CN103580806B - A kind of method of robustness efficiency transmission in cognition network - Google Patents

Abstract

The invention discloses a kind of method of robustness efficiency transmission in cognition network, measured the channel condition information of correlated channels by secondary user's receiving terminal, and feed back to secondary user's transmitting terminal；Secondary user's transmitting terminal determines the uncertain set of each channel gain by repetitive measurement；The problem that secondary user's maximizes secondary user's efficiency according to robustness optimization method modeling；Solve worst channel gain and the constraint of primary user's QoS requirement is converted into convex constraint；Solve the transmit power of optimum according to power distribution algorithm, and be transmitted according to this power on each channel, solve the optimal transmission power of robust when channel condition information uncertain, can be used for cognition network communication.The present invention ensure that the maximization of secondary user's efficiency in the uncertain situation of channel information；Strictly control primary user's receiving terminal jamming power, it is ensured that the service quality of primary user, effectively eliminate the information uncertainty deterioration to cognition network performance.

Description

A kind of method of robustness efficiency transmission in cognition network

Technical field

The invention belongs to wireless communication technology field, particularly relate to a kind of method of robustness efficiency transmission in cognition network.

Background technology

Cognitive radio networks allows secondary user's to utilize the mandate frequency spectrum of primary user to carry out frequency spectrum share under not reducing the premise of service quality QoS of primary user, improves the utilization rate of whole frequency spectrum.On the other hand, the Radio Transmission Technology maximizing efficiency becomes the hot issue of current green communications.Especially, in cognitive radio networks, how secondary user's maximizes, while improving the availability of frequency spectrum, the hot issue that self efficiency is current research.

Existing major part document, when maximizing secondary user's efficiency, assumes that greatly the status information CSI of channel is accurately.But, due to the time-varying characteristics of wireless transmission channel, the impact of the factors such as the finiteness of training sequence and the time delay of feedback channel, the channel condition information CSI that transmitting terminal obtains is inaccurate often.And, inaccurate channel condition information can affect the energetic efficiency characteristic of secondary user's and the service quality of primary user.

Owing to the existing maximization efficiency method for cognitive radio networks is designed under assuming accurate channel information state.These transmission methods are applied in practical wireless systems, frequently can lead to the deterioration of secondary user's efficiency, also can reduce the service quality of primary user.

Summary of the invention

The purpose of the embodiment of the present invention is in that to provide a kind of method of robustness efficiency transmission in cognition network, aim to solve the problem that existing cognitive radio networks maximizes efficiency method and causes the deterioration of secondary user's efficiency, the problem reducing primary user's service quality because not considering channel information uncertain.

The method of transmission that the embodiment of the present invention is achieved in that in a kind of cognition network robustness efficiency, in this cognition network, the method for robustness efficiency transmission comprises the following steps:

Step one, secondary user's receiving terminal SR when not receiving useful data or the interference signal that receives of the pre-test of reception, estimates the primary user of the transmission estimated value to interference on channel k；When primary user mourns in silence, secondary user's receiving terminal SR measures the white noise power σ on channel k², when receiving data, estimate the channel gain of secondary user's transmitting terminal ST according to training sequence, numerical value is also sent to secondary user's transmitting terminal ST by feedback channel by secondary user's receiving terminal SR；

Step 2, secondary user's transmitting terminal ST, when not having data to send, demodulates the public transmission signal of primary user, as confirmed response ACK, negative response NAK etc., estimates the channel gain of primary user receiving terminal PR accordingly

Step 3, before communication starts, by repeatedly repeating step one to three, it is determined that the uncertain set of relevant parameter；

Step 4, according to the parameter that step one to three obtains, the descending efficiency f (P) that can multiple secondary user's be transmitted over multiple channels is expressed as

$f\left(P\right)=\frac{\underset{k=1}{\overset{3}{\mathrm{\Σ}}}{B\log }_2(1+\frac{{\mathrm{\Γ}}_k{P}_k{H}_s^k}{{\mathrm{\σ}}^2+I_P^k})}{P_C+5\underset{k=1}{\overset{3}{\mathrm{\Σ}}}{p}_k}$

