CN1852043A - Power control method of multi-carrier-wave wireless honeycomb communication system - Google Patents

Abstract

This invention relates to a power control method for multi-carrier wireless cell phone communication system, in which, first of all, a base station end tests the sum of interference and noises on a first sub-carrier of the ith user, evaluates the channel link gain, determines the parameter values to compute the ration of a parameter b1 and the optimum signal/interference-noise ratio, if it is smaller than the target S/I-N ratio, then the user stops the transmission and orders the emission power to be zero, if it is greater than that, it computes the ratio of the two at the base station end and feeds it back to ith user, who computes out the Pi (n+1) and repeats the step till i=K, then orders 1=1+1, then repeats the above mentioned step till 1=L then continues the steps to get the optimum emission power.

Description

The Poewr control method of multi-carrier-wave wireless honeycomb communication system

Technical field

The invention belongs to field of wireless transmission, relate to a kind of multiple services distributed power control method of adaptation that adopts the design of non-cooperative game theory analysis at multicarrier (Multi-carrier) radio honeycomb communication system.

Background technology

Power control is very important for radio honeycomb communication system, and particularly in the system that adopts code division multiple access (CDMA) technology, power control more is absolutely necessary.Because utilize power control can resist near-far interference, improve channel capacity, improve service quality, reduce the biological effect of microwave radiation, prolong the travelling carriage battery life.The function of power control is to be user's assignment transmitting power, makes their service quality (QoS) require to be met, and suppresses the interference that is produced simultaneously.Mainly still SINR balanced algorithm or received power balanced algorithm are main to existing power control algorithm, but these two kinds of algorithms can not be according to the adaptive variation of the variation of Channel Transmission environment, also there is the positive feedback phenomenon in the SINR balanced algorithm, and these algorithms are mainly at speech business.Therefore, along with the application of a large amount of multimedia services, original power control scheme is can not satisfy this multiple services the requirement fully.Simultaneously, the use of many antennas, multi-transceiver technology, many new problems that made original systems face.Therefore, we are necessary at supporting multi-service and based on the radio honeycomb communication system of multi-antenna multi-carrier-wave technology, designing an effective power control scheme.

Game theoretic introducing solves this class problem to us and has brought a new method.Game theory is to analyze interactional mathematical tool between the participant with conflict of interest.The main purpose of traditional power control is to launch the power that just satisfies essential performance objective.And be that one of emission is to own best power under all multifactor restrictions and influence based on the main purpose of game theoretic power control, this power also can guarantee to satisfy essential performance objective certainly., self adaptation distributed because of having based on the control of game theoretic power, adjustment be flexible, satisfy characteristics such as multi-service, can overcome the systematic jitters of the positive feedback introducing of SINR, more and more receive publicity, and be highly suitable for a new generation the wireless communication system of emphasizing adaptive design in use.

Calendar year 2001, people such as Mingbo Xiao are on forefathers' basis, proposed to be the distributed non-cooperation power control algorithm of utility function, and be referred to as power control scheme (UBPC) based on utility function at (QoS) equivalence that QoS of customer is required of wireless cellular system.This algorithm has had certain progress again with non-cooperation power control game playing algorithm than other people, says that in a sense in fact all have been united based on game theoretic power control algorithm forefathers for it, has formed a general power controlling Design scheme.This algorithm can reduce the target Signal to Interference plus Noise Ratio (SINR) of certain customers automatically when system congestion, the convergence of promotion system and calling control, and have some adjustable parameters so that the user can obtain required fairness, adaptivity etc., can also satisfy the various blended service requirements (for example time delay and bit error rate etc.) of each voice-and-data user in the heterogeneous network simultaneously.In addition, compare with power control algorithm in the past, the algorithm of Mingbo Xiao can come adaptive adjustment target Signal to Interference plus Noise Ratio according to each transmission environment quality constantly, rather than specifies a fixing target Signal to Interference plus Noise Ratio by system, and this is a big advantage of this algorithm.This advantage has guaranteed that also this algorithm goes for multiple radio honeycomb communication system, and stronger applicability is also promptly arranged.In a word, that the power control algorithm that people such as Mingbo Xiao propose has is distributed, self adaptation, adjustment flexibly, satisfy advantages such as multi-service, applicability be strong, than forefathers' scheme many advantages are arranged.

