CN103857027A - Power distribution method in satellite communication system - Google Patents

Abstract

The invention discloses a power distribution method in a satellite communication system. The power distribution method comprises the steps of firstly, determining the relation between the capacity of each user and distributed power according to a satellite link budget equation in the satellite communication system; then modeling the power distribution problem in the satellite communication system into a non-linear convex optimization problem with constraints; finally, on the basis of the duality theorem, obtaining the optimal power distribution scheme through solution of the duality problem of the optimization problem. Compared with a traditional uniform or proportional power distribution method, the power distribution method has the advantage that on the premise that the whole system capacity is kept unchanged, the fairness of power distribution to each user is improved; in addition, due to the fact that the satellite link budget equation is adopted to calculate the capacity distributed to each user, the distribution result has higher practicability in the actual satellite communication system.

Description

Power distribution method in satellite communication system

Technical field

The invention belongs to the technical field that resource is distributed, be specifically related to the power distribution method in a kind of satellite communication system.

Background technology

In modern satellite communications system, due to the restriction of satellite platform, the power resource on star is limited and valuable.In order to better meet user's business demand, must improve the service efficiency of power resource.Due to each user's business applications difference, and the channel condition of different user is also different.Therefore must power resource be given to different users dynamically to improve resource utilization according to user's business applications and channel condition.

Because power division has influence on the performance of whole satellite communication system, therefore it has obtained research widely.As famous " water-filling algorithm ", this algorithm is take maximized system capacity as target, carry out power division according to each user's channel condition, the user that channel condition is good will distribute more power resource, and the user of bad channel conditions is by the power resource distributing still less.But this algorithm has an obvious shortcoming, do not consider each user's business demand, although a user's channel condition is good, its business demand amount is very little, therefore distributes more power resource significantly can cause waste to it; Although a user's channel condition is poor, its business demand is larger, and distributes less power resource significantly can cause unjustness between power division to it.For this reason, the people such as J.P.Choi and V.W.S.Chan are take the quadratic sum of difference that minimizes all customer service demands and partition capacity as optimization aim, power division problem is modeled as to a protruding optimization problem, and has analyzed the impact of different factors for power distribution result.But they do not propose concrete algorithm to address this problem.For this reason, the people such as Yang Hong and Qi Feng is applied to dichotomy and subgradient algorithm to solve this optimization problem respectively.A subject matter in above-mentioned research is, the relation of each user's the capacity that is assigned to and the power resource being assigned to is determined by shannon formula, but Shannon capacity formula cannot be realized in actual satellite communication system, it has theoretical reference meaning, therefore, in above research, power distribution result is not optimum result in actual satellite communication system.For this reason, the people such as Apostolos Destounis determine user's capacity by actual satellite link accounting equation, then, to maximize the number of users that meets capacity needs as criterion, carry out power division.But it does not have the whole volume of taking into account system and the fairness problem for each user assignment power.

The object of the invention is to make up the weak point in above-mentioned research.The link budget equation that the present invention adopts in communication system is via satellite determined user's capacity, take the quadratic sum that minimizes all users' partition capacity and the difference of business applications as criterion, carries out power optimization.The power division criterion that the present invention selects will be given the large more power resource of user assignment of business applications, has realized the fairness of power resource allocation.Meanwhile, the same maximization that realizes power system capacity that also can try one's best of this criterion.Therefore the power division criterion that, the present invention selects is got compromise between power system capacity maximum and the fairness of power resource allocation.In addition, owing to adopting link budget equation to determine that user assignment arrives to obtain capacity in the present invention, therefore, each user assignment to capacity in actual satellite communication system, can realize.

Summary of the invention

The object of the present invention is to provide the power distribution method in a kind of satellite communication system, solve the power division problem in satellite communication system, to improve the utilization ratio of system power resource.

