CN103781166A - Mobile terminal power distribution method in heterogeneous wireless network cooperative communication system - Google Patents

Abstract

The invention discloses a mobile terminal power distribution method in a heterogeneous wireless network cooperative communication system. The method comprises the following steps: the step 1 in which the mobile terminal is used to detect that L relay nodes which can be used to provide the cooperative communication service exist; the step 2 in which a particle population representing the power distribution scheme is generated; the step 3 in which a fitness function of the power distribution scheme is defined; the step 4 in which initial fitness values of particles are calculated; the step 5 in which particle individual extremum value and global extremum value initialization is performed; the step 6 in which a new generation of particle population is generated; the step 7 in which the fitness value of each particle in the new generation of particle population is recalculated; the step 8 in which particles are selected to undergo mutation and the mutated particles are judged whether to be legal; the step 9 in which the particle individual extremum value and global extremum value are updated; and the step 10 in which the global extremum value is selected as the transmitting power distribution scheme of the mobile terminal. According to the invention, the particle swarm algorithm is used to distribute the transmitting power of the mobile terminal in the heterogeneous cooperative communication system, and at the same time, the receiving end signal-to-noise ratio and the power consumption performance of the mobile terminal can be optimized.

Description

Mobile terminal power distribution method in heterogeneous wireless network cooperation communication system

Technical field

The invention belongs to wireless communication technology field, be specifically related to a kind of mobile terminal power distribution method based on particle cluster algorithm in heterogeneous wireless network cooperation communication system.

Background technology

At present, in wireless communications environment, there is multiple heterogeneous network, as GSM, CDMA, Wi-Fi, UWB, bluetooth etc.But these network technologies differ from one another, each heterogeneous network works alone, and mobile terminal can only access a kind of network at one time, can not make full use of idle Internet resources.Heterogeneous wireless network cooperative communication technology can be realized the mutual amalgamation and collaboration between multiple heterogeneous networks, make mobile terminal can utilize multiple heterogeneous network transmission data simultaneously, make full use of idle Internet resources, optimize the performance of whole communication system, therefore caused gradually domestic and international researcher's attention.

In heterogeneous wireless network cooperation communication system, in the time of mobile terminal and receiving terminal apart from each other, mobile terminal can by utilizing after packet shunting, multiple isomery collaboration relay node are parallel be transmitted to receiving terminal, thereby improves transmission quality and transmission rate simultaneously.But, mobile terminal utilizes multiple isomery relay node cooperation transmission data simultaneously, in obtaining better transmission quality and higher transmission rates, need to consume more power, how making mobile terminal consume alap power in obtaining high as far as possible transmission performance is a key issue, needs the power of terminal to divide effectively and rationally on each repeated link to be equipped with the object that reaches the performance aspect transmission performance and the power consumption two of simultaneously optimizing terminal for this reason.

The current existing much research about wireless cooperation communication system power distributing technique aspect, but great majority research mainly concentrates on the power division of homogeneous network, consideration via node rather than mobile terminal, and take single performance index as optimization aim.Publication number is CN102006654A, open day is the patent of invention " method and system of power division in a kind of many relay cooperative communications " on April 6th, 2011, providing according to channel gain is user terminal selecting via node the method for via node being carried out to power division, but the transmit power allocations problem of the mobile terminal in communication process is not considered in this invention, and plant the homogeneous network of standard and only take the transmission rate of user terminal as optimization aim for single.Publication number is CN101588627A, open day is the patent of invention " method of the power optimized co-allocation of source and via node in collaboration communication " on November 25th, 2009, what consider is under the prerequisite that the power sum of source node and via node is certain, adopt water-filling algorithm to carry out power division to source and via node, but this invention is equally for single homogeneous network, and only turn to optimization aim with power system capacity maximum, there is no to consider the optimization to mobile terminal (source node) power consumption performance.

Summary of the invention

The shortcoming that the object of the invention is to overcome prior art is with not enough, mobile terminal emitting power distribution method in a kind of heterogeneous wireless network cooperation communication system is provided, for the mobile terminal reasonable distribution transmitting power in heterogeneous wireless network cooperation communication system, take the power consumption of mobile terminal and the signal to noise ratio of receiving terminal as optimization aim, in reducing mobile terminal power consumption as far as possible, make receiving terminal obtain high as far as possible signal to noise ratio.

