the individual user pairing schemes in the next generation population reaches n; if the number of the N is less than n, repeating the step 5.1.2; if n, finishing the genetic operation of the agent;
step 5.1.4, judging the iteration times of the current population; if the iteration times do not reach T, taking the next generation population obtained in the step 5.1.3 as the population of the current agent and returning to the step 5.1.1; if the iteration times reach T, ending the iteration;
step five, the cross operation comprises the following steps:
step 5.2.1, roughly randomly selecting two individuals I in the current generation populationmAnd InAs a parent individual;
step 5.2.2, selection of ImAnd InDesignated portions of the gene sequence as matching sequences:
Figure FDA0003292387680000021
wherein l is the matching section length;
step 5.2.3, exchanging the matching sequences to obtain filial generation individuals
Figure FDA0003292387680000022
And
Figure FDA0003292387680000023
Figure FDA0003292387680000024
step 5.2.4, contemporary individuals
Figure FDA0003292387680000025
And
Figure FDA0003292387680000026
if the situation of gene duplication occurs, keeping the matched part unchanged after the exchange, and correcting the duplicated genes outside the matched part according to the one-to-one correspondence of the matched part until no duplicated genes exist;
step five, the mutation operation comprises the following steps: pairing scheme individuals I for mutated usersmIn other words, two genes with variations in the gene sequence are selected almost randomly
Figure FDA0003292387680000027
And
Figure FDA0003292387680000028
the position of the individual is exchanged to obtain the individual of the user pairing scheme after genotype variation
Figure FDA0003292387680000029
Expressed as:
Figure FDA00032923876800000210
wherein the content of the first and second substances,
Figure FDA00032923876800000211
and
Figure FDA00032923876800000212
namely two genes which are subjected to position exchange in the mutation operation,
Figure FDA00032923876800000213
and matching the scheme individuals for the users containing the new gene sequences obtained after the mutation.
2. The method for pairing users in a non-orthogonal multiple access scenario based on a genetic method as claimed in claim 1, wherein the specific step of encoding in step two comprises:
step 2.1, grouping users, expressed as:
M＝m1+m2+…+nK
wherein, K is the number of the packets to be grouped, M is the number of users, MkRepresenting the number of users in the kth group;
2.2, selecting user pairing schemes with different permutation and coding strategies for coding; for a given user set containing M users, numbering the users according to the descending order of the channel gain to obtain a user number set N, wherein the channel gain of communication between each user and a base station can be obtained by channel estimation;
step 2.3, for the user pairing scheme corresponding to the kth individual in the current iteration, the sequence I is obtained after the codingkComprises the following steps:
Figure FDA00032923876800000214
Figure FDA00032923876800000215
wherein, the sequence IkMiddle front m1The value is the number assigned to the user in the first group, followed by m2The value is the number assigned to the user in the second group and so on.
3. The method for pairing users in a non-orthogonal multiple access scenario based on a genetic method according to claim 1, wherein the specific step of generating the initial population in step three comprises:
step 3.1, roughly randomly selecting P schemes in a user pairing scheme set to be selected to form an initial population;
and 3.2, setting the evolution algebra counter t to be 0 to record the number of the calculated iterations.
CN201910743280.8A 2019-08-13 2019-08-13 Non-orthogonal multiple access scene user pairing method based on genetic method Active CN110505681B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910743280.8A CN110505681B (en) 2019-08-13 2019-08-13 Non-orthogonal multiple access scene user pairing method based on genetic method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910743280.8A CN110505681B (en) 2019-08-13 2019-08-13 Non-orthogonal multiple access scene user pairing method based on genetic method

Publications (2)

Publication Number Publication Date
CN110505681A CN110505681A (en) 2019-11-26
CN110505681B true CN110505681B (en) 2022-02-22

Family

ID=68586622

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910743280.8A Active CN110505681B (en) 2019-08-13 2019-08-13 Non-orthogonal multiple access scene user pairing method based on genetic method