The technical solution that realizes the object of the invention is: the power distribution method in a kind of satellite communication system, and allocation step is as follows:

Step 1: in the fairness power division model of satellite communication system, initialization dual variable λ and σ=[σ ₁, σ ₂..., σ _k] and the iteration step length of each dual variable; By each σ _ivalue be set to 0, wherein i ∈ 1,2 ..., K}, and its corresponding iteration step length

be set to 0.001; Choose the user U of business applications minimum _j, be P to the power of its distribution _j=P _totalt _j/ T _total, wherein T _totalfor the summation of all customer service applications, P _totalfor gross power on satellite communication system culminant star, T _jfor user U _jbusiness applications; Then determine the initial value λ of dual variable λ ₀, and its iteration step length is set to λ ₀/ 100000;

Step 2: the power resource of determining the each user assignment after optimizing

Step 3: according to the performance number of distributing to each user obtaining in previous step, upgrade dual variable λ and σ _ivalue;

Step 4: whether judgement below two conditions meets simultaneously:

with $\left|{\mathrm{\&sigma;}}_i^{n+1}(W_{B_i}-{\mathrm{\&Sigma;}}_{j\&Element;N_{B_i}}{W}_j)\right|<\mathrm{\&epsiv;},\&ForAll;i\&Element;\{1,...,K\},\mathrm{\&epsiv;}=0.001;$ If meet, finish whole assigning process simultaneously; Otherwise, return to step 2.

The fairness power division model of the satellite communication system in step 1 is:

$\underset{\left\{P_i\right\}}{\min }\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{(T_i-D_i)}^2$

s.t.

$D_i=\frac{P_i\&CenterDot;G_S\&CenterDot;{(G/T)}_i}{L_i\&CenterDot;{(E_b/N_0)}_i\&CenterDot;k}\&le;T_i$

$\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{P}_i\&le;P_{\mathrm{total}}$

$\underset{i\&Element;N_{B_j}}{\mathrm{\&Sigma;}}{W}_i\&le;W_{B_j}$

Step 1 and step 2 determine that according to following formula the initial value of dual variable λ and the each user assignment after optimizing arrive to obtain power resource

$\frac{2\&CenterDot;G_S\&CenterDot;{(G/T)}_i}{L_i\&CenterDot;{(E_b/n_0)}_i\&CenterDot;k}(T_i-\frac{P_i^{\mathrm{opt}}\&CenterDot;G_S\&CenterDot;{(G/T)}_i}{L_i\&CenterDot;{(E_b/n_0)}_i\&CenterDot;k})=\mathrm{\&lambda;}+{\mathrm{\&sigma;}}_j\frac{G_S\&CenterDot;{(G/T)}_i\&CenterDot;[1+\mathrm{\&rho;}\left({\mathrm{\&alpha;}}_i\right)]}{L_i\&CenterDot;{(E_b/n_0)}_i\&CenterDot;k\&CenterDot;\mathrm{\&eta;}\left({\mathrm{\&alpha;}}_i\right)},i\&Element;N_{B_j};$

Wherein, L _irepresent i user's downlink loss, it is mainly declined by free-space loss, rain and polarization loss etc. forms; (G/T) _irepresent the gain of i user receiving equipment and the ratio of equivalent noise temperature; G _srepresent the gain of satellite transmitting antenna; K represents Boltzmann constant, and its value is 1.379 × 10 ^-23w/KHz; α _irepresent the modulating-coding pattern that i user adopts; (E _b/ N ₀) _i, ρ (α _i) and η (α _i) represent respectively modulating-coding pattern α _icorresponding demodulation threshold bit signal to noise ratio, roll the factor and spectrum efficiency.

In step 3, the method for upgrading dual variable is

${\mathrm{\&lambda;}}^{n+1}={[{\mathrm{\&lambda;}}^n-{\mathrm{\&Delta;}}_{\mathrm{\&lambda;}}^n(P_{\mathrm{total}}-\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{P}_i^{\mathrm{opt}})]}^{+}$

${\mathrm{\&sigma;}}_i^{n+1}={[{\mathrm{\&sigma;}}_i^n-{\mathrm{\&Delta;}}_{\mathrm{\&sigma;}}^n(W_{B_i}-\underset{j\&Element;N_{B_i}}{\mathrm{\&Sigma;}}{W}_j)]}^{+}$

Wherein, [x] ⁺=max{0, x}; N represents iterations; Δ represents the iteration step length of each dual variable;

represent the power resource of the each user assignment after optimizing; P _totalthe power resource that expression system is total; represent the total bandwidth of i wave beam; W _jrepresent j the bandwidth resources that user is shared;

represent i user's set that wave beam is included.

