CN1968489B - Resource allocation method and system - Google Patents

Abstract

The invention relates to a method for distributing resource, which comprises that: 1, building the relation between distributed resource and solid; using gene algorism to obtain the solid with maximum adaptation, and the resource to obtain maximum distribution adaptation region; 2, calculating the resources distributed to each user in said region. The invention also discloses a resource distributor system. The invention can realize optimized resource distribution.

Description

A kind of method and system of resource allocation

Technical field

The present invention relates to communication field, relate in particular to a kind of method and system of resource allocation.

Background technology

In the communications field, there is the gain of certain hour diversity in user's behavior.In order to improve the utilance of resource, then need to carry out multiplexing to resource.So, introduced shared channel, different user distribution according to need.Existing dispatching algorithm, in user's fairness, user's present located environmental differences, the service rate of user's request, user's historical speed, local resource is mostly just considered one to two factor in the factors such as user's scheduling priority.According to the different factors of considering, amplify out different control algolithms then.PF strategy is for example only considered user's dispatching priority; The maxC/I strategy is only considered user's present located environmental differences; The Round-Robin strategy with, only consider user's fairness.

Because these several algorithms are mostly just considered one to two factor, be difficult in the allocation strategy that reaches optimum on the overall situation.

Summary of the invention

The technical problem to be solved in the present invention provides a kind of method and system of resource allocation, and these method and system can be taken all factors into consideration various distribution factors, reaches optimum allocation strategy on the overall situation.

The objective of the invention is to be achieved through the following technical solutions:

The invention provides a kind of method of resource allocation, comprise step:

1) set up sub-district resource to be allocated and individual corresponding relation, utilize genetic algorithm, according to

Calculate user's scheduling of resource fitness to be allocated, wherein, ARi is that dispatching priority of users, Hi are for history is obtained service speed, Ri is the fairness tolerance that channel circumstance, Fi obtain for the user for application speed, Ci for resource to be allocated, the Pi of user i; With each user scheduling of resource fitness to be allocated addition that calculates, obtain the scheduling of resource fitness to be allocated of sub-district, promptly obtain the ideal adaptation degree; Compare each individual fitness, obtain the individuality of maximum adaptation degree, promptly obtain the resource to be allocated of maximum scheduling fitness sub-district;

2) described maximum resource to be allocated of dispatching each user's correspondence in the fitness sub-district is identical figure place binary coding; Calculate the number of the binary coding correspondence of each user's correspondence, obtain each user's resource to be allocated;

The described corresponding relation of setting up community user resource to be allocated and individuality comprises:

Set up community user resource binary coding to be allocated and individual binary-coded corresponding relation.

The number that each user of described sub-district resource binary to be allocated coding is corresponding and allow resource to be allocated smaller or equal to maximum

Obtain the resource to be allocated of described user i in the following manner:

11) generate N individuality at random;

14) calculate the number of the binary coding correspondence of user's correspondence in each individual corresponding district, obtain the resource ARi to be allocated of user i.

Preferably, between described step 11) and step 14), also comprise:

111) calculate a bulk concentration, if a bulk concentration then enters step 14), otherwise enters step 112 greater than threshold value); Wherein, described bulk concentration be and should individuality between affinity greater than the individual number of default affinity constant and total number of individuals purpose quotient;

112) accumulation step 114) number of repetition, if number of repetition is greater than maximum number of repetitions, then enters step 14), otherwise enter step 113);

113) use cell scheduling fitness, individual bulk concentration, calculate the individual replicate ratio;

114) duplicate according to the individual replicate ratio and exchange pairing, variation, generate individually, enter step 111).