The power control scheme UBPC of people such as Mingbo Xiao design is formulated as following form to power control problem:

$p_i^{*}=\max [{\hat{p}}_i,0]---\left(1\right)$

${\hat{p}}_i=\underset{p_i\≥0}{\arg \max }{\mathrm{NU}}_i---\left(2\right)$

NU _i＝U _i(γ _i)-C _i(p _i) (3)

Wherein, γ _iBe the Signal to Interference plus Noise Ratio of user i, p _iBe the transmitting power of user i, p _i ^*Be the optimal transmit power of user i, i=1,2 ..., K is user's ordinal number; NU _iFor the network utility function (net utilityfunction) of user i, reasonable in order to make betting model, the network utility function should be a quasiconcave function; U _i(γ _i) be user i from utility function (self utility function), it is one is the function of independent variable with the Signal to Interference plus Noise Ratio, and it will satisfy following character (character 1): U _i(0)=0, U _i(∞)=1, U _i(γ _i) be worth with γ _iThe increase of value and increasing; C _i(p _i) be the cost function (pricing function) of user i, it has reflected the cost that the user pays for the transmission power level of oneself, and can restrict the user unrestrictedly increases transmitting power, and it will satisfy following character (character 2): C _i(0)=0, C _i(p _i) be worth with p _iThe increase of value and increasing.People such as Mingbo Xiao adopted following form from utility function:

$U_i\left({\mathrm{\γ}}_i\right)=\frac{1}{1+e^{-{\mathrm{\α}}_i({\mathrm{\γ}}_i-{\mathrm{\β}}_i)}}---\left(4\right)$

Wherein, parameter alpha _iAnd β _iBe real number, reflected the steepness and the center symmetric points of function respectively, their value can be adjusted according to business need.If SINR _MinFor system guarantees the desired minimum SINR value of transmission quality for each user.Parameter S INR then _Min, α _iAnd β _iAs long as know wherein two, definite the 3rd parameter that just can be unique.Usually, SINR _MinBe known, and α _iThe reflection steepness can be determined according to business need; Like this, at definite SINR _MinAnd α _iSituation under, β _iValue just can be obtained by following equation:

$U_i^{\′}\left({\mathrm{SINR}}_{\min }^i\right){\mathrm{SINR}}_{\min }^i=U_i\left({\mathrm{SINR}}_{\min }^i\right)---\left(5\right)$

If represent iteration renewal ordinal number with k, During the k time renewal of expression user i by $\frac{{\mathrm{dNU}}_i({\mathrm{\γ}}_i,p_i)}{{\mathrm{dp}}_i}=0$ The Signal to Interference plus Noise Ratio value that solves, then the optimal power allocation update scheme of UBPC can be written as:

${\hat{p}}_i(k+1)=\mathrm{\Ψ}\left(p_i\left(k\right)\right)=\frac{{\widehat{\mathrm{\γ}}}_i\left(k\right)}{{\mathrm{\γ}}_i\left(k\right)}{p}_i\left(k\right)---\left(6\right)$

People such as Mingbo Xiao have proved the power mapping

Be a standard interference function, the iterative algorithm of UBPC also is a calibration power control algolithm (standard power control).

The target of distributed power control (DCPC) algorithm of called optical imaging is to satisfy under the restrictive condition of minimum Signal to Interference plus Noise Ratio minimized transmitting power.Its power control algorithm can be written as:

${\hat{p}}_i(k+1)=\mathrm{\Ψ}\left(p_i\left(k\right)\right)=\frac{{\mathrm{\γ}}_i^l}{{\mathrm{\γ}}_i\left(k\right)}{p}_i\left(k\right)---\left(7\right)$

As indicated above, (7) formula is compared with (6) formula, and have only the target signal interference ratio difference of molecular moiety: the target signal interference ratio in (7) formula is given by system, and the target signal interference ratio in (6) formula calculates in real time according to transmission environment and business need.Obviously, UBPC has more flexibility.

Yet there is following problem in people's such as Mingbo Xiao research:

UBPC only is given in the application in the single antenna single carrier cdma system down link, and do not relate to up-to-date mobile cellular system now, such as the 3rd generation (3G) of adopting technology such as multicarrier, many antennas and afterwards 3 generations or the 4th generation (B3G or 4G) mobile cellular system.

The design of utility function is based on the core of the power scheme of theory of games, and people such as Mingbo Xiao have adopted widely used sigmoid function in neural net (Sigmoid function) to be used as the basis of UBPC utility function.This function is based on the e exponential function, and independent variable is on the index, and this has increased the computational complexity of algorithm.In addition, $U_i\left(0\right)=1/(1+e^{{\mathrm{\α}}_i{\mathrm{\β}}_i})\≠0,$ Only work as α _iβ _iWhen very big, U _i(0) just approximates 0, so it can not the strict requirement of satisfying character 1.

Summary of the invention

The objective of the invention is to solve the problems of the technologies described above, designed a kind of distributed power controlling schemes at multi-carrier-wave wireless honeycomb communication, this power control scheme can adapt to the multi-service requirement based on theory of games.