Object of the present invention is achieved through the following technical solutions:

Mobile terminal power distribution method in heterogeneous wireless network cooperation communication system, comprises the steps:

Total L the via node that can provide collaboration communication to serve of periphery is provided for step 1, mobile terminal, therefrom selects M the good via node access of channel condition according to the channel condition information of each repeated link;

Step 2, generation represent the particle population of power allocation scheme, produce N particle, and N is population scale, and wherein k particle is defined as Particle _k=(X _k, V _k) (k=1 ..., N), X _kbe the position of particle, represent k kind power allocation scheme, V _kthe movement velocity of particle, X _kand V _kbe and contain M, M is the one dimension real number array that via node is counted element, in addition, and N one dimension real number array X that contains M element of definition _{k_pbest}(k=1,2 ..., N) and the individual extreme value that is used for storing particle, define 1 one dimension real number array X that contains M element _gbestbe used for storing the global extremum of particle population;

The fitness function of step 3, definition power allocation scheme, using receiving terminal signal to noise ratio and mobile terminal power consumption as optimization aim, particle Particle _kfitness function be defined as wherein, γ _kfor receiving terminal signal to noise ratio, for mobile terminal is distributed to the performance number sum of each via node, α and β are for adjusting parameter, and each via node adopts the mode of amplifying forward pass to forward the packet from mobile terminal simultaneously;

Step 4, according to the initial fitness value of each particle described in the fitness function calculation procedure 2 described in step 3 $F_k^0=f\left(X_k^0\right)(k=\mathrm{1,2},\·\·\·,N),$ Wherein

it is the initial position of particle;

Step 5, by the initial position of each particle

be stored in individual extreme value X separately _{k_pbest}(k=1,2 ..., N) in, the initialization of individual extreme value completed; The initial individual extreme value of fitness value maximum is stored in to global extremum X _gbestin, complete the initialization of global extremum;

Step 6, by the speed V of new particle more _kwith position X _k, producing particle population of new generation, j is for the speed of k particle in particle population

and position

respectively by search equation ${V_k}^j=w{V_k}^{j-1}+c_1{R}_1(X_{k\_\mathrm{pbest}}^{j-1}-X_k^{j-1})+c_2{R}_2(X_{\mathrm{gbest}}^{j-1}-X_k^{i-1})$ With $X_k^j=X_k^{j-1}+{V_k}^j,(1\≤j\≤T)$ Produce, wherein

for Particle _kthe individual extreme value of the j time iteration,

for the global extremum of the j time iteration of particle population, w is the inertia weight factor, c ₁, c ₂for the study factor, R ₁, R ₂for the random number in (0,1) interval, T is greatest iteration number;

Step 7, recalculate the fitness value of the each particle in particle population of new generation according to the fitness function described in step 3 ${F_k}^j=f\left(X_k^j\right)(k=\mathrm{1,2},\·\·\·,N),(1\≤j\≤T);$

Step 8, choose the particle that in particle population, fitness value is lower and make a variation, judge that whether the particle after variation is legal, if variation particle is legal particle, replaces primary particle with variation particle, and calculate its fitness value; If variation particle is illegal particle, reject variation particle, retain original particle; Described illegal particle refers to the particle with following arbitrary feature: the performance number that (1) is distributed i via node

make signal to noise ratio lower than the required signal-noise ratio threshold value of this via node proper communication; (2)

or

described legal particle refers to the particle without above-mentioned arbitrary feature;

Step 9, the more individual extreme value of new particle and global extremum, through type

the more individual extreme value of new particle, the fitness value F of even current particle _k ^jbe better than the individual extreme value of its previous generation

fitness value

individual extreme value is updated to the position of current particle

otherwise individual extreme value is constant; Through type upgrade global extremum, wherein, position is

particle fitness value meet $F_{\mathrm{kk}\_\mathrm{pbest}}^j=\max \{F_{1\_\mathrm{pbest}}^j,\·\·\·\·\·\·,F_{N\_\mathrm{pbest}}^j\},$ It is the individual extreme value that global extremum is updated to fitness value maximum;

Step 10, iterations increase 1(j=j+1), repeating step 6～9, until the maximum adaptation degree value of particle population

the continuous several times of growth be less than given threshold value l _th, or iterations reaches maximum of T (j=T); Now select global extremum

as the transmit power allocations scheme of mobile terminal, power distribution algorithm finishes.

Preferably, in step 1, mobile terminal detects that the total number of the via node of periphery is L, and each via node only has an antenna that is used for sending and receiving; Described channel condition information represents with channel gain, supposes that mobile terminal is designated as h to the channel gain of i via node _i, i via node is designated as g to the channel gain of receiving terminal _i, from all L via node, selective channel gain meets h _i>=h _{i, th}and g _i>= _{gi, th}the access of M via node, wherein h _{i, th}with _{gi, th}that i article of repeated link proper communication requires channel gain threshold value.