The present invention also provides a kind of resource distributor system, comprising:

Obtain Resource Unit to be allocated, be used to set up sub-district resource to be allocated and individual corresponding relation, utilize genetic algorithm, according to

Calculate user's scheduling of resource fitness to be allocated, wherein, ARi is that dispatching priority of users, Hi are for history is obtained service speed, Ri is the fairness tolerance that channel circumstance, Fi obtain for the user for application speed, Ci for resource to be allocated, the Pi of user i; With each user scheduling of resource fitness to be allocated addition that calculates, obtain the scheduling of resource fitness to be allocated of sub-district, promptly obtain the ideal adaptation degree, relatively each individual fitness, obtain the individuality of maximum adaptation degree, promptly obtain the resource to be allocated of maximum scheduling fitness sub-district;

User's resource computing unit to be allocated, be used for when described maximum scheduling fitness each user's corresponding district of sub-district resource to be allocated is identical figure place binary coding, calculate the number of the binary coding correspondence of each user's correspondence, obtain each user's resource to be allocated;

Wherein, the described corresponding relation of setting up community user resource to be allocated and individuality is: set up community user resource binary coding to be allocated and individual binary-coded corresponding relation; The number that each user of described sub-district resource binary to be allocated coding is corresponding and allow resource to be allocated smaller or equal to maximum;

Wherein, obtain the resource to be allocated of described user i in the following manner:

11) generate N individuality at random, described individual corresponding binary coding of producing at random satisfies condition number and allow resource to be allocated smaller or equal to maximum of each user resource binary coding to be allocated correspondence;

14) calculate the number of the binary coding correspondence of user's correspondence in each individual corresponding district, obtain the resource ARi to be allocated of user i.

Preferably, describedly obtain Resource Unit to be allocated and comprise:

The corresponding relation unit is used to set up community user resource to be allocated and individual corresponding relation;

The heredity unit is used for the corresponding relation according to the foundation of described corresponding relation unit, utilizes genetic algorithm, according to

Calculate user's scheduling of resource fitness to be allocated, wherein, ARi is that dispatching priority of users, Hi are for history is obtained service speed, Ri is the fairness tolerance that channel circumstance, Fi obtain for the user for application speed, Ci for resource to be allocated, the Pi of user i; With each user scheduling of resource fitness to be allocated addition that calculates, obtain the scheduling of resource fitness to be allocated of sub-district, promptly obtain the ideal adaptation degree, relatively each individual fitness, obtain the individuality of maximum adaptation degree, promptly obtain the resource to be allocated of maximum scheduling fitness sub-district.

Preferably, described hereditary unit comprises:

Obtain maximum adaptation degree individual cell, be used for basis

Calculate user's scheduling of resource fitness to be allocated, wherein, ARi is that dispatching priority of users, Hi are for history is obtained service speed, Ri is the fairness tolerance that channel circumstance, Fi obtain for the user for application speed, Ci for resource to be allocated, the Pi of user i; With each user scheduling of resource fitness to be allocated addition that calculates, obtain the scheduling of resource fitness to be allocated of sub-district, promptly obtain the ideal adaptation degree, relatively each individual fitness, obtain the individuality of maximum adaptation degree, promptly obtain the resource to be allocated of maximum scheduling fitness sub-district;

Individual bulk concentration unit is used to calculate a bulk concentration, and relatively individual bulk concentration and threshold value size, if a bulk concentration is greater than threshold value, inform that then obtaining maximum adaptation degree individual cell obtains maximum adaptation degree individuality, if bulk concentration, is then informed the unit that the adds up number of repetition that adds up less than threshold value; Wherein, described bulk concentration be and should individuality between affinity greater than the individual number of default affinity constant and total number of individuals purpose quotient;

The unit adds up, be used to the number of repetition that adds up, relatively number of repetition is big or small with maximum repetition, if number of repetition is greater than maximum number of repetitions, inform that then obtaining maximum adaptation degree individual cell obtains maximum adaptation degree sub-district, obtain the sub-district resource to be allocated of maximum scheduling fitness, if number of repetition, informs then that the individual replicate unit calculates the individual replicate ratio less than maximum number of repetitions;

The individual replicate unit is used to use cell scheduling fitness, individual bulk concentration, calculates the individual replicate ratio; The unit that duplicates, exchanges, makes a variation is used for duplicating and exchange pairing, variation according to the individual replicate ratio, generates individual.