The present invention adopts following technical scheme to solve technical problem:

A kind of Poewr control method of multi-carrier-wave wireless honeycomb communication system comprises following steps:

Step 1 makes n=1, definition n _Max, each user's initialization power vector p ₁ ⁰..., p _K ⁰,

Step 2 makes i=1, l=1,

Step 3 definition $\underset{p_i}{\max }{\mathrm{NU}}_i^{\mathrm{MC}}=\underset{p_i^l,...,P_i^L}{\max }[\underset{l}{\overset{L}{\mathrm{\Σ}}}{U}_i^l\left({\mathrm{\γ}}_i^l\right)-\underset{l}{\overset{L}{\mathrm{\Σ}}}{C}_i^l\left(p_i^l\right)],$ Wherein $U\left({\mathrm{\γ}}_i^l\right)=1-\frac{1}{1+{({\mathrm{\γ}}_i^l/b_i)}^{a_i}}$ For from utility function, at first the base station end is measured the I on l subcarrier of i user _i ^l(n), estimate channel link gain h _i ^l(n), determine good each parameter value, utilize then $b_i={\mathrm{SINR}}_{\min }^i/{(a_i-1)}^{\frac{1}{a_i}}$ Calculate parameter b _i, utilize $\frac{{\∂U}_i^l}{{\∂p}_i^l}=\frac{a_i{({\mathrm{\γ}}_i^l/b_i)}^{a_i}}{p_i^l{(1+{({\mathrm{\γ}}_i^l/b_i)}^{a_i})}^2}-\frac{{\∂C}_i^l\left(p_i^l\right)}{\∂p_i^l}=0$ Calculate best Signal to Interference plus Noise Ratio value

If Less than target Signal to Interference plus Noise Ratio value, then this user stops transmission, even transmitting power is 0, and jumps to step 5; If greater than, then calculate at the base station end

And feeding back to user i, user i utilizes ${\hat{p}}_i(n+1)=p_i\left(n\right){\widehat{\mathrm{\γ}}}_i\left(k\right)/{\mathrm{\γ}}_i\left(k\right)$ Calculate p _i ^l(n+1),, repeat this step, till i=K, so just can obtain optimum transmitting power with seasonal i=i+1: ${\hat{p}}_i=\underset{p_i\≥0}{\arg \max }{\mathrm{NU}}_i^{\mathrm{MC}},$

Step 4 is if l=L then jumps to step 5; If not, make l=l+1, and get back to step 3,

Step 5 makes n=n+1, if n is greater than n _Max, then jump to step 6, if less than but p ₁ ^N+1-p ₁ ⁿ., p _K ^N+1-p _K ⁿSatisfy required precision, then enter step 6,, then get back to step 2 if do not satisfy required precision,

Step 6 stops iteration, and each user transmitting power vector at this moment is Nash Equilibrium strategy p ₁ ^*..., p _K ^*

Compared with prior art, the present invention has following advantage:

The present invention adopts the non-cooperative game model to describe the power control problem of multi-carrier-wave wireless honeycomb communication system, and the amplitude function of utilizing Butterworth filter has been constructed new for utility function, avoided the e exponential function among the UBPC, the strict requirement of having satisfied character 1 has also kept the advantage of UBPC simultaneously.The present invention is by game mechanism reasonable in design, realized effective control to transmitting power, and can break through the limitation of the SIR balance criterion constant power control algolithm of present practical application by adjusting the convenient, flexible various business need of adaptation of parameter on utility function.

Description of drawings

Fig. 1 is the power control algorithm flow chart that the present invention designs.

Fig. 2 be the present invention design from the curve chart of utility function about SINR.

Fig. 3 be the present invention design from utility function under identical cost function with UBPC from the convergence of utility function relatively (many antennas situation).

Fig. 4 is algorithm of the present invention convergence situation (many antennas situation) in instantiation.

Fig. 5 is that the SINR value after algorithm of the present invention and SINR balanced algorithm are restrained in example compares (many antennas situation).

Fig. 6 be the present invention design from utility function under identical cost function with UBPC from the convergence of utility function relatively (single-antenna case).

Fig. 7 is that the SINR value after algorithm of the present invention and SINR balanced algorithm are restrained in example compares (single-antenna case).

Embodiment

A kind of Poewr control method of multi-carrier-wave wireless honeycomb communication system (with reference to Fig. 1) comprises following steps:

Step 1 makes n=1, definition n _Max, each user's initialization power vector p ₁ ⁰..., p _K ⁰,

Step 2 makes i=1, l=1,

Step 3 definition $\underset{p_i}{\max }{\mathrm{NU}}_i^{\mathrm{MC}}=\underset{p_i^l,...,p_i^L}{\max }[\underset{l}{\overset{L}{\mathrm{\Σ}}}{U}_i^l\left({\mathrm{\γ}}_i^l\right)-\underset{l}{\overset{L}{\mathrm{\Σ}}}{C}_i^l\left(p_i^l\right)],$ Wherein $U\left({\mathrm{\γ}}_i^l\right)=1-\frac{1}{1+{({\mathrm{\γ}}_i^l/b_i)}^{a_i}}$ For from utility function, at first the base station end is measured the I on l subcarrier of i user _i ^l(n), estimate channel link gain h _i ^l(n), determine good each parameter value, utilize then $b_i={\mathrm{SINR}}_{\min }^i/{(a_i-1)}^{\frac{1}{a_i}}$ Calculate parameter b _i, utilize $\frac{{\∂U}_i^l}{\∂p_i^l}=\frac{a_i{({\mathrm{\γ}}_i^l/b_i)}^{a_i}}{p_i^l{(1+{({\mathrm{\γ}}_i^l/b_i)}^{a_i})}^2}-\frac{{\∂C}_i^l\left(p_i^l\right)}{\∂p_i^l}=0$ Calculate best Signal to Interference plus Noise Ratio value