Compared with prior art, its remarkable advantage: it will give the large more power resource of user assignment of business applications, has realized the fairness of power resource allocation in the present invention.Meanwhile, the same maximization that realizes power system capacity that also can try one's best of this algorithm.Therefore the present invention can obtain a compromise between power system capacity maximum and the fairness of power resource allocation.In addition, the calculation of capacity formula of employing via satellite link budget equation obtains, and therefore, can realize in actual satellite communication system for the capacity of each user assignment, has improved the practicality of allocation result.

Accompanying drawing explanation

Fig. 1 is the satellite communication system allocation plan of the power distribution method of satellite communication system of the present invention.

Fig. 2 is the flow chart of the power distribution algorithm that proposes of the present invention.

Fig. 3 is while adopting three kinds of different capacity distribution methods in the embodiment of power distribution method of satellite communication system of the present invention, the amount of capacity that each user assignment arrives.

Fig. 4 is while adopting different three kinds of power distribution algorithms in the embodiment of power distribution method of satellite communication system of the present invention, the difference of the capacity of each user assignment and its application traffic carrying capacity square.

Embodiment

Below in conjunction with accompanying drawing, the present invention is described in further detail.

In conjunction with Fig. 1～4, the power resource of establishing on this system culminant star can be distributed to the different user in different beams arbitrarily, but the total bandwidth of each wave beam is fixed.How the present invention mainly solves the problem for power on different user assignment stars in down link, to improve the overall performance of system.In this system, the system of setting up departments has K wave beam and M user.Wherein each wave beam B _irepresent, i ∈ 1,2 ..., K}; Each user U _irepresent, i ∈ 1,2 ..., M}, and at wave beam B _iin all users set use represent.On system culminant star, gross power is P _total.I user's business applications are T _i, for power resource on the star of its distribution is P _i.

In actual satellite communication system, the information that each user will send is by sending out after modulating-coding, the modulating-coding pattern adopting according to each user can be determined the relation between the capacity of its distribution and the power resource of needs, rather than determines both relations by the formula of Shannon capacity.If the modulating-coding pattern of i user's employing is α _i, and its corresponding demodulation threshold bit signal to noise ratio is (E _b/ N ₀) _i.Just can determine according to following formula is like this capacity D of i user assignment _i.

$D_i={\left(\frac{C}{N_0}\right)}_i/{\left(\frac{E_b}{N_0}\right)}_i---\left(1\right)$

(C/N in above formula ₀) _ibe i user's downlink carrier and the ratio of noise power spectral density, it can be obtained by following satellite link budget equation.

${\left(\frac{C}{N_0}\right)}_i=\frac{P_i\&CenterDot;G_S}{L_i\&CenterDot;k}\&CenterDot;{\left(\frac{G}{T}\right)}_i---\left(2\right)$

L in above formula _irepresent i user's downlink loss, it is mainly declined by free-space loss, rain and polarization loss etc. forms; (G/T) _irepresent the gain of i user receiving equipment and the ratio of equivalent noise temperature; G _srepresent the gain of satellite transmitting antenna; K represents Boltzmann constant, and its value is 1.379 × 10 ^-23w/KHz.

By in formula (2) substitution formula (1), just can obtain being the capacity of i user assignment and distributing the relation between power.

$D_i=\frac{P_i\&CenterDot;G_S\&CenterDot;{(G/T)}_i}{L_i\&CenterDot;{(E_b/N_0)}_i\&CenterDot;k}---\left(3\right)$

In the time that satellite communication system is fixing, the parameter in above formula (3) is except being the power P of each user assignment _ioutside, other parameters are all known quantities.The capacity that therefore can be found to be i user assignment by above-mentioned two formula depends on for power P on the star of its distribution _isize.In the time giving on the star of a user assignment the increase of output power, its capacity is also along with increase.But power is limited on star, therefore the total capacity of whole system is also limited.In addition be subject to equally, the restriction of bandwidth on each wave beam star for the capacity of each user assignment.As the capacity D of i user's application _iwith the modulating-coding pattern α adopting _iafter known, need to be for the bandwidth of its distribution:

$W_i=\frac{D_i\&CenterDot;[1+\mathrm{\&rho;}\left({\mathrm{\&alpha;}}_i\right)]}{\mathrm{\&eta;}\left({\mathrm{\&alpha;}}_i\right)}---\left(4\right)$

In above formula, η (α _i) and ρ (α _i) be respectively modulating-coding pattern α _ispectrum efficiency and roll the factor.If

be the bandwidth resources of i beam allocation, the total bandwidth of all user assignments is therein less than

The result that optimization aim is chosen power division has important impact, from in existing research common take maximized system capacity as optimization aim different, the present invention is take the quadratic sum of difference that minimizes all users' business applications and the capacity of distribution as optimization aim.This optimization aim can be the relatively large capacity of user assignment that business applications are large, to improve the fairness between power division.Meanwhile, the maximization that realizes power system capacity that this optimization aim also can be tried one's best.Therefore, the present invention selects optimization aim to get compromise between power system capacity maximum and the fairness of power resource allocation.According to above-mentioned optimization aim, set up the fairness power division model in following satellite communication system.