Above technical scheme as can be seen, owing to utilize global optimum's query capability of genetic algorithm, have the individuality of maximum adaptation degree by global search, thereby obtain the resource to be allocated of maximum scheduling fitness sub-district, realize the scheduling purpose of optimized resource distribution.

Further, set up sub-district resource to be allocated and individual corresponding relation by binary system, quote global optimum's assignment problem that Biological Principles solves resource allocation, connect two fields in succinct mode, operation is easily conveniently understood.

Further, calculate this step of bulk concentration by in the process of genetic algorithm, quoting, thus the restriction of jumping out locally optimal solution.

Further, when the calculating user dispatches fitness, take all factors into consideration the various factors that in scheduling, occurs, realize reaching the scheduling purpose of resource optimum allocation.

Description of drawings

Fig. 1 is total method flow diagram provided by the invention;

Fig. 2 is the method first embodiment flow chart provided by the invention;

Fig. 3 is the method second embodiment flow chart provided by the invention;

Fig. 4 is method the 3rd an embodiment flow chart provided by the invention;

Fig. 5 is that system first provided by the invention implements illustration;

Fig. 6 is that system second provided by the invention implements illustration.

Embodiment

The present invention is applied in the scheduling process of resource allocation, to realize the purpose of resource allocation global optimum.Its core concept is: set up sub-district resource to be allocated and individual corresponding relation, utilize genetic algorithm to obtain the sub-district of the maximum scheduling of resource to be allocated fitness, thereby calculate the resource to be allocated of each user in this sub-district.User's scheduling of resource fitness to be allocated refers in the scheduling of resource process, and each user's correspondence is obtained the quality of Resources allocation.

According to above core concept, the invention provides the method for resource allocation.

Seeing also Fig. 1 is total method flow diagram provided by the invention, comprising:

D1) set up sub-district resource to be allocated and individual corresponding relation, utilize genetic algorithm to obtain the individuality of maximum adaptation degree, correspondence is obtained the resource to be allocated of maximum scheduling fitness sub-district;

D2) calculate the resource to be allocated that maximum is dispatched each user's correspondence in the sub-district of fitness.

According to above-mentioned total method flow, the invention provides specific embodiment.

Following embodiment provides the detailed process of the corresponding relation of setting up sub-district resource to be allocated and individuality, thereby sets up getting in touch of resource allocation and genetic algorithm, utilizes genetic algorithm to obtain the sub-district of the maximum scheduling of resource to be allocated fitness.

Seeing also Fig. 2 is method first embodiment provided by the invention, comprising:

E1 sets up corresponding relation;

Set up sub-district resource to be allocated and individual corresponding relation, detailed process is as follows:

Each individual binary coding mode adopts following mode to characterize, suppose to have m user, the resource to be allocated of each user's correspondence is AR, the resource to be allocated of first user's correspondence is AR1, the resource to be allocated of second user's correspondence is AR2, and the resource to be allocated of m user's correspondence is ARm; M user's resource to be allocated and allow resource to be allocated smaller or equal to maximum, also be the corresponding number of m user's resource binary coding to be allocated and allow resource to be allocated smaller or equal to maximum.Wherein the order that is arranged in order of user generates at random.

The user 1	The user 2	......	User m
				AR1	AR2	......	?ARm

The length of ARi is allowed the binary system length decision of the resource correspondence of distribution by maximum, for example, suppose that maximum Resources allocation is 8, and then the length of ARi is 3, and each individual length is 3*m.

E2 obtains the resource to be allocated of maximum scheduling fitness sub-district;

Utilize genetic algorithm to obtain and have the resource to be allocated of the promptly maximum scheduling fitness of the individuality sub-district of maximum adaptation degree.

E3 calculates the resource to be allocated that maximum is dispatched each user's correspondence in the sub-district of fitness;

The sub-district of maximum scheduling fitness can have maximum scheduling fitness to user's resource to be allocated in should the sub-district.

