CN102149203A - Power allocation method in cognition orthogonal frequency division multiple access (OFDMA) system based on proportional fairness and interference constraints - Google Patents

Abstract

The invention discloses a power allocation method in a cognition orthogonal frequency division multiple access (OFDMA) system based on proportional fairness and interference constraints, comprising the following steps: firstly carrying out power allocation among cognitive users in accordance with proportional fairness factors, total power constraints and permanent speed values; then using a quick water injection method to carry out carrier power allocation by each cognitive user; if the carrier power of each cognitive user does not exceed a power constraint value, ending an algorithm, and otherwise, modifying allocated power of a carrier exceeding the power constraint to be the power constraint value; calculating the speed of the carrier and adding the speed to the permanent speed value of the user; eliminating the carrier from the carrier set of the user; and if the carrier set of each cognitive user is a null set, ending the algorithm, otherwise, subtracting the allocated power constraint value from a total power value, and repeating the above steps until the algorithm ends. The method is better in performances of the system capacity and the proportional fairness among the cognitive users, low in the algorithmic complexity and applied to engineering.

Description

In the cognitive OFDMA system based on the power distribution method of equitable proportion and interference constraints

Technical field

The present invention relates to wireless communication field, be specifically related in a kind of cognitive OFDMA system power distribution method based on equitable proportion and interference constraints.

Background technology

The key technology that cognitive radio technology is the current frequency spectrum resource scarcity of solution, the availability of frequency spectrum is low.In cognitive radio, cognitive user is used idle frequency range by dynamically adjusting parameters such as transmitting power, frequency and modulation system under the prerequisite of not disturbing authorized user, thereby has improved the utilization ratio of frequency spectrum effectively.It is the transmission technology that is easy to realize frequency spectrum resource control of generally acknowledging at present that OFDM inserts (OFDMA) technology.The OFDMA technology can be by frequency spectrum combination and cutting distributing system resource between the user neatly, this makes it to combine well with cognitive system.In cognitive OFDMA system, how under the prerequisite that satisfies various constraintss (interference constraints, equitable proportion constraint, total power constraint etc.), effectively resource (carrier wave, power etc.) is reasonably distributed to each cognitive user, become the key that improves cognitive OFDMA systematic function.

The existing resources distribution method mainly is divided into two big classes: static resource allocation method (see " MultiuserOFDM; " in IEEE International Symposium on Signal Processing and itsApplications, 1999) and dynamic resource allocation method (see " On the Use of Liner Programming forDynamic Subchannel and Bit Allocation in Multiuser OFDM; " in IEEE GLOBECOM, 2001).Static resource allocation method is as TDMA, FDMA etc., though simple owing to the dynamic-change information of reasonably not utilizing cognitive user and channel makes that the performance of this resource allocation method is limited.On the contrary, dynamic Resources allocation rule can reasonably be utilized the diversity gain between cognitive user and performance is improved a lot than static resource allocation method, thereby has obtained very big concern.At present the dynamic resource allocation method about cognitive OFDMA system has a lot, but two problems of basic existence, the one, the algorithm more complicated is not suitable for engineering application; The 2nd, seldom consider the constraint of equitable proportion, a lot of work all concentrate on the optimization of considering whole volume, and have ignored the fairness problem between cognitive user.When in the resource allocation problem that the fairness between cognitive user constraint is joined cognitive OFDMA system, the complexity of problem can improve greatly.

We are by carrying out modeling analysis to the cognitive OFDMA system power assignment problem of considering the equitable proportion constraint, utilize the architectural feature of nonlinear equation in the model to carry out rational simplification derivation, thereby with the power division problem reduction between cognitive user is the equation solution process of asking unitary variant, has simplified finding the solution of this problem greatly.Simultaneously, carrier power for single user is distributed, need iterative computation to obtain the defective of a rational water filling thresholding in order to overcome traditional water-filling algorithm, we have proposed a kind of quick water-filling algorithm that is based upon on the carrier signal to noise ratio ordering basis, and this algorithm can a step obtain the water filling thresholding.Ignore the computation complexity (this process can realize in the process of allocation of carriers) of carrier signal to noise ratio ordering, the complexity of this quick water-filling algorithm is O (N), only is equivalent to the complexity of an iteration of traditional waterflood algorithm.

Summary of the invention

The objective of the invention is to overcome the deficiencies in the prior art, propose in a kind of cognitive OFDMA system power distribution method based on equitable proportion and interference constraints.