$\underset{\left\{P_i\right\}}{\min }\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{(T_i-D_i)}^2---\left(5\right)$

s.t.

$D_i=\frac{P_i\&CenterDot;G_S\&CenterDot;{(G/T)}_i}{L_i\&CenterDot;{(E_b/N_0)}_i\&CenterDot;k}\&le;T_i---\left(6\right)$

$\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{P}_i\&le;P_{\mathrm{total}}---\left(7\right)$

$\underset{i\&Element;N_{B_j}}{\mathrm{\&Sigma;}}{W}_i\&le;W_{B_j}---\left(8\right)$

The capacity that constraints (6) is expressed as each user assignment can not exceed the traffic carrying capacity of its application, to prevent the situation of the emergent power wasting of resources, the gross power that constraints (7) is expressed as all user assignments should not exceed the gross power on star, and the total bandwidth that constraints (8) is expressed as all user assignments in same wave beam should not exceed the total bandwidth of wave beam.

Target function (5) is a convex function, and constraints (6)-(8) are linear functions.Therefore this optimization is a protruding optimization problem.Because this problem is non-linear belt restraining problem, its optimal solution of direct solution is more difficult, therefore can solve by its dual problem.Protruding optimization has a good character, and the optimal value of its dual problem is also primal problem optimal value, after solving the optimal solution of dual problem, just can on the basis of the optimal solution of dual problem, obtain easily the optimal solution of primal problem.For this reason, first introduce dual variable λ and σ=[σ ₁, σ ₂..., σ _k], produce Lagrangian:

$L(P,\mathrm{\&sigma;},\mathrm{\&lambda;})=\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{(T_i-D_i)}^2-\mathrm{\&lambda;}(P_{\mathrm{total}}-\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{P}_i)-\underset{i=1}{\overset{K}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}_i(W_{B_i}-\underset{j\&Element;N_{B_i}}{\mathrm{\&Sigma;}}{W}_j)---\left(9\right)$

P=[P in above formula ₁, P ₂..., P _n].

Can obtain Lagrange duality function according to formula (9), that is:

$D(\mathrm{\&sigma;},\mathrm{\&lambda;})=\underset{P}{\min }L(P,\mathrm{\&sigma;},\mathrm{\&lambda;})---\left(10\right)$

And the dual problem of former problem:

$d=\underset{\mathrm{\&lambda;}\&GreaterEqual;0,\mathrm{\&sigma;}\&GreaterEqual;0}{\max }D(\mathrm{\&sigma;},\mathrm{\&lambda;})---\left(11\right)$

Its dual problem can be divided into again following two subproblems and solve.

Subproblem 1-power division: given dual variable λ and σ=[σ ₁, σ ₂..., σ _k], for i ∈ arbitrarily 1,2 ..., M}, to formula (9) about P _icarry out differentiate, obtain following equation:

$\frac{2\&CenterDot;G_S\&CenterDot;{(G/T)}_i}{L_i\&CenterDot;{(E_b/n_0)}_i\&CenterDot;k}(T_i-\frac{P_i^{\mathrm{opt}}\&CenterDot;G_S\&CenterDot;{(G/T)}_i}{L_i\&CenterDot;{(E_b/n_0)}_i\&CenterDot;k})=\mathrm{\&lambda;}+{\mathrm{\&sigma;}}_j\frac{G_S\&CenterDot;{(G/T)}_i\&CenterDot;[1+\mathrm{\&rho;}\left({\mathrm{\&alpha;}}_i\right)]}{L_i\&CenterDot;{(E_b/n_0)}_i\&CenterDot;k\&CenterDot;\mathrm{\&eta;}\left({\mathrm{\&alpha;}}_i\right)},i\&Element;N_{B_j}---\left(12\right)$