Wherein, B is the bandwidth of channel, when adopting quadrature amplitude modulation MQAM when error rate BER=0.001 of secondary user's, and the now coefficient of variation Γ between channel speed and the channel capacity of actual transmissions system_k=0.1258, P_CThe cognitive base station CBS(secondary user's transmitting terminal ST being namely mentioned above) constant power expense, efficiency 1/ η of the power amplifier of cognitive base station CBS₀=0.20, P=[P₁,P₂,P₃] represent cognitive base station CBS transmit power vector on all channels, P here_k, { 1,2,3} represents the transmit power on channel k to k ∈；

Now, according to robustness optimization method, the problem maximizing secondary user's efficiency is modeled as

$P1:\underset{P_x}{\max }\·\underset{G_{\mathrm{Sj}},H_S^k,I_P^k}{\min }\·f\left(P\right)$

$s.t.C1:\underset{k=1}{\overset{3}{\mathrm{\Σ}}}{P}_k\≤P_{\max },$

$C2:P\·G_{\mathrm{Sj}}^T\≤{\mathrm{IT}}_j,\∀j\∈\left\{\mathrm{1,2}\right\},$

$C3:G_{\mathrm{Sj}}\∈Q_j,\∀j\∈\left\{\mathrm{1,2}\right\},$

$C4:H_S^k\∈F_k,\∀k\∈\left\{\mathrm{1,2,3}\right\},$

$C5:I_P^k\∈I_k,\∀k\∈\left\{\mathrm{1,2,3}\right\}\·$

Wherein, P_maxRepresent the maximum transmit power of cognitive base station CBS, IT_jIt it is the jth primary user receiving terminal PR interference threshold value that can tolerate；

Step 5, according to ${\left(H_S^k\right)}^{*}=\arg \underset{(H_S^k\∈F_k)}{\min }\·(1+\frac{{\mathrm{\Γ}}_k{P}_k{H}_S^k}{{\mathrm{BN}}_0+I_P^k})=\arg \underset{(H_S^k\∈F_k)}{\min }\·\left(H_S^k\right),\∀k\∈\left\{\mathrm{1,2,3}\right\},$ Solve secondary user's channel gain on channel kWorst-case valueWherein, worst-case value refers at given feasible transmit power vectorMake object function in problem P1MinimumValue；

Step 6, then solve the cumulative interference that secondary user's is caused by primary user on channel kWorst-case valueWherein, worst-case value refers at given feasible transmit power vectorMake object function in problem P1MinimumValue；；

Step 7, the correlation criterion according to robustness optimization, following equivalent convex constraint C6 can be converted into by P1 retraining C2 and C3:

$C6:P\·G_{\mathrm{Sj}}^T+{\mathrm{\ϵ}}_0{\left|\right|P\left|\right|}_2\≤{\mathrm{IT}}_j,\∀j\∈\left\{\mathrm{1,2}\right\}$

Wherein, weight matrix M_j=I_3×3, inverse matrixAdditionally, l₂Norm | | | |₂Dual norm be still l₂Norm | | | |₂, namely

Step 8, substitutes into the result in step 5 to seven in problem P1, can be converted into

$\begin{array}{c}P2:\underset{P_k}{\max }\·f_1\left(P\right)=\frac{U_R\left(P\right)}{U_{\mathrm{TP}}\left(P\right)}=\frac{\underset{k=1}{\overset{3}{\mathrm{\Σ}}}{B\log }_2(1+\frac{{\mathrm{\Γ}}_k{P}_k{\left(H_S^k\right)}^{*}}{{\mathrm{\σ}}^2+{\left(I_P^k\right)}^{*}})}{P_C+5\underset{k=1}{\overset{3}{\mathrm{\Σ}}}{P}_k}\\ s.t.C1,C6.\end{array};$

Step 9, adopts following power distribution algorithm, Solve problems P2:

The first step, given maximum iteration time L_max=50 and error tolerance threshold value κ=0.0001；

Second step, initializes and currently can index assignment n=0 to current iteration by valid value q=0 simultaneously；