Preferably, in step 2, the value of particle population scale N is definite by precision and the comprehensive measurement of complexity of algorithm, and N is larger, and the precision of algorithm is higher, but complexity is also higher; N is less, and the precision of algorithm is lower, but complexity is also lower; The initial position array of particle

element by random number in scope is filled, wherein P _maxfor the rated power of mobile terminal; To initial position array

in element add a constraints: when

or

time, give up this array, regenerate new array; Produce in the same way the initial velocity array of particle

Preferably, in step 2, described individual extreme value X _{k_pbest}refer to particle Particle _kin all previous iterative process the position of fitness value maximum (optimal location),

represent the initial value of individual extreme value,

represent by the j time iteration particle Particle _kindividual extreme value; Described global extremum X _gbestrefer to the individual extreme value of fitness value maximum in particle population,

represent the global extremum initial value of particle population,

represent the particle population global extremum of the j time iteration of cut-off, the optimal location searching by the j time iteration particle population.

Preferably, in step 3, described γ _krepresent the receiving terminal signal to noise ratio of k kind power allocation scheme, γ _kprovided by following formula: ${\mathrm{\γ}}_k={\left(\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\frac{\sqrt{X_k\left[i\right]}\·\sqrt{P_i}\·\left|h_i{g}_i\right|}{\sqrt{N_0+{\left|h_i\right|}^2\·X_k\left[i\right]}}\right)}^2/(1+\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\frac{P_i{\left|g_i\right|}^2}{N_0+{\left|h_i\right|}^2{X}_k\left[i\right]}),$ Wherein,

be the transmitting power of i via node, N ₀for noise variance; The present invention is optimization system received signal to noise ratio and two indexs of mobile terminal power consumption simultaneously: received signal to noise ratio γ _klarger, via node performance number summation

less, the fitness function value obtaining is larger, and corresponding power allocation scheme performance is more excellent; Otherwise power allocation scheme performance is poorer.

Preferably, in step 3, α and β are adjustment parameter, α >0, and β >0, α and β value can be according to the specific requirement of receiving terminal signal to noise ratio and mobile terminal power consumption is regulated; In the time that α is equal with β value, illustrates and give receiving terminal signal to noise ratio and these two same attention degrees of optimization index of mobile terminal power consumption; In the time that α value is greater than β, the attention degree that the attention degree of receiving terminal signal to noise ratio is greater than to mobile terminal power consumption is described, otherwise, in like manner.

Preferably, in step 4, described Particle _k=(X _k, V _k) (k=1 ..., N) initial fitness value be by defined fitness function f (X in step 3 _k) at initial position

time function value; In like manner, particle Particle _k=(X _k, V _k) (k=1 ..., N) and fitness value after the j time iteration is

fitness value is the standard for passing judgment on power allocation scheme quality, and fitness value is higher, and corresponding power allocation scheme performance is better; Otherwise, poorer.

Preferably, in step 5, by the initial position of particle

directly be stored in individual extreme value separately

in, as the initial value of individual extreme value, by initial fitness value

maximum individual extreme value is stored in global extremum

in, as the initial value of global extremum,

wherein, position is

particle fitness value meet $F_{\mathrm{kk}\_\mathrm{pbest}}^0=\max \{F_{1\_\mathrm{pbest}}^0,\·\·\·\·\·\·,F_{N\_\mathrm{pbest}}^0\}.$

Preferably, in step 6, w is the inertia weight factor, for regulating the reserving degree of the original speed of particle; c ₁, c ₂for the study factor, for regulating particle to fly to the step-length of personal best particle and global optimum position, c in classical particle population algorithm ₁, c ₂interval be (0,4), set c herein ₁=c ₂=2, R ₁, R ₂for the random number in (0,1) interval; The inertia weight factor is provided by following formula: w=w _max/ (1+e ^{-λ η}), in formula, λ is for adjusting parameter, for adjusting the variation speed of w value; w _maxthe parameter for adjusting inertia weight factor size, w _maxwhen value is larger, be conducive to improve ability of searching optimum, but convergence precision is lower, w _maxwhile getting smaller value, be conducive to improve convergence precision, but be easily absorbed in local optimum;

wherein $F_{\mathrm{gbest}}^{j-1}=\max \{F_{1\_\mathrm{pbest}}^{j-1},\·\·\·\·\·\·,F_{N\_\mathrm{pbest}}^{j-1}\},$ J-1 is for the fitness value of the global extremum of particle population; When the fitness value of particle

more hour, η is larger, and w is larger, regulates the size of the inertia weight factor according to the fitness value of particle, makes the particle that fitness value is less obtain stronger ability of searching optimum; In like manner, make the particle that fitness value is larger obtain stronger local search ability, restrain to extreme value.