Step based on the power distribution method of equitable proportion and interference constraints in the cognitive OFDMA system is as follows:

1) following variable is carried out initialization,

P _{K, total}=0,

R _{K, fixed}=0, Ω _{K, left}=Ω _k, for k=2,3 ..., K, wherein P _TotalRepresent total power constraint value,

Represent current total power constraint value, P _{K, total}The power of k cognitive user, Ω are distributed in expression _kThe carrier set of k cognitive user, Ω are distributed in expression _{K, fixed}Expression belongs to the carrier set of the fixed rate of k cognitive user, Ω _{K, left}Expression Ω _kIn remove Ω _{K, fixed}In element after remaining element set, K is the cognitive user number;

2) according to the equitable proportion factor gamma of cognitive user speed ₁: γ ₂...: γ _K, Ω _{K, left}In carrier channel noise ratio, current fixed rate R _{K, fixed}And current gross power

Calculate the power P of distributing to each cognitive user _{K, total}

3) each cognitive user is according to the performance number P that gets _{K, total}, utilize quick water flood to Ω _{K, left}In carrier wave carry out power division;

4) each cognitive user detects Ω _{K, left}In carrier power whether all satisfy interference constraints condition: P _{K, n}≤ P _{K, max}, n ∈ Ω _{K, left}, P wherein _{K, n}Represent the power that k cognitive user distributed on n carrier wave, P _{K, max}The expression set omega _{K, left}The power constraint value of middle carrier wave all satisfies the power constraint condition if the carrier power of all cognitive user is distributed, and then algorithm finishes, otherwise, detect P _{K, n}＞P _{K, max}K cognitive user n carrier wave put under the carrier set Ω of fixed rate _{K, fixed}In, and with n carrier wave from set omega _{K, left}K cognitive user fixed rate R upgraded in middle deletion _{K, fixed}For: R _{K, fixed}=R _{K, fixed}+ log ₂(1+P _{K, max}H _{K, n}), H wherein _{K, n}=| h _{K, n}| ²/ (N ₀BN ^-1), h _{K, n}Be Ω _{K, left}In the carrier gains of n carrier wave, N ₀Be the channel noise power spectrum density, B is total bandwidth, and N is total carrier wave number, upgrades current total power constraint value

For:

N wherein _{K, fixed}Be set omega _{K, fixed}Middle carrier wave number changes step 2 over to), finish until algorithm.

Step 2) the equitable proportion factor gamma described in according to cognitive user speed ₁: γ ₂...: γ _K, Ω _{K, left}In carrier channel noise ratio, current fixed rate R _{K, fixed}And current gross power

Calculate the power P of distributing to each cognitive user _{K, total}Step is:

The power P of cognitive user _{K, total}Computing formula is as follows

$F\left(P_{1,\mathrm{total}}\right)=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{a}_k{\left(P_{1,\mathrm{total}}\right)}^{b_k}-{\tilde{P}}_{\mathrm{total}}=0$

$P_{k,\mathrm{total}}=a_k{\left(P_{1,\mathrm{total}}\right)}^{b_k}$

Wherein,

$a_k=\left\{\begin{array}{cc}1,& k=1\\ {\left(\frac{H_{1,\mathrm{mid}}{M}_{1,\mathrm{left}}}{N_{1,\mathrm{left}}}{2}^{\frac{R_{1,\mathrm{fixed}}}{N_{1\_\mathrm{left}}}}\right)}^{\frac{N_{1,\mathrm{left}}{\mathrm{\γ}}_k}{N_{k,\mathrm{left}}{\mathrm{\γ}}_1}}\·\frac{N_{k,\mathrm{left}}}{H_{k,\mathrm{mid}}{M}_{k,\mathrm{left}}}{2}^{\frac{R_{k,\mathrm{fixed}}}{N_{k,\mathrm{left}}}},& k=\mathrm{2,3},...,K\end{array}\right.$

$b_k=\left\{\begin{array}{cc}1,& k=1\\ \frac{N_{1,\mathrm{left}}{\mathrm{\γ}}_k}{N_{k,\mathrm{left}}{\mathrm{\γ}}_1},& k=\mathrm{2,3},...,K\end{array}\right.,$

a _kWith b _kIn variable declaration: H _{K, mid}=| h _{K, mid}| ²/ (N ₀BN ^-1), h _{K, mid}Be Ω _{K, left}The median of middle carrier gains, N ₀Be the channel noise power spectrum density, B is total bandwidth, and N is total carrier wave number.

N _{K, left}Be Ω _{K, left}Middle carrier wave number.