3rd step, as convergence identifier F==0 and current iteration index n≤L_maxTime, order performs following statement；Otherwise, the 7th step is jumped to；

4th step, when maximum efficiency q gives timing, solves following convex optimization problem P3, it is thus achieved that power allocation vector P ' now；

$\begin{array}{c}P3:\underset{P}{\max }.U_R\left(P\right)-q{U}_{\mathrm{TP}}\left(P\right)\\ s.t.C1,C6.\end{array}$

5th step, if inequality U_R(P′)-qU_TP(P ') < κ set up, then assignment P*=P ' andConvergence mark assignment F=1；Otherwise, assignmentAnd n=n+1, convergence mark assignment F=0；

6th step, returns to the 3rd step, and continuation order performs correlative；

7th step, returns optimum energy valid value q^*With optimal power allocation vector P^*=[P₁ ^*,P₂ ^*,P₃ ^*]；

Step 10, cognitive base station CBS adopts given vector power P^*=[P₁ ^*,P₂ ^*,P₃ ^*], on channel 1 with power P₁ ^*Send data to secondary user's S1, channel 2 and 3 is respectively adopted P₂ ^*And P₃ ^*Data are sent to secondary user's S2.

Further, in step 3, it is determined that the uncertain set of relevant parameter is as follows:

Make ε₀=ε_j=τ_k=δ_k, { 1,2}, { 1,2, the 3} border representing unified uncertain set, generalized norm adopts l to k ∈ to j ∈ in addition₂Norm | | | |₂, for vector X ∈ R^1×n, l₂Norm | | | |₂It is represented by ${\left|\right|X\left|\right|}_2={({\left|x_1\right|}^2+...+{\left|x_n\right|}^2)}^{\frac{1}{2}}:$

The first step, for parameter G_Sj, for the receiving terminal of primary user P1, set up following uncertain set

$Q_1=\left\{G_{S1}\right|{\left|\right|{(G_{S1}-{\hat{G}}_{S1})}^T\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\}$

Wherein,Namely when primary user P1 is not transmitted on channel 2, nowMaking weight matrix value is unit matrix M₁=I_3×3∈R^3×3,It is G_S1Current estimated value；

For the receiving terminal of primary user P2, similar uncertain set can be obtained

$Q_2=\left\{G_{S2}\right|{\left|\right|{(G_{S2}-{\hat{G}}_{S2})}^T\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\}$

Wherein,Namely when primary user P1 is not transmitted on channel 1 and 3, nowMaking weight matrix value is unit matrix M₂=I_3×3∈R^3×3,It is G_S2Current estimated value；

Second step, about parameterFollowing uncertain set is set up for secondary user's S1

$F_1=\left\{H_S^1\right|{\left|\right|H_S^1-{\hat{H}}_S^1\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\}$

Following uncertain set is set up about secondary user's S2

$F_2=\left\{H_S^2\right|{\left|\right|H_S^2-{\hat{H}}_S^2\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\}$

$F_3=\left\{H_S^3\right|{\left|\right|H_S^3-{\hat{H}}_S^3\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\}$

Wherein, F₁It it is secondary user's S1 channel gain on channel 1Uncertain set, F₂And F₃It is secondary user's S2 channel gain on channel 2 and 3 respectivelyWithUncertain set, now weight coefficient assignment is It is secondary user's at secondary user's channel gain on channel kEstimated value；

3rd step, for parameterFollowing uncertain set can be obtained

$L_k=\left\{I_P^k\right|{\left|\right|I_P^k-{\hat{I}}_P^k\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\},\&ForAll;k\&Element;\{\mathrm{1,2,3}\}$

Wherein, L_kBe primary user on channel k to secondary user's produce cumulative interferenceUncertain set, Z_kBeing weight coefficient assignment is Be primary user on channel k to secondary user's produce cumulative interferenceEstimated value.