Preferably, in step 8, the variation of particle mainly contains following steps:

Step 8.1, chooses particle to be made a variation; The choosing method of particle to be made a variation is: by the fitness value size of particle by descending, choose fitness value in particle population less after

individual particle is as particle to be made a variation, fitness value larger before

individual particle remains unchanged;

Step 8.2, to the particle to be made a variation of choosing ${\mathrm{Particle}}_k^j=(X_k^j,V_k^j)$ With probability $p_k^j=\frac{|F_k^j-F_{\mathrm{avg}}^j|}{F_{\max }^j}$ Make a variation, wherein

be the average fitness value of j for particle in particle population,

be the maximum adaptation degree value of j for particle in particle population;

Step 8.3, the variation method of particle to be made a variation is: to the particle to be made a variation of choosing random its position array of selecting element in one or several element make a variation, the element after variation

by formula $X_k^{j^{\′}}\left[i\right]=X_k^j\left[i\right]+\mathrm{\Δ}\·P_{\max }\·\mathrm{rand}(-\mathrm{1,1})$ Produce, wherein, △ is that power is adjusted step-length, is the real number between 0～1, and the less arithmetic accuracy of value of △ is higher, but convergence of algorithm speed also can be slack-off, and △ is larger, and arithmetic accuracy is lower, and convergence rate accelerates; Particle ${\mathrm{Particle}}_k^j=(X_k^j,V_k^j)$ After variation, be denoted as ${\mathrm{Particle}}_{k^{\′}}^j=(X_{k^{\′}}^j,V_{k^{\′}}^j).$

The present invention has following advantage and effect with respect to prior art:

1, in the present invention mobile terminal utilize the multiple heterogeneous nodes of periphery as relaying (as used the mobile terminal of the heterogeneous networks such as Wi-Fi, bluetooth, 3G around, each mobile terminal possesses the function as relay station) carry out cooperation transmission data, the in the situation that of, Bandwidth-Constrained congested in single network, can make full use of the resource of other idle network around, improve the transmission rate of data.

2, the packet that in the present invention, mobile terminal sends sends to each isomery via node after data distribution, by multiple isomery relay node cooperations transmission data, the in the situation that of transmitting terminal and receiving terminal apart from each other, dtr signal, can effectively improve the communication quality of system.

3, the present invention uses particle cluster algorithm to distribute the transmitting power of the mobile terminal in isomery cooperation communication system, take the power consumption of mobile terminal and the signal to noise ratio of receiving terminal as optimization aim, can optimize the power consumption performance of receiving terminal signal to noise ratio and mobile terminal simultaneously.

4, the present invention uses a kind of power distribution method based on improved particle swarm optimization algorithm, in population renewal process, add Variation mechanism, avoid the result of algorithm search to be absorbed in local optimum, more be conducive to the direction convergence of population to globally optimal solution, transmitting power to mobile terminal in heterogeneous wireless network cooperation communication system is carried out reasonable distribution, better the receiving terminal signal to noise ratio of optimization system and mobile terminal power consumption performance.

5, the improved particle swarm optimization algorithm of one that the present invention uses, in population renewal process, the less particle of fitness value is made a variation, guarantee the diversity of population in renewal process, can avoid population to converge on prematurely local optimum, make search procedure constantly be tending towards the region of fitness value maximum.

6, the present invention, with the position array representation power allocation scheme of population, carries out iteration by the position array of population and upgrades searching extreme value, and its algorithmic procedure complexity is low, easy operating, and practicality is high.

Accompanying drawing explanation

Fig. 1 is many relayings of the present invention power division flow chart.

Fig. 2 is many relayings of the present invention Distributed Power Architecture schematic diagram.

Embodiment

Below in conjunction with embodiment and accompanying drawing, the present invention is described in further detail, but embodiments of the present invention are not limited to this.

Embodiment

Implementation process of the present invention mainly comprises two parts, access control (via node that selection will access) and the transmitting power of mobile terminal is distributed.

In this embodiment, Figure 1 shows that mobile terminal emitting power allocation flow figure, Figure 2 shows that the mobile terminal emitting power in heterogeneous wireless network cooperation communication system distributes and data transfer mode.

In the present embodiment, the performing step of access control is as follows:

Step 1, heterogeneous wireless network model as shown in Figure 2, intelligent radio access point (AP) carries out Real-Time Monitoring to the channel statistic of each repeated link.

Step 2, the mobile terminal shown in Fig. 2 is to the channel statistic of each repeated link of intelligent AP acquisition request, and the channel statistic of each repeated link is sent to mobile terminal by intelligent AP.

Step 3, mobile terminal has been learnt L the available free channel of via node by intelligent AP, supposes that mobile terminal is designated as h to the channel gain of i via node _i, i via node is designated as g to the channel gain of receiving terminal _i.Suppose that in L via node, channel gain meets h _i>=h _{i, th}and g _i>= _{gi, th}totally 10 of via nodes, wherein h _{i, th}with _{gi, th}be the i article of desired channel gain threshold value of repeated link proper communication, these 10 via nodes are selected as cooperating relay, are mobile terminal parallel forwarding data.