Each cognitive user described in the step 3) is according to the performance number P that gets _{K, total}, utilize quick water flood to Ω _{K, left}In carrier wave carry out the power division step and be:

(1) determines interval [α _n, α _N+1], wherein

Satisfy in this interval

Thereby obtain gathering Ф _k={ i|i≤n, i ∈ Ω _{K, left}, if for the first time to Ω _{K, left}In carrier wave carry out the fast power water filling, and Ω _{K, left}In the carrier channel noise ratio satisfy

Then according to interval left side dividing value α _nWith interval the right dividing value α _N+1Definition directly calculate and satisfy P _{K, total}∈ [α _n, α _N+1] the carrier wave sequence number n that requires, if not for the first time to Ω _{K, left}In carrier wave carry out the fast power water filling, then only need on previous basis interval left side dividing value α _nWith interval the right dividing value α _N+1Adjust, do not need to recomputate, suppose that the water injection power of l carrier wave in last quick the injecting process has once exceeded the power constraint value, then l carrier wave will be from set omega _{K, left}Middle rejecting do not participated in the quick the injecting process of epicycle, then to interval left side dividing value α _nWith interval the right dividing value α _N+1Adjust: at first revise interval left side dividing value α _nValue be

Revise interval the right dividing value α _N+1Value be

Rule of judgment

Whether still satisfy, be then interval left side dividing value α _nWith interval the right dividing value α _N+1Value revise and to finish, otherwise if

Then n+1 carrier wave added set Φ _kIn, i.e. Φ _k={ i|i≤n+1, i ∈ Ω _{K, left}The interval left side dividing value α of modification _nValue be Revise interval the right dividing value α _N+1Value be

Repeat above step until condition

Satisfy;

(2) calculate fast power water filling thresholding:

M wherein _kBe set Φ _kThe number of middle element.

The present invention makes full use of the model feature of power division between cognitive user under equitable proportion and the interference constraints, with the power division between cognitive user effectively abbreviation be the equation solution process of unitary variant, greatly reduce the computation complexity of this problem.The quick water flood that is directed to single user's carrier power assignment problem proposition has simultaneously also been avoided iterative process, can determine the water filling thresholding by once calculating.

Description of drawings

Fig. 1 is based on the flow chart of the power distribution method of equitable proportion and interference constraints in the cognitive OFDMA system;

Fig. 2 is that proportional fairness between cognitive user is with the situation of change of total power constraint value, compared in the analogous diagram that algorithm of the present invention and traversal are separated (traversal all situations obtain optimal solution), TDMA algorithm and with reference to algorithm (" Increasing in Capacity of Multiuser OFDM System Using DynamicSubchannel Allocation ", in Proc.IEEE VTC, 2000) fairness performance

Fig. 3 is the situation of change of the power system capacity gain of cognitive user with the total power constraint value, compared in the analogous diagram algorithm of the present invention and traversal separate (traversal all situations obtain optimal solution), with reference to algorithm (" Increasing in Capacity of Multiuser OFDM System Using Dynamic SubchannelAllocation ", in Proc.IEEE VTC, 2000) with respect to the capacity gain of TDMA algorithm, its value is obtained by the power system capacity of each algorithm capacity divided by the TDMA algorithm under the equal conditions.

Embodiment

Step based on the power distribution method of equitable proportion and interference constraints in the cognitive OFDMA system is as follows:

1) following variable is carried out initialization, P _{K, total}=0,

R _{K, fixed}=0, Ω _{K, left}=Ω _k, fork=2,3 ..., K, wherein P _TotalRepresent total power constraint value,

Represent current total power constraint value, P _{K, total}The power of k cognitive user, Ω are distributed in expression _kThe carrier set of k cognitive user, Ω are distributed in expression _{K, fixed}Expression belongs to the carrier set of the fixed rate of k cognitive user, Ω _{K, left}Expression Ω _kIn remove Ω _{K, fixed}In element after remaining element set, K is the cognitive user number;

2) according to the equitable proportion factor gamma of cognitive user speed ₁: γ ₂...: γ _K, Ω _{K, left}In carrier channel noise ratio, current fixed rate R _{K, fixed}And current gross power

Calculate the power P of distributing to each cognitive user _{K, total}

3) each cognitive user is according to the performance number P that gets _{K, total}, utilize quick water flood to Ω _{K, left}In carrier wave carry out power division;