So far, this time finishing scheduling is waited for scheduling next time.

Community user resource to be allocated and individual corresponding relation process have been increased in the present embodiment, concrete binary coded system foundation corresponding relation between the two that adopts, be understandable that, can adopt other coded system equally, as real coding, it is identical that its process of setting up corresponding relation and present embodiment are set up the corresponding relation process.

Seeing also Fig. 3 is method second embodiment provided by the invention,

Following embodiment increase utilizes genetic algorithm to obtain the individuality of maximum adaptation degree, the promptly corresponding detailed process of obtaining the sub-district of the maximum scheduling of resource to be allocated fitness, in N individuality, seek the individuality of maximum adaptation degree, promptly seek the resource to be allocated of maximum scheduling fitness sub-district, enlarge the searching scope, find out optimal solution.

Comprise step:

F1 sets up corresponding relation;

Set up community user resource to be allocated and individual corresponding relation, specific as follows:

Each individual binary coding mode adopts following mode to characterize, suppose to have m user, the resource to be allocated of each user's correspondence is AR, the resource to be allocated of first user's correspondence is AR1, the resource to be allocated of second user's correspondence is AR2, and the resource to be allocated of m user's correspondence is ARm; M user's resource to be allocated and allow resource to be allocated smaller or equal to maximum, also be the corresponding number of m user's resource binary coding to be allocated and allow resource to be allocated smaller or equal to maximum.Wherein the order that is arranged in order of user generates at random.

The user 1	The user 2	......	User m
				AR1	AR2	......	ARm

The length of ARi is allowed the binary system length decision of the resource correspondence of distribution by maximum, for example, suppose that maximum Resources allocation is 8, and then the length of ARi is 3, and each individual length is 3*m, i.e. L.

F2 generates N individuality at random;

The individual corresponding binary coding that generates at random satisfies corresponding condition number and allow resource to be allocated smaller or equal to maximum of each user resource binary coding to be allocated.

F3 calculates the resource to be allocated of user in each individual corresponding district;

Computational process is as follows:

Suppose to have 5 users, individual corresponding binary coding is 0,100 0,001 0,001 0,101 0101, ARi such as following table that then final each user distributes:

UE	The ARi encoded radio	The ARi actual value
			1	0100	4
2	0001	1
			3	0001	1
4	0101	5
			5	0101	5

In the table, promptly corresponding each user's of ARi actual value resource to be allocated.

F4 calculates user's scheduling fitness, and is specific as follows:

Suppose that the user's number that participates in dispatching is m, wherein the dispatching priority of user i is Pi (human configuration), it is Hi (drawing by historical statistical data) that history is obtained service speed, application speed is Ri (human configuration), channel circumstance is Ci (passing through apparatus measures), and the fairness tolerance that the user obtains is Fi (passing through apparatus measures); Resource to be allocated is ARi (above-mentioned steps draws).Wherein, Pi, Hi, Ri, Ci, Fi, all normalization.

The formula of particular user scheduling fitness (Fitness (i)) is as follows:

$\mathrm{Fitness}\left(i\right)=\mathrm{Fi}*\frac{\mathrm{Hi}-\mathrm{Ri}}{\mathrm{Ri}}*\mathrm{Ci}*\mathrm{Pi}*\mathrm{ARi}$

M user's resource to be allocated and wherein smaller or equal to maximum resource to be allocated.

A corresponding in esse m user, corresponding N the individual resource difference to be allocated of each user, other parameter is identical.

The scheduling of resource fitness to be allocated of F5 calculation plot, correspondence obtains the ideal adaptation degree;

Specific as follows:

With each user scheduling of resource fitness to be allocated Fitness (i) addition that step F 4 calculates, obtain the scheduling of resource fitness to be allocated (Fcell) of sub-district.

$\mathrm{Fcell}=\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Fitness}\left(i\right)$

F6 repeating step F3, F4, F5 obtains N individual fitness;

F7 obtains the resource to be allocated of maximum scheduling fitness sub-district;

Compare N individual fitness, obtain the individuality of maximum adaptation degree, promptly obtain the resource to be allocated of maximum scheduling fitness sub-district.