The assigning process of the transmitting power to mobile terminal is as follows:

Step 1, the one dimension real number array X(particle position that definition contains 10 elements) and V(particle rapidity), wherein X is as the transmit power allocations scheme of mobile terminal, element value X[i] (i=1,2 ..., 10) and represent that this scheme distributes to the performance number of i via node.In like manner, in V, also comprise 10 element V[i] (i=1,2 ..., 10).

Step 2, the fitness function of definition power allocation scheme wherein $\mathrm{\γ}={\left(\underset{i=1}{\overset{10}{\mathrm{\Σ}}}\frac{\sqrt{X\left[i\right]}\·\sqrt{P_i}\·\left|h_i{g}_i\right|}{\sqrt{N_0+{\left|h_i\right|}^2\·X\left[i\right]}}\right)}^2/(1+\underset{i=1}{\overset{10}{\mathrm{\Σ}}}\frac{P_i{\left|g_i\right|}^2}{N_0+{\left|h_i\right|}^{2}X\left[i\right]}),$ N ₀for noise variance, be made as in the present embodiment 1, X[i] represent that mobile terminal distributes to the performance number of each via node,

be the transmitting power of i via node, P _maxfor the rated power of mobile terminal, in the present embodiment, establish P _max=120.Via node adopts the mode data of transmitting mobile terminal simultaneously of amplifying forward pass.In the present embodiment, establish α=β, represent system received signal to noise ratio and power consumption to give equal attention.

Step 3, the array of the expression power allocation scheme described in generation step 1, process is as follows: use

the element X[1 of position array X described in random number filling step 1 in scope], X[2] ... X[10], if

or

reuse random number and fill array; The element V[1 of speed array V described in filling step 1 in the same way], V[2] ... V[10].

Step 4, produces the particle population that represents multiple power allocation schemes.Utilize the method for step 3, produce N particle that represents power allocation scheme, its initial position array and number of speed group are respectively

with

wherein N is population scale, and in the present embodiment, setting population scale is N=100.

Step 5, the position array of 100 particles that step 4 is produced

be stored in respectively the initial value of the individual extreme value of middle conduct; Relatively

fitness value draw the individual extreme value with maximum adaptation degree value, be stored in

the middle initial value as global extremum.

Step 6, the more speed of new particle and position, produces particle population of new generation.By search equation ${V_k}^j=w{V_k}^{j-1}+c_1{R}_1(X_{k\_\mathrm{pbest}}^{j-1}-X_k^{j-1})+c_2{R}_2(X_{\mathrm{gbest}}^{j-1}-X_k^{i-1}),(1\≤j\≤T;k=\mathrm{1,2},\·\·\·,100)$ The more speed of new particle, by

the more position of new particle.Wherein,

be respectively Particle _kthe speed of the j time iteration and position, for Particle _kthe individual extreme value of the j time iteration,

for the global extremum of the j time iteration of population, T is greatest iteration number, establishes T=1000 in the present embodiment, and w is the inertia weight factor, and its expression formula is provided by following formula: w=w _max/ (1+e ^{-λ η}), in formula $\mathrm{\η}=\frac{F_{\mathrm{gbest}}^{j-1}-F_k^{j-1}}{F_{\mathrm{gbest}}^{j-1}},(1\≤j\≤1000),$ Wherein $F_{\mathrm{gbest}}^{j-1}=\max \{F_{1\_\mathrm{pbest}}^{j-1},\·\·\·\·\·\·,F_{100\_\mathrm{pbest}}^{j-1}\},$ In the present embodiment, get w _max=1.5, λ=2.5, c ₁, c ₂for the study factor, get c ₁=c ₂=2, R ₁, R ₂for the random number in (0,1) interval.

Step 7, recalculates the fitness value of each particle according to the fitness function described in step 2 ${F_k}^j=f\left(X_k^j\right)(1\≤j\≤1000;k=\mathrm{1,2},\·\·\·,100).$

Step 8, chooses particle the particle population of new generation producing make a variation from step 6, mainly contains following steps:

Step 8.1, chooses particle to be made a variation.The choosing method of particle to be made a variation is: by the fitness value size of particle by descending, choose fitness value in particle population less after

individual particle is as particle to be made a variation, fitness value larger before

individual particle remains unchanged.