4) each cognitive user detects Ω _{K, left}In carrier power whether all satisfy interference constraints condition: P _{K, n}≤ P _{K, max}, n ∈ Ω _{K, left}, P wherein _{K, n}Represent the power that k cognitive user distributed on n carrier wave, P _{K, max}The expression set omega _{K, left}The power constraint value of middle carrier wave all satisfies the power constraint condition if the carrier power of all cognitive user is distributed, and then algorithm finishes, otherwise, detect P _{K, n}＞P _{K, max}K cognitive user n carrier wave put under the carrier set Ω of fixed rate _{K, fixed}In, and with n carrier wave from set omega _{K, left}K cognitive user fixed rate R upgraded in middle deletion _{K, fixed}For: R _{K, fixed}=R _{K, fixed}+ log ₂(1+P _{K, max}H _{K, n}), H wherein _{K, n}=| h _{K, n}| ²/ (N ₀BN ^-1), h _{K, n}Be Ω _{K, left}In the carrier gains of n carrier wave, N ₀Be the channel noise power spectrum density, B is total bandwidth, and N is total carrier wave number, upgrades current total power constraint value

For:

N wherein _{K, fixed}Be set omega _{K, fixed}Middle carrier wave number changes step 2 over to), finish until algorithm.

Step 2) the equitable proportion factor gamma described in according to cognitive user speed ₁: γ ₂...: γ _K, Ω _{K, left}In carrier channel noise ratio, current fixed rate R _{K, fixed}And current gross power

Calculate the power P of distributing to each cognitive user _{K, total}Step is:

The power P of cognitive user _{K, total}Computing formula is as follows

$F\left(P_{1,\mathrm{total}}\right)=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{a}_k{\left(P_{1,\mathrm{total}}\right)}^{b_k}-{\tilde{P}}_{\mathrm{total}}=0$

$P_{k,\mathrm{total}}=a_k{\left(P_{1,\mathrm{total}}\right)}^{b_k}$

Wherein,

$a_k=\left\{\begin{array}{cc}1,& k=1\\ {\left(\frac{H_{1,\mathrm{mid}}{M}_{1,\mathrm{left}}}{N_{1,\mathrm{left}}}{2}^{\frac{R_{1,\mathrm{fixed}}}{N_{1\_\mathrm{left}}}}\right)}^{\frac{N_{1,\mathrm{left}}{\mathrm{\γ}}_k}{N_{k,\mathrm{left}}{\mathrm{\γ}}_1}}\·\frac{N_{k,\mathrm{left}}}{H_{k,\mathrm{mid}}{M}_{k,\mathrm{left}}}{2}^{\frac{R_{k,\mathrm{fixed}}}{N_{k,\mathrm{left}}}},& k=\mathrm{2,3},...,K\end{array}\right.$

$b_k=\left\{\begin{array}{cc}1,& k=1\\ \frac{N_{1,\mathrm{left}}{\mathrm{\γ}}_k}{N_{k,\mathrm{left}}{\mathrm{\γ}}_1},& k=\mathrm{2,3},...,K\end{array}\right.,$

a _kWith b _kIn variable declaration: H _{K, mid}=| h _{K, mid}| ²/ (N ₀BN ^-1), h _{K, mid}Be Ω _{K, left}The median of middle carrier gains, N ₀Be the channel noise power spectrum density, B is total bandwidth, and N is total carrier wave number.

N _{K, left}Be Ω _{K, left}Middle carrier wave number.

Each cognitive user described in the step 3) is according to the performance number P that gets _{K, total}, utilize quick water flood to Ω _{K, left}In carrier wave carry out the power division step and be:

(1) determines interval [α _n, α _N+1], wherein Satisfy in this interval

Thereby obtain gathering Φ _k={ i|i≤n, i ∈ Ω _{K, left}, if for the first time to Ω _{K, left}In carrier wave carry out the fast power water filling, and Ω _{K, left}In the carrier channel noise ratio satisfy

Then according to interval left side dividing value α _nWith interval the right dividing value α _N+1Definition directly calculate and satisfy P _{K, total}∈ [α _n, α _N+1] the carrier wave sequence number n that requires, if not for the first time to Ω _{K, left}In carrier wave carry out the fast power water filling, then only need on previous basis interval left side dividing value α _nWith interval the right dividing value α _N+1Adjust, do not need to recomputate, suppose that the water injection power of l carrier wave in last quick the injecting process has once exceeded the power constraint value, then l carrier wave will be from set omega _{K, left}Middle rejecting do not participated in the quick the injecting process of epicycle, then to interval left side dividing value α _nWith interval the right dividing value α _N+1Adjust: at first revise interval left side dividing value α _nValue be

Revise interval the right dividing value α _N+1Value be Rule of judgment

Whether still satisfy, be then interval left side dividing value α _nWith interval the right dividing value α _N+1Value revise and to finish, otherwise if

Then n+1 carrier wave added set Φ _kIn, i.e. Ф _k={ i|i≤n+1, i ∈ Ω _{K, left}The interval left side dividing value α of modification _nValue be

Revise interval the right dividing value α _N+1Value be

Repeat above step until condition

Satisfy;

(2) calculate fast power water filling thresholding:

M wherein _kBe set Ф _kThe number of middle element.