F8 calculates the resource to be allocated that maximum is dispatched each user's correspondence in the sub-district of fitness;

The sub-district of maximum scheduling fitness can have maximum scheduling fitness to the resource to be allocated of user in should the sub-district.

Concrete process is as follows:

Suppose to have 5 users, the individual corresponding binary coding with maximum adaptation degree is 0,110 00,010,000 0,111 0010, ARi such as following table that then final each user distributes:

UE	The ARi encoded radio	The ARi actual value
			1	0110	6
2	0001	1
			3	0000	0
4	0111	7
			5	0010	2

In the table, promptly corresponding each user's of ARi actual value resource to be allocated.So far this finishing scheduling is waited for scheduling next time.

The present embodiment increase utilizes genetic algorithm to obtain the process of the individuality of maximum adaptation degree, is understandable that, also can utilize other method in the genetic algorithm to obtain the individuality of maximum adaptation degree.

Seeing also Fig. 4 is method the 3rd embodiment provided by the invention,

This embodiment increase utilizes genetic algorithm to generate N individuality, utilizes Biological Principles to obtain optimum population individuality, thereby provides prerequisite for seeking global optimum's individuality.

Comprise step:

G1 sets up corresponding relation;

Set up community user resource to be allocated and individual corresponding relation, specific as follows:

Each individual binary coding mode adopts following mode to characterize, suppose to have m user, the resource to be allocated of each user's correspondence is AR, the resource to be allocated of first user's correspondence is AR1, the resource to be allocated of second user's correspondence is AR2, and the resource to be allocated of m user's correspondence is ARm; M user's resource to be allocated and allow resource to be allocated smaller or equal to maximum, also be the corresponding number of m user's resource binary coding to be allocated and allow resource to be allocated smaller or equal to maximum.Wherein the order that is arranged in order of user generates at random.

The user 1	The user 2	......	User m
				AR1	AR2	......	ARm

The length of ARi is allowed the binary system length decision of the resource correspondence of distribution by maximum, for example, suppose that maximum Resources allocation is 8, and then the length of ARi is 3, and each individual length is 3*m, i.e. L.

G2 generates N individuality at random;

The individual corresponding binary coding that generates at random satisfies corresponding condition number and allow resource to be allocated smaller or equal to maximum of each user resource binary coding to be allocated.

G3 calculates a bulk concentration;

Detailed process is as follows:

In N the individuality, the long L of the coding of each individuality, the character set of employing size for S (to adopt binary system compile individual character set for 0,1}, S=2), and the number of each individual binary coding correspondence allows resource to be allocated smaller or equal to maximum.The comentropy H (N) of genes of individuals seat j is defined as

$H_j\left(N\right)=\underset{i=1}{\overset{s}{\mathrm{\Σ}}}-P_{\mathrm{ij}}\log \left(P_{\mathrm{ij}}\right),$ Pij is that i symbol appears at the probability on the locus j, and Pij=locus j goes up the total number/N that i symbol occur.Mean entropy is $H\left(N\right)=\frac{1}{L}\underset{j=1}{\overset{L}{\mathrm{\Σ}}}{H}_j\left(N\right)$ 。The affinity degree that defines between two individual u and v is

$A_{u,v}=\frac{1}{1+H\left(2\right)}$ , Au, the span of v is (0,1), Au, v are big more, represent that two individualities are affine or similar.Au, v=1 represent that then both encoding genes are identical.So the concentration of individual i is defined as number of individuals/N of Cdi=and individual i affinity＞λ.λ is the affinity constant, general value 0.9≤λ≤1.

G4 bulk concentration and predetermined threshold value are relatively;

If a bulk concentration then enters step G10, otherwise enters step G5 greater than predetermined threshold value (0.8);

The invention provides predetermined threshold value is 0.8, but the present invention do not limit this value, and those skilled in the art can decide according to concrete condition.