Step 8.2, makes a variation to the particle to be made a variation of choosing, and concrete grammar is: to the particle to be made a variation of choosing

random its position array of selecting

element

in one or several element make a variation, the element after variation

by formula

produce, wherein, rand (1,1) is the random number in (1,1).Particle

after variation, be denoted as ${\mathrm{Particle}}_{k^{\′}}^j=(X_{k^{\′}}^j,V_{k^{\′}}^j).$

Step 8.3, the particle after judgement variation

whether be legal particle, if ${\mathrm{Particle}}_{k^{\′}}^j=(X_{k^{\′}}^j,V_{k^{\′}}^j)$ For legal particle, j for population in ${\mathrm{Particle}}_{k^{\′}}^j=(X_{k^{\′}}^j,V_{k^{\′}}^j)$ Replace ${\mathrm{Particle}}_k^j=(X_k^j,V_k^j),$ And calculate the fitness value of particle after variation; If ${\mathrm{Particle}}_{k^{\′}}^j=(X_{k^{\′}}^j,V_{k^{\′}}^j)$ For illegal particle, reject ${\mathrm{Particle}}_{k^{\′}}^j=(X_{k^{\′}}^j,V_{k^{\′}}^j),$ J retains for population ${\mathrm{Particle}}_k^j=(X_k^j,V_k^j).$

In step 8.1, to the particle to be made a variation of choosing the probability of its variation is wherein

be the average fitness value of j for particle in particle population,

be the maximum adaptation degree value of j for particle in particle population.

In step 8.3, described illegal particle refers to the particle with following arbitrary feature: the performance number that (1) is distributed i via node

make signal to noise ratio lower than the required signal-noise ratio threshold value of this via node proper communication; (2)

or

described legal particle refers to the particle without above-mentioned arbitrary feature.

Step 9, the more individual extreme value of new particle and global extremum.Through type

the more individual extreme value of new particle; Through type

upgrade global extremum, wherein, position is

particle fitness value meet $F_{\mathrm{kk}\_\mathrm{pbest}}^j=\max \{F_{1\_\mathrm{pbest}}^j,\·\·\·\·\·\·,F_{100\_\mathrm{pbest}}^j\}.$

Step 10, iterations increase 1(j=j+1), repeating step 6～9, when continuous 5 generation particle population maximum adaptation degree value

growth be less than 5%, or iterations j=T, power distribution algorithm finishes, and selects global extremum now

as the allocative decision of mobile terminal emitting power, the power division value of each via node is respectively

Above-described embodiment is preferably execution mode of the present invention; but embodiments of the present invention are not restricted to the described embodiments; other any do not deviate from change, the modification done under Spirit Essence of the present invention and principle, substitutes, combination, simplify; all should be equivalent substitute mode, within being included in protection scope of the present invention.

Claims (10)

1. the mobile terminal power distribution method in heterogeneous wireless network cooperation communication system, is characterized in that, comprises the steps:

Total L the via node that can provide collaboration communication to serve of periphery is provided for step 1, mobile terminal, therefrom selects M the good via node access of channel condition according to the channel condition information of each repeated link;

Step 2, generation represent the particle population of power allocation scheme, produce N particle, and N is population scale, and wherein k particle is defined as Particle _k=(X _k, V _k) (k=1 ..., N), X _kbe the position of particle, represent k kind power allocation scheme, V _kthe movement velocity of particle, X _kand V _kbe and contain M, M is the one dimension real number array that via node is counted element, in addition, and N one dimension real number array X that contains M element of definition _{k_pbest}(k=1,2 ..., N) and the individual extreme value that is used for storing particle, define 1 one dimension real number array X that contains M element _gbestbe used for storing the global extremum of particle population;

The fitness function of step 3, definition power allocation scheme, using receiving terminal signal to noise ratio and mobile terminal power consumption as optimization aim, particle Particle _kfitness function be defined as

wherein, γ _kfor receiving terminal signal to noise ratio,

for mobile terminal is distributed to the performance number sum of each via node, α and β are for adjusting parameter, and each via node adopts the mode of amplifying forward pass to forward the packet from mobile terminal simultaneously;

Step 4, according to the initial fitness value of each particle described in the fitness function calculation procedure 2 described in step 3 $F_k^0=f\left(X_k^0\right)(k=\mathrm{1,2},\·\·\·,N),$ Wherein

it is the initial position of particle;

Step 5, by the initial position of each particle

be stored in individual extreme value X separately _{k_pbest}(k=1,2 ..., N) in, the initialization of individual extreme value completed; The initial individual extreme value of fitness value maximum is stored in to global extremum X _gbestin, complete the initialization of global extremum;

Step 6, by the speed V of new particle more _kwith position X _k, producing particle population of new generation, j is for the speed V of k particle in particle population _k ^jwith position X _k ^jrespectively by search equation ${V_k}^j=w{V_k}^{j-1}+c_1{R}_1(X_{k\_\mathrm{pbest}}^{j-1}-X_k^{j-1})+c_2{R}_2(X_{\mathrm{gbest}}^{j-1}-X_k^{i-1})$ With $X_k^j=X_k^{j-1}+{V_k}^j,(1\≤j\≤T)$ Produce, wherein

for Particle _kthe individual extreme value of the j time iteration,

for the global extremum of the j time iteration of particle population, w is the inertia weight factor, c ₁, c ₂for the study factor, R ₁, R ₂for the random number in (0,1) interval, T is greatest iteration number;