Embodiment

Allocation of carriers process in the present embodiment does not belong to the research range of this algorithm, employing is with reference to algorithm (" Increasing in Capacity of Multiuser OFDM System Using Dynamic SubchannelAllocation ", in Proc.IEEE VTC, 2000) the allocation of carriers mode in, be by reference algorithm (" Increasing in Capacity of Multiuser OFDMSystem Using Dynamic Subchannel Allocation " with the power on each carrier wave in the allocation of carriers process, in Proc.IEEE VTC, 2000) mean allocation in changes the carrier power binding occurrence in this example into.Understand for convenience, now with amended list of references (" Increasing in Capacity of Multiuser OFDM System Using Dynamic SubchannelAllocation ", in Proc.IEEE VTC, 2000) the allocation of carriers mode in simply is described below:

(1) carrier set of each cognitive user of initialization

The speed R of each cognitive user of initialization _k=0, k=1,2 ... K, Ω are total available carrier wave set, wherein total N=14 available carrier wave, the carrier gains h of cognitive user 1 _{1, n}Between [0.84,0.99], produce the carrier gains h of cognitive user 2 at random _{2, n}Between [0.85,1.04], produce the noise power spectral density N of each carrier wave at random ₀=1mW, the value of total bandwidth B is made as 1;

(2) find out cognitive user k, satisfy

Selecting carrier wave n by cognitive user k satisfies: H _{K, n}〉=H _{K, m}, N ∈ Ω, H _{K, n}=| h _{K, n}| ²/ (N ₀BN ^-1), the speed of upgrading cognitive user k is R _k=R _k+ Blog ₂(1+H _{K, n}P _{K, max}), the carrier set Ω of renewal cognitive user k _kBe Ω _k=Ω _k∪ { n};

(3) carrier wave n is rejected from set omega.If

Then allocation of carriers finishes, otherwise gets back to step (2), finishes until allocation of carriers, obtains the carrier set Ω of each cognitive user _k, k=1,2 ... K.

Step based on the power distribution method of equitable proportion and interference constraints in the cognitive OFDMA system is as follows:

1) initializing variable:

P _{K, total}=0, R _{K, fixed}=0, Ω _{K, left}=Ω _k, for k=2,3 ..., K.P wherein _TotalRepresent total power constraint value, value is observed the performance of algorithm under the different total power constraint respectively from 10mW to 15mW,

Represent current total power constraint value, P _{K, total}The power of k cognitive user, Ω are distributed in expression _kThe carrier set of k cognitive user is distributed in expression, obtains Ω by foregoing allocation of carriers process _{K, fixed}Expression belongs to the carrier set of the fixed rate of k cognitive user, Ω _{K, left}Expression Ω _kIn remove Ω _{K, fixed}In element after remaining element set, K=2 is the cognitive user number, the carrier power binding occurrence P of cognitive user _{1, max}=P _{2, max}=1mW;

2) according to the equitable proportion factor gamma of cognitive user speed ₁: γ ₂=1: 1, Ω _{K, left}In carrier channel noise ratio, current fixed rate R _{K, fixed}And current gross power

Calculate the power P of distributing to each cognitive user _{K, total}

3) each cognitive user is according to the performance number P that gets _{K, total}, utilize quick water flood to Ω _{K, left}In carrier wave carry out power division;

4) each cognitive user detects Ω _{K, left}In the distribution power of carrier wave whether all satisfy interference constraints condition: P _{K, n}≤ P _{K, max}, n ∈ Ω _{K, left}, P wherein _{K, n}Represent the power that k cognitive user distributed on n carrier wave, P _{K, max}The expression set omega _{K, left}The power constraint value of middle carrier wave.All satisfy the power constraint condition if the carrier power of all cognitive user is distributed, then algorithm finishes, otherwise, detect P _{K, n}＞P _{K, max}K cognitive user n carrier wave put under the carrier set Ω of fixed rate _{K, fixed}In, and with n carrier wave from set omega _{K, left}K cognitive user fixed rate R upgraded in middle deletion _{K, fixed}For: R _{K, fixed}=R _{K, fixed}+ log ₂(1+P _{K, max}H _{K, n}), H _{K, n}=| h _{K, n}| ²/ (N ₀BV ^-1), h _{K, n}Be the carrier gains value of n the carrier wave of cognitive user k, the carrier gains h of cognitive user 1 _{1, n}Between [0.84,0.99], produce the carrier gains h of cognitive user 2 at random _{2, n}Between [0.85,1.04], produce at random.N ₀=1mW is the channel noise power spectrum density, and B is total bandwidth, and its value is assumed to be 1, and N=14 is total carrier wave number.Upgrade current total power constraint value

For: N wherein _{K, fixed}Be set omega _{K, fixed}Middle carrier wave number changes step 2 over to), finish until algorithm.