The G5 number of repetition that adds up, relatively number of repetition and maximum number of repetitions;

If number of repetition surpasses maximum number of repetitions, then enter step G10, otherwise enter step G6;

Wherein, obtain the sub-district of the promptly maximum scheduling fitness of individuality with maximum adaptation degree; Maximum number of repetitions provided by the invention is 20, but the present invention do not limit this value, and those skilled in the art can decide according to concrete condition.The number of times that the number of repetition that present embodiment provides is meant and duplicates, exchanges or makes a variation.

G6 calculates the individual replicate ratio;

Concrete computational process is as follows:

The individual replicate ratio is represented with Fcopy (i), Fcell (i) expression cell scheduling fitness, Cd (i) expression bulk concentration.

$\mathrm{Fcopy}\left(i\right)=\frac{\frac{\mathrm{Fcell}\left(i\right)}{\mathrm{Cd}\left(i\right)}}{\underset{i}{\overset{N}{\mathrm{\Σ}}}\frac{\mathrm{Fcell}\left(i\right)}{\mathrm{Cd}\left(i\right)}}$

G7 duplicates;

Specifically generate N second generation individuality according to the individual replicate scale, reproduction process is as follows:

At first calculate each individual cumulative probability: by ordering from small to large, Fcopy then adds up with Fcopy.For example, 3 individuality: A (1) are arranged, A (2), A (3), their Fcopy are respectively 0.5,0.3,0.2, accumulated probability B (1) then, B (2), B (3) is respectively 0.5,0.8,1.0;

Secondly at [0,1] interval interior equally distributed pseudo random number r that produces;

If r＜B (1) then selects individual A (1); Otherwise, select to satisfy the following individual k that concerns, i.e. B (k-1)＜r＜=B (k).

G8 exchange pairing;

Second generation individuality carries out exchange pairing, generates N third generation individuality, and is specific as follows:

With second generation individuality is male parent, carry out individual pairing according to certain exchange probability (0.8～0.9), matching method: two individualities choosing are exchanged 1 cell encoding (the binary coding length of each user resource correspondence to be allocated) at random, generate N third generation individuality, if the number of the individual binary coding correspondence that generates does not satisfy the constraints that allows Resources allocation smaller or equal to maximum, then need to exchange again generation.

Exchange probability wherein provided by the invention is (0.8～0.9), but the present invention do not limit this exchange probability, and those skilled in the art can decide according to concrete condition.

The G9 variation;

Third generation individuality makes a variation, and generates N the 4th generation individuality, and is specific as follows:

Third generation individuality is carried out individual variation according to certain variation probability (less than 0.1), variation method: the individuality of choosing is adopted the operation of the negate of figure place at random, and the figure place at random of present embodiment variation is 1, generate the 4th generation individuality, enter step G3; If the number of the individual binary coding correspondence that generates does not satisfy the constraints that allows Resources allocation smaller or equal to maximum, then need the generation that makes a variation again.

Variation probability wherein provided by the invention is (less than 0.1), but the present invention do not limit this variation probability, and those skilled in the art can decide according to concrete condition; In addition figure place at random provided by the invention is 1, same the present invention not to this at random figure place limit, those skilled in the art can decide according to concrete condition.

G10 calculates the resource to be allocated of user in each individual corresponding district;

Computational process is as follows:

Suppose to have 5 users, individual corresponding binary coding is 0,100 0,001 0,001 0,101 0101, ARi such as following table that then final each user distributes:

UE	The ARi encoded radio	The ARi actual value
			1	0100	4
2	0001	1
			3	0001	1
4	0101	5
			5	0101	5

In the table, promptly corresponding each user's of ARi actual value resource to be allocated.

G11 calculates user's scheduling fitness;

Specific as follows:

Suppose that the user's number that participates in dispatching is m, wherein the dispatching priority of user i is Pi (human configuration), it is Hi (drawing by historical statistical data) that history is obtained service speed, application speed is Ri (human configuration), channel circumstance is Ci (passing through apparatus measures), and the fairness tolerance that the user obtains is Fi (passing through apparatus measures); Resource to be allocated is ARi (above-mentioned steps draws).Wherein, Pi, Hi, Ri, Ci, Fi, all normalization.