Step 7, recalculate the fitness value of the each particle in particle population of new generation according to the fitness function described in step 3 ${F_k}^j=f\left(X_k^j\right)(k=\mathrm{1,2},\·\·\·,N),(1\≤j\≤T);$

Step 8, choose the particle that in particle population, fitness value is lower and make a variation, judge that whether the particle after variation is legal, if variation particle is legal particle, replaces primary particle with variation particle, and calculate its fitness value; If variation particle is illegal particle, reject variation particle, retain original particle; Described illegal particle refers to the particle with following arbitrary feature: the performance number that (1) is distributed i via node

make signal to noise ratio lower than the required signal-noise ratio threshold value of this via node proper communication; (2)

or

described legal particle refers to the particle without above-mentioned arbitrary feature;

Step 9, the more individual extreme value of new particle and global extremum, through type

the more individual extreme value of new particle, the fitness value F of even current particle _k ^jbe better than the individual extreme value of its previous generation

fitness value

individual extreme value is updated to the position of current particle

otherwise individual extreme value is constant; Through type

upgrade global extremum, wherein, position is particle fitness value meet $F_{\mathrm{kk}\_\mathrm{pbest}}^j=\max \{F_{1\_\mathrm{pbest}}^j,\·\·\·\·\·\·,F_{N\_\mathrm{pbest}}^j\},$ It is the individual extreme value that global extremum is updated to fitness value maximum;

Step 10, iterations increase 1(j=j+1), repeating step 6～9, until the maximum adaptation degree value of particle population

the continuous several times of growth be less than given threshold value l _th, or iterations reaches maximum of T (j=T); Now select global extremum

as the transmit power allocations scheme of mobile terminal, power distribution algorithm finishes.

2. the mobile terminal power distribution method in heterogeneous wireless network cooperation communication system according to claim 1, it is characterized in that, in step 1, mobile terminal detects that the total number of the via node of periphery is L, and each via node only has an antenna that is used for sending and receiving; Described channel condition information represents with channel gain, supposes that mobile terminal is designated as h to the channel gain of i via node _i, i via node is designated as g to the channel gain of receiving terminal _i, from all L via node, selective channel gain meets h _i>=h _{i, th}and g _i>=g _{i, th}the access of M via node, wherein h _{i, th}with _{gi, th}that i article of repeated link proper communication requires channel gain threshold value.

3. the mobile terminal power distribution method in heterogeneous wireless network cooperation communication system according to claim 1, is characterized in that, in step 2, the value of particle population scale N is determined by precision and comprehensive measurement of complexity of algorithm, N is larger, and the precision of algorithm is higher, but complexity is also higher; N is less, and the precision of algorithm is lower, but complexity is also lower; The initial position array of particle

element by

random number in scope is filled, wherein P _maxfor the rated power of mobile terminal; To initial position array in element add a constraints: when

or

time, give up this array, regenerate new array; Produce in the same way the initial velocity array of particle

4. the mobile terminal power distribution method in heterogeneous wireless network cooperation communication system according to claim 1, is characterized in that, in step 2, and described individual extreme value X _{k_pbest}refer to particle Particle _kthe position of fitness value maximum in all previous iterative process,

represent the initial value of individual extreme value,

represent by the j time iteration particle Particle _kindividual extreme value; Described global extremum X _gbestrefer to the individual extreme value of fitness value maximum in particle population,

represent the global extremum initial value of particle population, represent the particle population global extremum of the j time iteration of cut-off, the optimal location searching by the j time iteration particle population.

5. the mobile terminal power distribution method in heterogeneous wireless network cooperation communication system according to claim 1, is characterized in that, in step 3, and described γ _krepresent the receiving terminal signal to noise ratio of k kind power allocation scheme, γ _kprovided by following formula: ${\mathrm{\γ}}_k={\left(\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\frac{\sqrt{X_k\left[i\right]}\·\sqrt{P_i}\·\left|h_i{g}_i\right|}{\sqrt{N_0+{\left|h_i\right|}^2\·X_k\left[i\right]}}\right)}^2/(1+\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\frac{P_i{\left|g_i\right|}^2}{N_0+{\left|h_i\right|}^2{X}_k\left[i\right]}),$ Wherein,

be the transmitting power of i via node, N ₀for noise variance; And while optimization system received signal to noise ratio and two indexs of mobile terminal power consumption: received signal to noise ratio γ _klarger, via node performance number summation

less, the fitness function value obtaining is larger, and corresponding power allocation scheme performance is more excellent; Otherwise power allocation scheme performance is poorer.