Step 2) the equitable proportion factor gamma described in according to cognitive user speed ₁: γ ₂=1, Ω _{K, left}In carrier channel noise ratio, current fixed rate R _{K, fixed}And current gross power

Calculate the power P of distributing to each cognitive user _{K, total}Step is: solve an equation

$F\left(P_{1,\mathrm{total}}\right)=\underset{k=1}{\overset{2}{\mathrm{\Σ}}}{a}_k{\left(P_{1,\mathrm{total}}\right)}^{b_k}-{\tilde{P}}_{\mathrm{total}}=0$

$P_{2,\mathrm{total}}=a_2{\left(P_{1,\mathrm{total}}\right)}^{b_2}$

Wherein,

$a_k=\left\{\begin{array}{cc}1,& k=1\\ {\left(\frac{H_{1,\mathrm{mid}}{M}_{1,\mathrm{left}}}{N_{1,\mathrm{left}}}{2}^{\frac{R_{1,\mathrm{fixed}}}{N_{1,\mathrm{left}}}}\right)}^{\frac{N_{1,\mathrm{left}}}{N_{k,\mathrm{left}}}}\·\frac{N_{k,\mathrm{left}}}{H_{k,\mathrm{mid}}{M}_{k,\mathrm{left}}}{2}^{\frac{R_{k,\mathrm{fixed}}}{N_{k,\mathrm{left}}}},& k=2\end{array}\right.$

$b_k=\left\{\begin{array}{cc}1,& k=1\\ \frac{N_{1,\mathrm{left}}}{N_{k,\mathrm{left}}},& k=2\end{array}\right.,$

a _kWith b _kIn variable declaration: H _{K, mid}=| h _{K, mid}| ²/ (N ₀BN ^-1), h _{K, mid}Be Ω _{K, left}The median of middle carrier gains, the carrier gains h of cognitive user 1 _{1, n}Between [0.84,0.99], produce the carrier gains h of cognitive user 2 at random _{2, n}Between [0.85,1.04], produce at random.N ₀=1mW is the channel noise power spectrum density, and B is total bandwidth, and its value is assumed to be 1, and N=14 is total carrier wave number. N _{K, left}Be Ω _{K, left}Middle carrier wave number.

Finding the solution of equation can utilize the Newton method or the method for false position (to see " Handbook ofMathematical Functions with Formulas, Graphs, and Mathematical Tables ", 9 in this step ^ThPrinting.New York:Dover, 1972.)

Each cognitive user described in the step 3) is according to the performance number P that gets _{K, total}, utilize quick water flood to Ω _{K, left}In carrier wave carry out the power division step and be:

(1) determines interval [α _n, α _N+1], wherein

Satisfy in this interval

Thereby obtain gathering Ф _k={ i|i≤n, i ∈ Ω _{K, left}, if for the first time to Ω _{K, left}In carrier wave carry out the fast power water filling, and Ω _{K, left}In the carrier channel noise ratio satisfy

Then according to interval left side dividing value α _nWith interval the right dividing value α _N+1Definition directly calculate and satisfy P _{K, total}∈ [α _n, α _N+1] the carrier wave sequence number n that requires.If not for the first time to Ω _{K, left}In carrier wave carry out the fast power water filling, then only need on previous basis interval left side dividing value α _nWith interval the right dividing value α _N+1Adjust, do not need to recomputate.The water injection power of supposing l carrier wave in last quick the injecting process has once exceeded the power constraint value, and then l carrier wave will be from set omega _{K, left}Middle rejecting do not participated in the quick the injecting process of epicycle, therefore revises interval left side dividing value α _nValue be

Revise interval the right dividing value α _N+1Value be

Rule of judgment

Whether still satisfy, be then interval left side dividing value α _nWith interval the right dividing value α _N+1Value revise and to finish, otherwise if

Then n+1 carrier wave added set Φ _kIn, i.e. Ф _k={ i|i≤n+1, i ∈ Ω _{K, left}, revise interval left side dividing value α _nValue be

Revise interval the right dividing value α _N+1Value be

Repeat above step until condition

Satisfy.

(2) the water filling thresholding of the quick water flood of calculating:

M wherein _kBe set Φ _kThe number of middle element.