If Less than target Signal to Interference plus Noise Ratio value, then this user stops transmission, even transmitting power is 0, and jumps to step 5; If greater than, then calculate at the base station end And feeding back to user i, user i utilizes ${\hat{p}}_i(n+1)=p_i\left(n\right){\widehat{\mathrm{\γ}}}_i\left(k\right)/{\mathrm{\γ}}_i\left(k\right)$ Calculate p _i ^l(n+1),, repeat this step, till i=K, so just can obtain optimum transmitting power with seasonal i=i+1: ${\hat{p}}_i=\underset{p_i\≥0}{\arg \max }{\mathrm{NU}}_i^{\mathrm{MC}},$

Step 4 is if l=L then jumps to step 5; If not, make l=l+1, and get back to step 3,

Step 5 makes n=n+1, if n is greater than n _Max, then jump to step 6, if less than but p ₁ ^N+1-p ₁ ⁿ..., p _K ^N+1-p _K ⁿSatisfy required precision, then enter step 6,, then get back to step 2 if do not satisfy required precision,

Step 6 stops iteration, and each user transmitting power vector at this moment is Nash Equilibrium strategy p ₁ ^*..., p _K ^*

The above-mentioned of the Poewr control method of above-mentioned multi-carrier-wave wireless honeycomb communication system from utility function is $U\left(\mathrm{\γ}\right)=1-\frac{1}{1+{(\mathrm{\γ}/b)}^a},$ Strict satisfied $U\left(0\right)=1-\frac{1}{1+0}=0,$ U (∞)=∞, U (γ) is the requirement about the increasing function of γ.

Above-mentioned from utility function $U\left(\mathrm{\γ}\right)=1-\frac{1}{1+{(\mathrm{\γ}/b)}^a},$ The value of a is relevant with type of service, for the demanding professional a of real-times such as speech business 〉=10, real-time is required business 4≤a≤8 suitable with the data stream class service, and real-time is required and professional suitable business 1≤a≤3 of interactive type; And the value of b can be utilized equation by given minimum Signal to Interference plus Noise Ratio value and a $b_i={\mathrm{SINR}}_{\min }^i/{(a_i-1)}^{\frac{1}{a_i}}$ Calculate.

The present invention has designed new from utility function, to avoid occurring the form of e exponential function.What the present invention designed from utility function is:

$U\left(\mathrm{\γ}\right)=1-\frac{1}{1+{(\mathrm{\γ}/b)}^a},a>0,b>0---\left(8\right)$

Wherein, a and b are undetermined constant, and a has reflected the steepness from utility function, is similar to the parameter alpha among the UBPC, and a is big more, and curve is precipitous more.So to real time business, as speech business, a will get greatly, this user just is not easy to be closed more like this, and a is big more certainly, and operand is also bigger comparatively speaking, because the order of function has increased; And to the data business, because so not high to delay requirement, it is littler that a can get relatively, but this user also relatively easily is closed.B has reflected the center position from utility function, is similar to the parameter beta among the UBPC, and we can pass through (5) formula, at parameter a and SINR _MinUnder the situation about determining, calculate a unique b value.As seen by adjusting the value of parameter a and b, we just can adapt to various business demand easily.This abstract Mathematical Modeling is owing to avoided the contact of some technical schemes (as modulation, coded system etc.) in direct and the various real systems, often has stronger applicability, because the expression formula of these technology correspondences is very complicated sometimes, even possibly can't explicit expression.Simultaneously, the present invention design both avoided adopting the form of e exponential function from utility function, again can the strict requirement of satisfying character 1.

From utility function, we have designed following non-cooperation power control betting model at the multi-antenna multi-carrier-wave wireless cellular system based on this.

Consider K user's multiple antenna and carrier system, the number of transmit antennas of establishing each user is t, and the reception antenna number is r, total L subcarrier, and complete quadrature between each subcarrier, then the Signal to Interference plus Noise Ratio of l subcarrier is on i user j root antenna:

${\mathrm{\γ}}_{\mathrm{ij}}^l=\frac{h_{\mathrm{ij}}^l{p}_{\mathrm{ij}}^l}{{\mathrm{\Σ}}_{k=1,k\≠i}^K{\mathrm{\Σ}}_{m=1}^t{h}_{\mathrm{km}}^l{p}_{\mathrm{km}}^l+{\mathrm{\Σ}}_{v=1,v\≠j}^t{h}_{\mathrm{iv}}^l{p}_{\mathrm{iv}}^l+{\mathrm{\σ}}^2}---\left(9\right)$

Wherein, h _Ij ^lBe pairing link gain on l subcarrier j of i the user root antenna, p _Ij ^lBe the transmitting power on l subcarrier j of i the user root antenna, σ ²Be pairing received noise power.We use I _Ij ^lRepresent the interference and the noise sum that are subjected on l subcarrier j of i the user root antenna, it comprises inter-user interference and inter-antenna interference sum, also promptly:

$I_{\mathrm{ij}}^l={\mathrm{\Σ}}_{k=1,k\≠i}^K{\mathrm{\Σ}}_{m=1}^t{h}_{\mathrm{km}}^l{p}_{\mathrm{km}}^l+{\mathrm{\Σ}}_{v=1,v\≠j}^t{h}_{\mathrm{iv}}^l{p}_{\mathrm{iv}}^l+{\mathrm{\σ}}^2---\left(10\right)$

We design following non-cooperative game model, and each user maximizes separately effectiveness (utility) by adjusting transmitting power on L the subcarrier.This betting model can be represented with following mathematical form: $G_{\mathrm{MC}}=\{\mathrm{\&kappa;},\{A_i^{\mathrm{MC}}\},\{{\mathrm{NU}}_i^{\mathrm{MC}}\left\}\right\},$ Wherein, К=1 ..., K is that the game participant also is each user's a ordinal number, A _i ^MCThe strategy of expression user i can be written as: $p_i=[p_{i1}^1,...,{\mathrm{<figure-callout\; id="1"\; label="p\; it"\; filenames="A20061004038500181.png,A20061004038500191.png"\; state="\{\{state\}\}">p</figure-callout>}}_{\mathrm{it}}^{},p_{i1}^2,...,{\mathrm{<figure-callout\; id="2"\; label="p\; it"\; filenames="A20061004038500212.png"\; state="\{\{state\}\}">p</figure-callout>}}_{\mathrm{it}}^{},...,p_{\mathrm{it}}^L].$ $A_i^{\mathrm{MC}}={[0,P_{\max }]}^L$ Be the value set of user i strategy, P _MaxBe maximum transmit power limit.The network utility function of definition user i is:

${\mathrm{NU}}_i^{\mathrm{MC}}=\underset{l}{\overset{L}{\mathrm{\&Sigma;}}}{U}_i^l\left({\mathrm{\&gamma;}}_i^l\right)-\underset{l}{\overset{L}{\mathrm{\&Sigma;}}}{C}_i^l\left(p_i^l\right)---\left(11\right)$

Wherein,

$U_i^l\left({\mathrm{\&gamma;}}_i^l\right)=(1-\frac{1}{1+{({\mathrm{\&gamma;}}_i^l/b_i)}^{a_i}})---\left(12\right)$

This game can be expressed as following maximization problems:

$\underset{p_i}{\max }{\mathrm{NU}}_i^{\mathrm{MC}}=\underset{p_i^l,...,p_i^L}{\max }{\mathrm{NU}}_i^{\mathrm{MC}},$ , to all i=1 ..., K (13)

And optimal transmit power more new formula can be expressed as:

${\hat{p}}_i=\underset{p_i\&GreaterEqual;0}{\arg \max }{\mathrm{NU}}_i^{\mathrm{MC}}---\left(14\right)$

The betting model of our definition, its Nash Equilibrium strategy are a vector power set p ₁ ^*..., p _K ^*, wherein ${\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="A20061004038500181.png,A20061004038500191.png"\; state="\{\{state\}\}">p</figure-callout>}}_1^{*}=(p_1^{l*},...p_1^{l*}),p_K^{*}=(p_K^{l*},...p_K^{l*})$ When each user adopted the Nash Equilibrium strategy, the neither one user can obtain higher benefit by the transmitting power strategy of independent change oneself.So the Nash Equilibrium strategy of this game is exactly an optimal transmit power value of the presently claimed invention.

Because each subcarrier quadrature, so can ignore interference between subcarrier.If each subcarrier is non-orthogonal, as long as in the denominator of Signal to Interference plus Noise Ratio, introduce the distracter between subcarrier.For user i, the Signal to Interference plus Noise Ratio on the subcarrier l is:

${\mathrm{\&gamma;}}_i^l=\frac{h_i^l{p}_i^l}{{\mathrm{\&Sigma;}}_{j=1,j\&NotEqual;i}^K{h}_j^l{p}_j^l+{\mathrm{\&sigma;}}^2}---\left(15\right)$

The network utility function is:

${\mathrm{NU}}_{\mathrm{il}}^{\mathrm{MC}}=U_i^l\left({\mathrm{\&gamma;}}_i^l\right)-C_i^l\left(p_i^l\right)---\left(16\right)$

${\mathrm{NU}}_i^{\mathrm{MC}}=\underset{l}{\overset{L}{\mathrm{\&Sigma;}}}{\mathrm{NU}}_{\mathrm{il}}^{\mathrm{MC}}---\left(17\right)$

Definition according to Nash Equilibrium can get:

${\mathrm{NU}}_{\mathrm{il}}^{\mathrm{MC}}(p_i^{l*},p_{-i}^{l*})\&GreaterEqual;{\mathrm{NU}}_{\mathrm{il}}^{\mathrm{MC}}(p_i^l,p_{-i}^{l*}),$ To all p _i ^l(18)

Wherein, p _-i ^L*Transmitting power for other user except user i.