6. the mobile terminal power distribution method in heterogeneous wireless network cooperation communication system according to claim 1, it is characterized in that, in step 3, α and β are for adjusting parameter, α >0, β >0, α and β value can be according to the specific requirement of receiving terminal signal to noise ratio and mobile terminal power consumption is regulated; In the time that α is equal with β value, illustrates and give receiving terminal signal to noise ratio and these two same attention degrees of optimization index of mobile terminal power consumption; In the time that α value is greater than β, the attention degree that the attention degree of receiving terminal signal to noise ratio is greater than to mobile terminal power consumption is described, otherwise, in like manner.

7. the mobile terminal power distribution method in heterogeneous wireless network cooperation communication system according to claim 1, is characterized in that, in step 4, and described Particle _k=(X _k, V _k) (k=1 ..., N) initial fitness value be by defined fitness function f (X in step 3 _k) at initial position

time function value; In like manner, particle Particle _k=(X _k, V _k) (k=1 ..., N) and fitness value after the j time iteration is

fitness value is the standard for passing judgment on power allocation scheme quality, and fitness value is higher, and corresponding power allocation scheme performance is better; Otherwise, poorer.

8. the mobile terminal power distribution method in heterogeneous wireless network cooperation communication system according to claim 1, is characterized in that, in step 5, by the initial position of particle

directly be stored in individual extreme value separately

in, as the initial value of individual extreme value,

by initial fitness value

maximum individual extreme value is stored in global extremum

in, as the initial value of global extremum,

wherein, position is

particle fitness value meet $F_{\mathrm{kk}\_\mathrm{pbest}}^0=\max \{F_{1\_\mathrm{pbest}}^0,\·\·\·\·\·\·,F_{N\_\mathrm{pbest}}^0\}.$

9. the mobile terminal power distribution method in heterogeneous wireless network cooperation communication system according to claim 1, is characterized in that, in step 6, w is the inertia weight factor, for regulating the reserving degree of the original speed of particle; c ₁, c ₂for the study factor, for regulating particle to fly to the step-length of personal best particle and global optimum position, c in classical particle population algorithm ₁, c ₂interval be (0,4), set c herein ₁=c ₂=2, R ₁, R ₂for the random number in (0,1) interval; The inertia weight factor is provided by following formula: w=w _max/ (1+e ^{-λ η}), in formula, λ is for adjusting parameter, for adjusting the variation speed of w value; w _maxthe parameter for adjusting inertia weight factor size, w _maxwhen value is larger, be conducive to improve ability of searching optimum, but convergence precision is lower, w _maxwhile getting smaller value, be conducive to improve convergence precision, but be easily absorbed in local optimum; $\mathrm{\η}=\frac{F_{\mathrm{gbest}}^{j-1}-F_k^{j-1}}{F_{\mathrm{gbest}}^{j-1}},(1\≤j\≤T),$ Wherein $F_{\mathrm{gbest}}^{j-1}=\max \{F_{1\_\mathrm{pbest}}^{j-1},\·\·\·\·\·\·,F_{N\_\mathrm{pbest}}^{j-1}\},$ J-1 is for the fitness value of the global extremum of particle population; When the fitness value of particle

more hour, η is larger, and w is larger, regulates the size of the inertia weight factor according to the fitness value of particle, makes the particle that fitness value is less obtain stronger ability of searching optimum; In like manner, make the particle that fitness value is larger obtain stronger local search ability, restrain to extreme value.

10. the mobile terminal power distribution method in heterogeneous wireless network cooperation communication system according to claim 1, is characterized in that, in step 8, the variation of particle mainly contains following steps:

Step 8.1, chooses particle to be made a variation; The choosing method of particle to be made a variation is: by the fitness value size of particle by descending, choose fitness value in particle population less after

individual particle is as particle to be made a variation, fitness value larger before

individual particle remains unchanged;

Step 8.2, to the particle to be made a variation of choosing

with probability

make a variation, wherein

be the average fitness value of j for particle in particle population,

be the maximum adaptation degree value of j for particle in particle population;

Step 8.3, the variation method of particle to be made a variation is: to the particle to be made a variation of choosing random its position array of selecting

element

in one or several element make a variation, the element after variation

by formula

produce, wherein, △ is that power is adjusted step-length, is the real number between 0～1, and the less arithmetic accuracy of value of △ is higher, but convergence of algorithm speed also can be slack-off, and △ is larger, and arithmetic accuracy is lower, and convergence rate accelerates; Particle ${\mathrm{Particle}}_k^j=(X_k^j,V_k^j)$ After variation, be denoted as ${\mathrm{Particle}}_{k^{\′}}^j=(X_{k^{\′}}^j,V_{k^{\′}}^j).$

Images