Fig. 2 is that proportional fairness between cognitive user is with the situation of change of gross power, compared method of the present invention and traversal in the analogous diagram and separated (optimal solution that the traversal all situations obtains), TDMA method and reference method (" Increasing in Capacity of Multiuser OFDM System Using Dynamic SubchannelAllocation ", in Proc.IEEE VTC, 2000) proportional fairness performance, as can be seen from the figure, method of the present invention is compared to TDMA method and reference method when total power constraint value is lower than 14mW, can well guarantee the equitable proportion requirement between cognitive user.When total power constraint value exceeds 14mW, because all carrier powers have all exceeded the carrier power binding occurrence, the capacity of each cognitive user is by carrier power binding occurrence decision, thereby caused that traversal is separated, the coincidence of the inventive method and reference method three curve.And the TDMA method is different with former three owing to its allocation of carriers mode, thereby the equitable proportion performance of cognitive user is also different with former three.In addition, when total power constraint value during less than 12.5mW, from simulation result as can be seen, method of the present invention can be approached traversal well and be separated, but obtain by the traversal all situations owing to travel through the solution procedure of separating, so complexity is higher than method of the present invention far away.When total power constraint value greater than 12.5mW during less than 14mW, because the distribution power of part carrier wave has exceeded the power constraint value of carrier wave, need utilize method of the present invention that it is upgraded finds the solution, caused the result to have some to depart from traversal separating in order to simplify the part approximate calculation of calculating and adopting in the procedure, but in general, method of the present invention has one well to trade off on equitable proportion performance and method complexity.Fig. 3 is the situation of change of the power system capacity gain of cognitive user with gross power, compared method of the present invention and traversal in the analogous diagram and separated (optimal solution that the traversal all situations obtains), reference method (" Increasing in Capacity of Multiuser OFDMSystem Using Dynamic Subchannel Allocation ", in Proc.IEEE VTC, 2000) with respect to the capacity gain of TDMA method, its value is obtained by the power system capacity of each method capacity divided by the TDMA method under the equal conditions, and the downward trend of curve is because capacity difference reduces along with the increase of total power constraint.From figure curve as can be seen, when total power constraint value was lower than 14mW, the inventive method was better than TDMA method and reference method on power system capacity.When total power constraint value is higher than 14mW, because all carrier powers have all exceeded the power constraint value of carrier wave, the capacity of each cognitive user is by carrier power binding occurrence decision, thereby caused that traversal is separated, the coincidence of the inventive method and reference method three curve.Descend though the power system capacity of the inventive method is separated to some extent than traversal, the complexity that the complexity of the inventive method is separated well below traversal is more suitable in engineering application.

Claims (3)

1. In the cognitive OFDMA system based on the power distribution method of equitable proportion and interference constraints, it is characterized in that its step is as follows:

   1) following variable is carried out initialization, P _{K, total}=0,

   

   R _{K, fixed}=0, Ω _{K, left}=Ω _k, for k=2,3 ..., K, wherein P _TotalRepresent total power constraint value,

   

   Represent current total power constraint value, P _{K, total}The power of k cognitive user, Ω are distributed in expression _kThe carrier set of k cognitive user, Ω are distributed in expression _{K, fixed}Expression belongs to the carrier set of the fixed rate of k cognitive user, Ω _{K, left}Expression Ω _kIn remove Ω _{K, fixed}In element after remaining element set, K is the cognitive user number;

   2) according to the equitable proportion factor gamma of cognitive user speed ₁: γ ₂...: γ _K, Ω _{K, left}In carrier channel noise ratio, current fixed rate R _{K, fixed}And current gross power

   

   Calculate the power P of distributing to each cognitive user _{K, total}

   3) each cognitive user is according to the performance number P that gets _{K, total}, utilize quick water flood to Ω _{K, left}In carrier wave carry out power division;

   4) each cognitive user detects Ω _{K, left}In carrier power whether all satisfy interference constraints condition: P _{K, n}≤ P _{K, max}, n ∈ Ω _{K, left}, P wherein _{K, n}Represent the power that k cognitive user distributed on n carrier wave, P _{K, max}The expression set omega _{K, left}The power constraint value of middle carrier wave all satisfies the power constraint condition if the carrier power of all cognitive user is distributed, and then algorithm finishes, otherwise, detect P _{K, n}＞P _{K, max}K cognitive user n carrier wave put under the carrier set Ω of fixed rate _{K, fixed}In, and with n carrier wave from set omega _{K, left}K cognitive user fixed rate R upgraded in middle deletion _{K, fixed}For: R _{K, fixed}=R _{K, fixed}+ log ₂(1+P _{K, max}H _{K, n}), H wherein _{K, n}=| h _{K, n}| ²/ (N ₀BN ^-1), h _{K, n}Be Ω _{K, left}In the carrier gains of n carrier wave, N ₀Be the channel noise power spectrum density, B is total bandwidth, and N is total carrier wave number, upgrades current total power constraint value