The formula of particular user scheduling fitness (Fitness (i)) is as follows:

$\mathrm{Fitness}\left(i\right)=\mathrm{Fi}*\frac{\mathrm{Hi}-\mathrm{Ri}}{\mathrm{Ri}}*\mathrm{Ci}*\mathrm{Pi}*\mathrm{ARi}$

M user's resource to be allocated and wherein smaller or equal to maximum resource to be allocated.

A corresponding in esse m user, the resource difference to be allocated in corresponding N the individuality of each user, other parameter is identical.

The scheduling of resource fitness to be allocated of G12 calculation plot, correspondence obtains the ideal adaptation degree;

Specific as follows:

With each user scheduling of resource fitness to be allocated Fitness (i) addition that step F 4 calculates, obtain the scheduling of resource fitness to be allocated (Fcell) of sub-district.

$\mathrm{Fcell}=\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Fitness}\left(i\right)$

G13 repeating step G10, G12, G13 obtains N individual fitness;

G14 obtains the resource to be allocated of maximum scheduling fitness sub-district;

Compare N individual fitness, obtain the individuality of maximum adaptation degree, promptly obtain the resource to be allocated of maximum scheduling fitness sub-district.

G15 calculates the resource to be allocated that maximum is dispatched each user's correspondence in the sub-district of fitness;

The sub-district of the maximum scheduling of resource to be allocated fitness can have maximum scheduling fitness to user's resource to be allocated in should the sub-district.Concrete process is as follows:

Suppose to have 5 users, the individual corresponding binary coding with maximum adaptation degree is 0,110 00,010,000 0,111 0010, ARi such as following table that then final each user distributes:

UE	The ARi encoded radio	The ARi actual value
			1	0110	6
2	0001	1
			3	0000	0
4	0111	7
			5	0010	2

In the table, promptly corresponding each user's of ARi actual value resource to be allocated.So far this finishing scheduling is waited for scheduling next time.

Increase the process of calculating user's resource to be allocated in the present embodiment, the binary coding of the individuality that finally obtains by genetic algorithm is the binary coding of respective cell user resource to be allocated, and the pairing actual value of the binary coding of each user's correspondence is this user's resource numerical value to be allocated.

Seeing also Fig. 5 is that system first provided by the invention implements illustration.

As shown in the figure: obtain Resource Unit 100 to be allocated, be used to set up sub-district and individual corresponding relation, and utilize genetic algorithm to obtain the resource to be allocated of maximum scheduling fitness sub-district;

User's resource computing unit 200 to be allocated is used for calculating maximum resource to be allocated of dispatching fitness each user of sub-district.

Obtaining Resource Unit 100 to be allocated comprises:

Corresponding relation unit 110 is used to set up community user resource to be allocated and individual corresponding relation;

Heredity unit 120 is used for the corresponding relation according to the foundation of corresponding relation unit, utilizes genetic algorithm to obtain the sub-district resource to be allocated of maximum scheduling fitness.

Heredity unit 120 comprises:

Obtain maximum adaptation degree individual cell 121, be used for obtaining maximum adaptation degree individuality according to the corresponding relation that the corresponding relation unit is set up, correspondence is obtained the resource to be allocated of maximum scheduling fitness sub-district;

Seeing also Fig. 6 is that system second provided by the invention implements illustration.

As shown in the figure: obtain Resource Unit 100 to be allocated, be used to set up sub-district and individual corresponding relation, and utilize genetic algorithm to obtain the resource to be allocated of maximum scheduling fitness sub-district;

User's resource computing unit 200 to be allocated is used for calculating maximum resource to be allocated of dispatching fitness each user of sub-district.