By (16), (17) and (18) Shi Kede:

${\mathrm{NU}}_i^{\mathrm{MC}}(p_i^{*},p_{-i}^{*})\&GreaterEqual;{\mathrm{NU}}_i^{\mathrm{MC}}(p_i,p_{-i}^{*}),$ To all p _i(19)

Also promptly, maximize the effectiveness on each subcarrier of user, also just maximized the total effectiveness of user.

Below we provide a kind of iterative algorithm that obtains the Nash Equilibrium strategy.(12) formula differentiate is got:

$\frac{\&PartialD;U_i^l}{{\&PartialD;p}_i^l}=\frac{a_i{({\mathrm{\&gamma;}}_i^l/b_i)}^{a_i}}{p_i^l{(1+{({\mathrm{\&gamma;}}_i^l/b_i)}^{a_i})}^2}-\frac{{\&PartialD;C}_i^l\left(p_i^l\right)}{{\&PartialD;p}_i^l}---\left(20\right)$

Order $\frac{{\&PartialD;U}_i^l}{\&PartialD;p_i^l}=0,$ Then can solve by (20) formula

And determining parameter a and SINR _MinSituation under, according to (5) formula and (12) Shi Kede:

$b_i={\mathrm{SINR}}_{\min }^i/{(a_i-1)}^{\frac{1}{a_i}}---\left(21\right)$

Algorithm 1

Step 1 makes n=1, definition n _Max, each user's initialization power vector p ₁ ⁰..., p _K ⁰

Step 2 makes i=1, l=1.

Step 3 base station end is measured interference and the noise sum I on l subcarrier of i user _i ^l(n), estimate channel link gain h _i ^l(n), determine good each parameter value, (21) formula of utilization calculates b _i, (20) formula of utilization calculates If Less than target Signal to Interference plus Noise Ratio value, then this user stops transmission, even transmitting power is 0, and jumps to step 5; Otherwise end calculates in the base station And feeding back to user i, user i utilizes (6) formula to calculate p _i ^l(n+1), with seasonal i=i+1, repeat this step, till i=K.

Step 4 is if l=L then jumps to step 5; Otherwise, make l=l+1, and get back to step 3.

Step 5 makes n=n+1, judges that whether n is greater than n earlier _MaxIf, greater than, then jump to step 6, otherwise, judge p ₁ ^N+1-p ₁ ⁿ..., p _K ^N+1-p _K ⁿWhether satisfy required precision, if all satisfy, then enter step 6, otherwise get back to step 2.

Step 6 stops iteration, and each user transmitting power strategy at this moment is Nash Equilibrium strategy p ₁ ^*..., p _K ^*

Consider the situation of many antennas.In order to make analysis easy, we have adopted the notion of Virtual User.Promptly every transmit antennas is regarded as a Virtual User, and the user in the real system is a set of these Virtual User.On formulation, promptly to do a Virtual User mapping, K user is mapped to Kt user.

Mapping relations following (mapping 1):

$p_x^l=p_{\mathrm{ms}\left(x\right)\mathrm{an}\left(x\right)}^l,{\mathrm{\&gamma;}}_x^l={\mathrm{\&gamma;}}_{\mathrm{ms}\left(x\right)\mathrm{an}\left(x\right)}^l,h_x^l=h_{\mathrm{ms}\left(x\right)\mathrm{an}\left(x\right)}^l---\left(22\right)$

Wherein, x=1 ... K ... Kt, ms (x)=ceil (x/t), expression user ordinal number; An (x)=mod (x-1, t)+1, expression transmitting antenna ordinal number.By shining upon 1, we just can simplify the power control problem in the resolution system.Specific as follows:

By shining upon 1, (8) formula and (9) formula can be rewritten as:

${\mathrm{\&gamma;}}_x^l=\frac{h_x^l{p}_x^l}{{\mathrm{\&Sigma;}}_{y=1,y\&NotEqual;x}^K{h}_y^l{p}_y^l+{\mathrm{\&sigma;}}^2}---\left(23\right)$

$I_x^l={\mathrm{\&Sigma;}}_{y=1,y\&NotEqual;x}^K{h}_y^l{p}_y^l+{\mathrm{\&sigma;}}^2---\left(24\right)$

According to (20) and (21) formula, calculate the Nash Equilibrium strategy according to algorithm 1 afterwards, just number of users becomes Kt, and ordinal number i becomes x.After calculating the Nash Equilibrium strategy of each Virtual User, do the reflection of mapping 1 again and penetrate, just can obtain each actual user's Nash Equilibrium strategy.

When r=t=1, multiaerial system has just deteriorated to a single aerial system, and do not need just can directly use algorithm 1 by shining upon 1 this moment, we can say that single-antenna case is a special case of many antennas situation.

Technical scheme for a better understanding of the present invention, below provide the instantiation of the distributed power control of single sub-district single antenna and multi-antenna multi-carrier-wave wireless cellular system up link, the power control algorithm of down link just can obtain by simple conversion, and the situation of many sub-districts is similarly in addition.