   

   For: N wherein _{K, fixed}Be set omega _{K, fixed}Middle carrier wave number changes step 2 over to), finish until algorithm.
2. 2. based on the power distribution method of equitable proportion and interference constraints, it is characterized in that step 2 in a kind of cognitive 0FDMA according to claim 1 system) described in the equitable proportion factor gamma according to cognitive user speed ₁: γ ₂...: γ _K, Ω _{K, left}In carrier channel noise ratio, current fixed rate R _{K, fixed}And current gross power

   

   Calculate the power P of distributing to each cognitive user _{K, total}Step is: the power P of cognitive user _{K, total}Computing formula is as follows

   $F\left(P_{1,\mathrm{total}}\right)=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{a}_k{\left(P_{1,\mathrm{total}}\right)}^{b_k}-{\tilde{P}}_{\mathrm{total}}=0$

   $P_{k,\mathrm{total}}=a_k{\left(P_{1,\mathrm{total}}\right)}^{b_k}$

   Wherein,

   $a_k=\left\{\begin{array}{cc}1,& k=1\\ {\left(\frac{H_{1,\mathrm{mid}}{M}_{1,\mathrm{left}}}{N_{1,\mathrm{left}}}{2}^{\frac{R_{1,\mathrm{fixed}}}{N_{1\_\mathrm{left}}}}\right)}^{\frac{N_{1,\mathrm{left}}{\mathrm{\γ}}_k}{N_{k,\mathrm{left}}{\mathrm{\γ}}_1}}\·\frac{N_{k,\mathrm{left}}}{H_{k,\mathrm{mid}}{M}_{k,\mathrm{left}}}{2}^{\frac{R_{k,\mathrm{fixed}}}{N_{k,\mathrm{left}}}},& k=\mathrm{2,3},...,K\end{array}\right.$

   $b_k=\left\{\begin{array}{cc}1,& k=1\\ \frac{N_{1,\mathrm{left}}{\mathrm{\γ}}_k}{N_{k,\mathrm{left}}{\mathrm{\γ}}_1},& k=\mathrm{2,3},...,K\end{array}\right.,$

   a _kWith b _kIn variable declaration: H _{K, mid}=| h _{K, mid}| ²/ (N ₀BN ^-1), h _{K, mid}Be Ω _{K, left}The median of middle carrier gains, N ₀Be the channel noise power spectrum density, B is total bandwidth, and N is total carrier wave number.

   

   N _{K, left}Be Ω _{K, left}Middle carrier wave number.
3. 3. based on the power distribution method of equitable proportion and interference constraints, it is characterized in that each cognitive user described in the step 3) is according to the performance number P that gets in a kind of cognitive OFDMA according to claim 1 system _{K, total}, utilize quick water flood to Ω _{K, left}In carrier wave carry out the power division step and be:

   (1) determines interval [α _n, α _N+1], wherein

   

   Satisfy in this interval

   

   Thereby obtain gathering Φ _k={ i|i≤n, i ∈ Ω _{K, left}, if for the first time to Ω _{K, left}In carrier wave carry out the fast power water filling, and Ω _{K, left}In the carrier channel noise ratio satisfy

   

   Then according to interval left side dividing value α _nWith interval the right dividing value α _N+1Definition directly calculate and satisfy P _{K, total}∈ [α _n, α _N+1] the carrier wave sequence number n that requires, if not for the first time to Ω _{K, left}In carrier wave carry out the fast power water filling, then only need on previous basis interval left side dividing value α _nWith interval the right dividing value α _N+1Adjust, do not need to recomputate, suppose that the water injection power of l carrier wave in last quick the injecting process has once exceeded the power constraint value, then l carrier wave will be from set omega _{K, left}Middle rejecting do not participated in the quick the injecting process of epicycle, then to interval left side dividing value α _nWith interval the right dividing value α _N+1Adjust: at first revise interval left side dividing value α _nValue be Revise interval the right dividing value α _N+1Value be

   

   Rule of judgment

   

   Whether still satisfy, be then interval left side dividing value α _nWith interval the right dividing value α _N+1Value revise and to finish, otherwise if

   

   Then n+1 carrier wave added set Φ _kIn, i.e. Φ _k={ i|i≤n+1, i ∈ Ω _{K, left}The interval left side dividing value α of modification _nValue be

   

   Revise interval the right dividing value α _N+1Value be Repeat above step until condition

   

   Satisfy;

   (2) calculate fast power water filling thresholding:

   

   M wherein _kBe set Ф _kThe number of middle element.

Images