Obtaining Resource Unit 100 to be allocated comprises:

Corresponding relation unit 110 is used to set up community user resource to be allocated and individual corresponding relation;

Heredity unit 120 is used for the corresponding relation according to the foundation of corresponding relation unit, utilizes genetic algorithm to obtain the sub-district resource to be allocated of maximum scheduling fitness.

Heredity unit 120 comprises:

Obtain maximum adaptation degree individual cell 121, be used for obtaining maximum adaptation degree individuality according to the corresponding relation that the corresponding relation unit is set up, correspondence is obtained the resource to be allocated of maximum scheduling fitness sub-district;

Individual bulk concentration unit 122 is used to calculate a bulk concentration, and relatively individual bulk concentration and threshold value size, if a bulk concentration is greater than threshold value, inform that then obtaining maximum adaptation degree individual cell obtains maximum adaptation degree individuality, if bulk concentration, is then informed the unit that the adds up number of repetition that adds up less than threshold value;

Unit 123 adds up, be used to the number of repetition that adds up, relatively number of repetition is big or small with maximum repetition, if number of repetition is greater than maximum number of repetitions, inform that then obtaining maximum adaptation degree individual cell obtains maximum adaptation degree sub-district, obtain the sub-district resource to be allocated of maximum scheduling fitness, if number of repetition, informs then that the individual replicate unit calculates the individual replicate ratio less than maximum number of repetitions;

Individual replicate unit 124 is used to use cell scheduling fitness, individual bulk concentration, calculates the individual replicate ratio;

The unit 125 that duplicates, exchanges, makes a variation is used for duplicating and exchange pairing, variation according to the individual replicate ratio, generates individual.

More than the method and system of a kind of resource allocation provided by the present invention are described in detail, used specific case herein principle of the present invention and execution mode are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.

Claims (1)

1. the method for a resource allocation is characterized in that, comprises step:

1) set up community user resource to be allocated and individual corresponding relation, utilize genetic algorithm, according to

Calculate user's scheduling of resource fitness to be allocated, wherein, ARi is that dispatching priority of users, Hi are for history is obtained service speed, Ri is the fairness tolerance that channel circumstance, Fi obtain for the user for application speed, Ci for resource to be allocated, the Pi of user i;

With each user scheduling of resource fitness to be allocated addition that calculates, obtain the scheduling of resource fitness to be allocated of sub-district, promptly obtain the ideal adaptation degree;

Compare each individual fitness, obtain the individuality of maximum adaptation degree, promptly obtain the resource to be allocated of maximum scheduling fitness sub-district;

2) each user's corresponding district resource to be allocated is identical figure place binary coding in the maximum scheduling fitness sub-district; Calculate the number of the binary coding correspondence of each user's correspondence, obtain each user's resource to be allocated; The described corresponding relation of setting up community user resource to be allocated and individuality comprises:

Set up community user resource binary coding to be allocated and individual binary-coded corresponding relation;

The number that each user of described sub-district resource binary to be allocated coding is corresponding and allow resource to be allocated smaller or equal to maximum;

Obtain the resource to be allocated of described user i in the following manner:

11) generate N individuality at random, described individual corresponding binary coding of producing at random satisfies condition number and allow resource to be allocated smaller or equal to maximum of each user resource binary coding to be allocated correspondence;

14) calculate the number of the binary coding correspondence of user's correspondence in each individual corresponding district, obtain the resource ARi to be allocated of user i;

Between described step 11) and step 14), also comprise:

111) calculate a bulk concentration, if a bulk concentration then enters step 14), otherwise enters step 112 greater than threshold value), wherein, described bulk concentration be and should individuality between affinity greater than the individual number of presetting the affinity constant and total number of individuals purpose quotient;

112) accumulation step 114) number of repetition, if number of repetition is greater than maximum number of repetitions, then enters step 14), otherwise enter step 113);

113) use cell scheduling fitness, individual bulk concentration, calculate the individual replicate ratio;

114) duplicate according to the individual replicate ratio and exchange pairing, variation, generate individually, enter step 111).

Images