(embodiment one)

Consider a up link that adopts single sub-district wireless cellular system of multi-antenna multi-carrier-wave Direct-Spread code division multiple access (MIMO MCDS-CDMA) technology, this system has K=6 user, r=t=4.Cost function adopts linear forms commonly used, promptly $C_i^l\left(p_i^l\right)={\mathrm{\&lambda;}}_i{p}_i^l,$ λ _iFor greater than 0 real number.Channel model adopts the costa207 standard.Three class business are arranged, conversation type (Class1), data stream type (type 2), interactive type (type 3) in the supposing the system.At the different demands of this three classes business, the parameter of power control is set to respectively:

Class1: a=10, λ=200, SINR _Min=5;

Type 2:a=5, λ=500, SINR _Min=10;

Type 3:a=2, λ=1000, SINR _Min=7.

Suppose among 6 users that user 1,2 is conversation type (Class1) user, user 3,4 is data stream type (type 2) user, and user 5,6 is interactive type (type 3) user.That is: a ₁=a ₂=10, λ ₁=λ ₂=200, ${\mathrm{<figure-callout\; id="1"\; label="SINR\; min"\; filenames="A20061004038500181.png,A20061004038500191.png"\; state="\{\{state\}\}">SINR</figure-callout>}}_{\min }^{}={\mathrm{<figure-callout\; id="2"\; label="SINR\; min"\; filenames="A20061004038500212.png"\; state="\{\{state\}\}">SINR</figure-callout>}}_{\min }^{}=5;$ a ₃＝a ₄＝5，λ ₃＝λ ₄＝500， ${\mathrm{<figure-callout\; id="3"\; label="SINR\; min"\; filenames="A20061004038500212.png"\; state="\{\{state\}\}">SINR</figure-callout>}}_{\min }^{}={\mathrm{<figure-callout\; id="4"\; label="SINR\; min"\; filenames="A20061004038500212.png"\; state="\{\{state\}\}">SINR</figure-callout>}}_{\min }^{}=10;$ a ₅＝a ₆＝2，λ ₅＝λ ₆＝1000， ${\mathrm{<figure-callout\; id="5"\; label="SINR\; min"\; filenames="A20061004038500191.png,A20061004038500192.png"\; state="\{\{state\}\}">SINR</figure-callout>}}_{\min }^{}={\mathrm{<figure-callout\; id="6"\; label="SINR\; min"\; filenames="A20061004038500212.png"\; state="\{\{state\}\}">SINR</figure-callout>}}_{\min }^{}=7.$ Concrete implementation step then of the present invention is as follows:

1, each user's initialization power vector p ₁ ⁰..., p ₆ ⁰, make it for being 1.Carry out the Virtual User mapping, obtain

p _V1 ⁰..., p _V24 ⁰, make n=1, n _Max=50.

2, make i=1, l=1.

3, the base station end is measured the I on l subcarrier of i Virtual User _i ^l(n), estimate channel link gain h _i ^l(n), determine good each parameter value, (21) formula of utilization calculates b _i, (20) formula of utilization calculates

If

Less than target Signal to Interference plus Noise Ratio value, then this Virtual User stops transmission, even transmitting power is 0, and jumps to step 5; Otherwise end calculates in the base station

And feeding back to Virtual User i, Virtual User i utilizes (6) formula to calculate p _i ^l(n+1), with seasonal i=i+1, repeat this step, till i=Kt=24.

4, if l=L=64 then jumps to 5; Otherwise, make l=l+1, and get back to 3.

5, make n=n+1, judge: if n=n _Max=50, then enter 6; Otherwise, judge p _V1 ^N+1-p _V1 ⁿ..., p _V24 ^N+1-p _V24 ⁿWhether less than 10-10, if all less than, then enter step 6, otherwise get back to 2.

6, stop iteration, each Virtual User transmitting power strategy at this moment is Nash Equilibrium strategy p _V1 ^*..., p _V24 ^*

7, shine upon 1 reflection and penetrate, obtain: p ₁ ⁰..., p ₆ ⁰

(embodiment two)

In addition, when r=t=1, other parameter and hypothesis are constant, when promptly considering single antenna multicarrier system, following concrete implementation step are arranged:

1, each user's initialization power vector p ₁ ⁰..., p ₆ ⁰, make it for being 1.Make n=1, n _Max=50.

2, make i=1, l=1.

3, the base station end is measured the I on l subcarrier of i Virtual User _i ^l(n), estimate channel link gain h _i ^l(n), determine good each parameter value, (21) formula of utilization calculates b _i, (20) formula of utilization calculates

If

Less than target Signal to Interference plus Noise Ratio value, then this Virtual User stops transmission, even transmitting power is 0, and jumps to step 5; Otherwise end calculates in the base station

And feeding back to Virtual User i, Virtual User i utilizes (6) formula to calculate p _i ^l(n+1), with seasonal i=i+1, repeat this step, till i=K=6.