Subproblem 2-dual variable is upgraded: optimum dual variable can be tried to achieve by addressing the problem:

$({\mathrm{\&sigma;}}^{\mathrm{opt}},{\mathrm{\&lambda;}}^{\mathrm{opt}})=\arg \underset{\mathrm{\&sigma;},\mathrm{\&lambda;}}{\max }\min \left[L(P^{\mathrm{opt}},\mathrm{\&sigma;},\mathrm{\&lambda;})\right]---\left(13\right)$

Because dual function is convex function, therefore can adopt following sub-gradient method to upgrade:

${\mathrm{\&lambda;}}^{n+1}={[{\mathrm{\&lambda;}}^n-{\mathrm{\&Delta;}}_{\mathrm{\&lambda;}}^n(P_{\mathrm{total}}-\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{P}_i^{\mathrm{opt}})]}^{+}---\left(14\right)$

${\mathrm{\&sigma;}}_i^{n+1}={[{\mathrm{\&sigma;}}_i^n-{\mathrm{\&Delta;}}_{\mathrm{\&sigma;}}^n(W_{B_i}-\underset{j\&Element;N_{B_i}}{\mathrm{\&Sigma;}}{W}_j)]}^{+}---\left(15\right)$

In above formula [x] ⁺=max{0, x}, n is iterations, Δ is the iteration step length of each dual variable.

The allocation step of whole power resource is as follows:

Step 1: first initialization dual variable λ and σ=[σ ₁, σ ₂..., σ _k] and the step-length of each dual variable.By each σ _ivalue be set to 0, wherein i ∈ 1,2 ..., K}, and its iteration step length

be set to 0.001.Choose the user U of business applications minimum _j, be P to the power of its distribution _j=P _totalt _j/ T _total, wherein T _totalfor the summation of all customer service applications.Then determine the initial value λ of dual variable λ according to formula (12) ₀, and its iteration step length is set to λ ₀/ 100000.

Step 2: according to formula (12), obtain the power resource of the each user assignment after optimizing.

Step 3: the performance number of distributing to each user obtaining by step (2) is brought into formula (14) and formula (15), upgrades the value of dual variable.

Step 4: whether judgement below two conditions meets simultaneously:

with $\left|{\mathrm{\&sigma;}}_i^{n+1}(W_{B_i}-{\mathrm{\&Sigma;}}_{j\&Element;N_{B_i}}{W}_j)\right|<\mathrm{\&epsiv;},\&ForAll;i\&Element;\{1,...,K\},\mathrm{\&epsiv;}=0.001.$ If met simultaneously, finish whole assigning process, otherwise skip to step 2, proceed iteration and upgrade.

By above 4 steps, can finally obtain the power resource for each user assignment after optimizing.The flow chart of whole power distribution algorithm as shown in Figure 2.

Below effect of the present invention to be described as next embodiment.The downlink loss L of each user in embodiment Satellite communication system _i, receiving equipment gain with equivalent noise temperature than (G/T) _iand the modulating-coding pattern α selecting _iall the same, in system, the occurrence of each parameter is as shown in the table:

The parameter of table 1 satellite communication system

Parameter	Parameter value
		Wave beam number	4
Number of users	20
		Number of users in each wave beam	5
Each user's business applications	From 1Mbps to 20Mbps, stepping 1Mbps
		Satellite gross power [P total]	20W
Satellite transmitting antenna gain [G S]	20000
		The bandwidth of each wave beam	100MHz
The gain of user receiving equipment with equivalent noise temperature than [G/T]	20
		User's downlink loss [L i]	2e 15
Spectrum efficiency [η (the α of modulating-coding pattern i)]	1.5
		Rolling coefficient [ρ (α of modulating-coding pattern i)]	1
Thresholding demodulation bit the signal to noise ratio [(E that modulating-coding pattern is corresponding b/N 0) i]	2.63

Process in accordance with the present invention 1 can be determined initial dual variable λ ₀=9.61e ⁵, the iteration step length of selection then carry out iteration according to allocation step, in the time iterating to the 36th time, meet algorithm iteration termination condition, now finishing iteration process, obtains as each user power resource value.

The validity of power distribution method proposing in order to embody the present invention, contrasts it and following two kinds of conventional power allocation algorithms:

1, even power distribution method: power resource is averagely allocated to each user.