Country Status (1)

Country Link
CN (1) CN110505681B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110958044B (en) * 2019-12-02 2022-07-29 东南大学 Non-orthogonal multiple access user clustering method based on density clustering
CN111294959B (en) * 2020-02-07 2022-03-18 安徽大学 Optimization method and optimization device for joint user grouping and power distribution

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2016165095A1 (en) * 2015-04-16 2016-10-20 富士通株式会社 Information transmission method, device and communication system
WO2018064582A1 (en) * 2016-09-30 2018-04-05 Intel IP Corporation Grant-free uplink non-orthogonal multiple access transmissions
CN108023842A (en) * 2017-12-01 2018-05-11 重庆邮电大学 The pilot design method of extensive mimo system
CN109195214A (en) * 2018-09-06 2019-01-11 南京邮电大学 A kind of NOMA power distribution method based on genetic algorithm

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2016165095A1 (en) * 2015-04-16 2016-10-20 富士通株式会社 Information transmission method, device and communication system
WO2018064582A1 (en) * 2016-09-30 2018-04-05 Intel IP Corporation Grant-free uplink non-orthogonal multiple access transmissions
CN108023842A (en) * 2017-12-01 2018-05-11 重庆邮电大学 The pilot design method of extensive mimo system
CN109195214A (en) * 2018-09-06 2019-01-11 南京邮电大学 A kind of NOMA power distribution method based on genetic algorithm

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
Joint Power and Resource Allocation for Non-Uniform Topologies in Heterogeneous Networks;由瀚良;《IEEE》;20160707;全文 *

Also Published As

Publication number Publication date
CN110505681A (en) 2019-11-26

Similar Documents

Publication Publication Date Title
CN110267338B (en) Joint resource allocation and power control method in D2D communication
Kalathil et al. Decentralized learning for multiplayer multiarmed bandits
CN111628855B (en) Industrial 5G dynamic multi-priority multi-access method based on deep reinforcement learning
CN111132264B (en) User clustering method for downlink of multi-user MIMO-NOMA system
CN111294959B (en) Optimization method and optimization device for joint user grouping and power distribution
CN110505681B (en) Non-orthogonal multiple access scene user pairing method based on genetic method
CN105873214A (en) Resource allocation method of D2D communication system based on genetic algorithm
CN106792451B (en) D2D communication resource optimization method based on multi-population genetic algorithm
CN113242601B (en) NOMA system resource allocation method based on optimized sample sampling and storage medium
CN110167176B (en) Wireless network resource allocation method based on distributed machine learning
Hao et al. Efficient and robust emergence of norms through heuristic collective learning
CN114330715A (en) Intelligent ammunition co-evolution task allocation method
CN107483079B (en) Double-population genetic ant colony routing method for low-voltage power line carrier communication
CN114158105B (en) Routing method and device of wireless sensor network applied to comprehensive energy system
CN116744311B (en) User group spectrum access method based on PER-DDQN
CN110995399B (en) Large-scale MIMO pilot frequency distribution method based on user grouping
CN109195222B (en) Power distribution method based on statistical characteristic reference
CN105376185A (en) Constant modulus blind equalization processing method based on optimization of DNA shuffled frog leaping algorithm in communication system
CN108260193B (en) Joint resource allocation method and device based on channel aggregation in heterogeneous network
CN115103372A (en) Multi-user MIMO system user scheduling method based on deep reinforcement learning
CN114938543A (en) Honeycomb heterogeneous network resource allocation method based on deep reinforcement learning
CN112737726A (en) Static and dynamic combined TDMA time slot allocation method
Wu et al. An improved resource allocation scheme for plane cover multiple access using genetic algorithm
CN111290849A (en) Method, device, equipment and medium for dynamically adjusting service resources
CN114745087B (en) Pilot frequency distribution method, device, equipment and storage medium

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