4, if l=L=64 then jumps to 5; Otherwise, make l=l+1, and get back to 3.

5, make n=n+1, judge: if n=n _Max=50, then enter 6; Otherwise, judge p _V1 ^N+1-p _V1 ⁿ..., p _V6 ^N+1-p _V6 ⁿWhether less than 10 ^-10If, all less than, then enter step 6, otherwise get back to 2.

6, stop iteration, each Virtual User transmitting power strategy at this moment is Nash Equilibrium strategy p _V1 ^*..., p _V6 ^*

The present invention utilizes the Virtual User mapping by the new utility function of design, has solved the distributed power control problem in the multi-carrier-wave wireless honeycomb communication system, has effectively utilized system resource.

Claims (3)

1, a kind of Poewr control method of multi-carrier-wave wireless honeycomb communication system is characterized in that, comprises following steps:

Step 1 makes n=1, definition n _Max, each user's initialization power vector p ₁ ⁰..., p _K ⁰,

Step 2 makes i=1, l=1,

Step 3 definition $\underset{p_i}{\max }{\mathrm{NU}}_i^{\mathrm{MC}}=\underset{p_i^1,\&CenterDot;\&CenterDot;\&CenterDot;,p_i^L}{\mathrm{ma}x}[\underset{l}{\overset{L}{\mathrm{\&Sigma;}}}{U}_i^l\left({\mathrm{\&gamma;}}_i^l\right)-\underset{l}{\overset{L}{\mathrm{\&Sigma;}}}{C}_i^l\left(p_i^l\right)]$ , wherein $U\left({\mathrm{\&gamma;}}_i^l\right)=1-\frac{1}{1+{({\mathrm{\&gamma;}}_i^l/b_i)}^{a_i}}$ For from utility function, at first the base station end is measured interference and the noise sum I on l subcarrier of i user _i ^l(n), estimate channel link gain h _i ^l(n), determine good each parameter value, utilize then $b_i={\mathrm{SINR}}_{\min }^i/{(a_i-1)}^{\frac{1}{a_i}}$ Calculate parameter b _i, utilize $\frac{{\&PartialD;U}_i^l}{\&PartialD;p_i^l}=\frac{a_i{({\mathrm{\&gamma;}}_i^l/b_i)}^{a_i}}{{(1+{({\mathrm{\&gamma;}}_i^l/b_i)}^{a_i})}^2}-\frac{{\&PartialD;C}_i^l\left(p_i^l\right)}{\&PartialD;p_i^l}=0$ Calculate best Signal to Interference plus Noise Ratio value

If

Less than target Signal to Interference plus Noise Ratio value, then this user stops transmission, even transmitting power is 0, and jumps to step 5; If greater than, then calculate at the base station end

, and feeding back to user i, user i utilizes ${\hat{p}}_i(n+1)=p_i\left(n\right){\widehat{\mathrm{\&gamma;}}}_i\left(k\right)/{\mathrm{\&gamma;}}_i\left(k\right)$ Calculate p _i ^l(n+1),, repeat this step, till i=K, so just can obtain optimum transmitting power with seasonal i=i+1 ${\hat{p}}_i=\underset{p_i\&GreaterEqual;0}{\arg \max }{\mathrm{NU}}_i^{\mathrm{MC}},$

Step 4 is if l=L then jumps to step 5: if not, make l=l+1, and get back to step 3,

Step 5 makes n=n+1, if n is greater than n _Max, then jump to step 6, if less than but p ₁ ^N+1-p ₁ ⁿ..., p _K ^N+1-p _K ⁿSatisfy required precision, then enter step 6,, then get back to step 2 if do not satisfy required precision,

Step 6 stops iteration, and each user transmitting power vector at this moment is Nash Equilibrium strategy p ₁ ^*..., p _K ^*

2, the Poewr control method of multi-carrier-wave wireless honeycomb communication system according to claim 1 is characterized in that, above-mentionedly from utility function is $U\left(\mathrm{\&gamma;}\right)=1-\frac{1}{1+{(\mathrm{\&gamma;}/b)}^a},$ Strict satisfied $U\left(0\right)=1-\frac{1}{1+0}=0,$ U (∞)=∞, U (γ) is the requirement about the increasing function of γ.

3, the Poewr control method of multi-carrier-wave wireless honeycomb communication system according to claim 1 is characterized in that, and is above-mentioned from utility function $U\left(\mathrm{\&gamma;}\right)=1-\frac{1}{1+{(\mathrm{\&gamma;}/b)}^a},$ The value of a is relevant with type of service, for the demanding professional a of real-times such as speech business 〉=10, real-time is required business 4≤a≤8 suitable with the data stream class service, and real-time is required and professional suitable business 1≤a≤3 of interactive type; And the value of b can be utilized equation by given minimum Signal to Interference plus Noise Ratio value and a $b_i={\mathrm{SINR}}_{\min }^i/{(a_i-1)}^{\frac{1}{a_i}}$ Calculate.

Images