2, ratio power distribution method: the ratio that power resource is taken to all business applications according to each customer service applications is distributed.

The total capacity of system when table 2 adopts three kinds of different capacity allocation algorithms

Distribution method	∑C i
		Evenly power division	109.1Mbps
Ratio power division	109.1Mbps
		The power division that the present invention proposes	109.1Mbps

The quadratic sum of the difference of the capacity of all user assignments and its application traffic carrying capacity when table 3 adopts three kinds of different capacity allocation algorithms

Distribution method	∑(T i-C i) 2
		Evenly power division	1.134E15
Ratio power division	6.627E14
		The power division that the present invention proposes	5.470E14

Fig. 2 has shown while adopting three kinds of different capacity distribution methods, the amount of capacity that each user assignment arrives.Table 2 has shown the overall system capacity while adopting three kinds of different capacity allocation algorithms.Fig. 3 shown while adopting different three kinds of power distribution algorithms, the difference of the capacity of each user assignment and its application traffic carrying capacity square.Table 3 has shown while adopting three kinds of different capabilities allocation algorithms, the quadratic sum of the difference of the capacity of all user assignments and its application traffic carrying capacity.Can find by Fig. 2, the power distribution algorithm that the present invention proposes provides more capacity will to the large user of business applications.And owing to being linear relationship between calculation of capacity function and the power of distribution, the overall system capacity that therefore three kinds of power distribution algorithms obtain is the same, as shown in table 2.Because the power distribution algorithm that the present invention proposes can be given the large more capacity of user assignment of business applications, therefore for the large user of business applications, the difference of its business applications and partition capacity square can be less than other two kinds of distribution methods.In addition, the difference of customer service applications used and partition capacity square is also minimum, as shown in table 3.In other words, the power algorithm that the present invention proposes is better than other two power distribution algorithms aspect fairness.

Claims (4)

1. the power distribution method in satellite communication system, is characterized in that, allocation step is as follows:

Step 1: in the fairness power division model of satellite communication system, initialization dual variable λ and σ=[σ ₁, σ ₂..., σ _k] and the iteration step length of each dual variable; By each σ _ivalue be set to 0, wherein i ∈ 1,2 ..., K}, and its corresponding iteration step length

be set to 0.001; Choose the user U of business applications minimum _j, be P to the power of its distribution _j=P _totalt _j/ T _total, wherein T _totalfor the summation of all customer service applications, P _totalfor gross power on satellite communication system culminant star, T _jfor user U _jbusiness applications; Then determine the initial value λ of dual variable λ ₀, and its iteration step length is set to λ ₀/ 100000;

Step 2: the power resource of determining the each user assignment after optimizing

Step 3: according to the performance number of distributing to each user obtaining in previous step, upgrade dual variable λ and σ _ivalue;

Step 4: whether judgement below two conditions meets simultaneously:

with if meet, finish whole assigning process simultaneously; Otherwise, return to step 2.

2. the power distribution method in satellite communication system according to claim 1, is characterized in that: the fairness power division model of the satellite communication system in step 1 is:

s.t.

。

3. the power distribution method in satellite communication system according to claim 1, is characterized in that: step 1 and step 2 determine that according to following formula the initial value of dual variable λ and the each user assignment after optimizing arrive to obtain power resource

Wherein, L _irepresent i user's downlink loss, it is mainly declined by free-space loss, rain and polarization loss etc. forms; (G/T) _irepresent the gain of i user receiving equipment and the ratio of equivalent noise temperature; G _srepresent the gain of satellite transmitting antenna; K represents Boltzmann constant, and its value is 1.379 × 10 ^-23w/KHz; α _irepresent the modulating-coding pattern that i user adopts; (E _b/ N ₀) _i, ρ (α _i) and η (α _i) represent respectively modulating-coding pattern α _icorresponding demodulation threshold bit signal to noise ratio, roll the factor and spectrum efficiency.

4. the power distribution method in satellite communication system according to claim 1, is characterized in that: in step 3, the method for upgrading dual variable is

Wherein, [x] ⁺=max{0, x}; N represents iterations; Δ represents the iteration step length of each dual variable;

represent the power resource of the each user assignment after optimizing; P _totalthe power resource that expression system is total;

represent the total bandwidth of i wave beam; W _jrepresent j the bandwidth resources that user is shared;

represent i user's set that wave beam is included.

Images