Further, in step 5, worst-case valueExpression formula be:

${\left(H_S^k\right)}^{*}=\arg \underset{(H_S^k\&Element;F_k)}{\min }\&CenterDot;(1+\frac{{\mathrm{\&Gamma;}}_k{\stackrel{\&OverBar;}{P}}_k{H}_S^k}{{\mathrm{BN}}_0+I_P^k})=\arg \underset{(H_S^k\&Element;F_k)}{\min }\&CenterDot;\left(H_S^k\right),\&ForAll;k\&Element;\left\{\mathrm{1,2,3}\right\}.$

Further, in step 6, worst-case valueExpression formula be:

${\left(I_S^k\right)}^{*}=\arg \underset{(I_S^k\&Element;L_k)}{\min }\&CenterDot;(1+\frac{{\mathrm{\&Gamma;}}_k{\stackrel{\&OverBar;}{P}}_k{H}_S^k}{{\mathrm{BN}}_0+I_P^k})=\arg \underset{(I_S^k\&Element;L_k)}{\min }\&CenterDot;\left(I_S^k\right),\&ForAll;k\&Element;\left\{\mathrm{1,2,3}\right\}.$

In cognition network provided by the invention, the method for robustness efficiency transmission, is measured the channel condition information of correlated channels, and feeds back to secondary user's transmitting terminal by secondary user's receiving terminal；Secondary user's transmitting terminal determines the uncertain set of each channel gain by repetitive measurement；The problem that secondary user's maximizes secondary user's efficiency according to robustness optimization method modeling；Solve worst channel gain and the constraint of primary user's QoS requirement is converted into convex constraint；The transmit power of optimum is solved according to power distribution algorithm, and be transmitted according to this power on each channel, the optimal transmission power of robust is solved when channel condition information uncertain, the service quality of strict guarantee primary user maximises the efficiency of secondary user's simultaneously, can be used for cognition network communication；

Present invention have the advantage that

1. in the present invention, by measuring, secondary user's determines that its channel information and primary user disturb the uncertain set of information to secondary user's, optimizes the power of self accordingly, ensure that the maximization of secondary user's efficiency in the uncertain situation of channel information；

2. in the present invention, secondary user's utilizes the uncertain set of the channel information of primary user to determine transmit power, strictly control primary user's receiving terminal jamming power, and then in the uncertain situation of channel information, strictly ensure that the service quality (QoS) of primary user；

3. the present invention is by considering the uncertainty of channel information and jamming power, adopts the power transmission method of robustness, effectively eliminates the uncertain deterioration to cognition network performance of channel information.

Accompanying drawing explanation

Fig. 1 is the method flow diagram of robustness efficiency transmission in the cognition network that the embodiment of the present invention provides；

Fig. 2 is the use cognition network downlink transfer scene graph that the embodiment of the present invention provides；

Fig. 3 is the method flowchart of robustness efficiency transmission in the cognition network that the embodiment of the present invention provides；

Fig. 4 is the sub-process figure of the power distribution algorithm that the embodiment of the present invention provides.

Detailed description of the invention

In order to make the purpose of the present invention, technical scheme and advantage clearly understand, below in conjunction with embodiment, the present invention is further elaborated.Should be appreciated that specific embodiment described herein is only in order to explain the present invention, is not intended to limit the present invention.

Below in conjunction with drawings and the specific embodiments, the application principle of the present invention is further described.

As it is shown in figure 1, the method for robustness efficiency transmission comprises the following steps in the cognition network of the embodiment of the present invention:

S101: secondary user's receiving terminal measures the channel condition information of correlated channels, and feeds back to secondary user's transmitting terminal；

S102: secondary user's transmitting terminal determines the uncertain set of each channel gain by repetitive measurement；

S103: the problem that secondary user's maximizes secondary user's efficiency according to robustness optimization method modeling；

S104: solve worst channel gain and the constraint of primary user's QoS requirement is converted into convex constraint；

S105: solve the transmit power of optimum according to power distribution algorithm, and be transmitted according to this power on each channel.

The present invention is described further in conjunction with specific embodiments:

As shown in Figure 2, the simulating scenes of the present invention comprises two pairs of primary user's transmission nodes to PT and PR i.e. P=2, primary user is numbered P1 and P2, one cognitive base station CBS and two secondary user's SR, it is numbered S1 and S2 respectively, assume that namely always have three channels takes K=3, wherein secondary user's receiving terminal S1 busy channel 1, secondary user's receiving terminal S1 busy channel 2 and 3, primary user's P1 busy channel 1 and channel 3, primary user's P2 busy channel 2(secondary user's and primary user how busy channel does not affect the use of the present invention, here the example of a concrete channel occupancy is only provided)；

As it is shown on figure 3, the method maximizing the robust transmission of efficiency in the present invention, specifically include step as follows:

Step one, secondary user's receiving terminal S1 when not receiving useful data or the interference signal that receives of the pre-test of reception, estimates the primary user transmitted the on channel 1 estimated value to its interferenceSecondary user's receiving terminal S2 estimates the primary user's estimated value to its interference in channel 2 and 3 transmissionWithWhen primary user mourns in silence, secondary user's receiving terminal S1 and S2 measures the white noise power σ on three channels²(without loss of generality, it is assumed here that on three channels, white noise power is identical), secondary user's receiving terminal S1 is according to training sequence cognitive base station CBS on channel 1 to its channel gainSecondary user's receiving terminal S2 estimates on channel 2 and 3WithThe numerical value of measurement is also sent to cognitive base station CBS by feedback channel by secondary user's receiving terminal；

Step 2, according to primary user's public information, cognitive base station CBS estimates that on three channels it arrives the channel gain between primary user receiving terminal PRWith

Step 3, before communication starts, by repeatedly repeating step (to three), it is determined that the uncertain set of relevant parameter is following (herein for convenient statement, makes ε₀=ε_j=τ_k=δ_k, { 1,2}, { 1,2, the 3} border representing unified uncertain set, generalized norm adopts l to k ∈ to j ∈ in addition₂Norm | | | |₂, for vector X ∈ R^1×n, l₂Norm | | | |₂It is represented by ${\left|\right|X\left|\right|}_2={({\left|x_1\right|}^2+...+{\left|x_n\right|}^2)}^{\frac{1}{2}}:$

The first step, for parameter G_Sj, for the receiving terminal of primary user P1, set up following uncertain set

$Q_1=\left\{G_{S1}\right|{\left|\right|{(G_{S1}-{\hat{G}}_{S1})}^T\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\}$

Wherein,Namely when primary user P1 is not transmitted on channel 2, nowMaking weight matrix value is unit matrix M₁=I_3×3∈R^3×3,It is G_S1Current estimated value；

For the receiving terminal of primary user P2, similar uncertain set can be obtained

$Q_2=\left\{G_{S2}\right|{\left|\right|{(G_{S2}-{\hat{G}}_{S2})}^T\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\}$

Wherein,Namely when primary user P1 is not transmitted on channel 1 and 3, nowMaking weight matrix value is unit matrix M₂=I_3×3∈R^3×3,It is G_S2Current estimated value；

Second step, about parameterFollowing uncertain set is set up for secondary user's S1

$F_1=\left\{H_S^1\right|{\left|\right|H_S^1-{\hat{H}}_S^1\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\}$

Following uncertain set is set up about secondary user's S2

$F_2=\left\{H_S^2\right|{\left|\right|H_S^2-{\hat{H}}_S^2\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\}$

$F_3=\left\{H_S^3\right|{\left|\right|H_S^3-{\hat{H}}_S^3\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\}$

Wherein, F₁It it is secondary user's S1 channel gain on channel 1Uncertain set, F₂And F₃It is secondary user's S2 channel gain on channel 2 and 3 respectivelyWithUncertain set, now weight coefficient assignment is It is secondary user's at secondary user's channel gain on channel kEstimated value；

3rd step, for parameterFollowing uncertain set can be obtained

$L_k=\left\{I_P^k\right|{\left|\right|I_P^k-{\hat{I}}_P^k\left|\right|}_2\&le;{\mathrm{\&epsiv;}}_0\},\&ForAll;k\&Element;\{\mathrm{1,2,3}\}$

Wherein, L_kBe primary user on channel k to secondary user's produce cumulative interferenceUncertain set, Z_kBeing weight coefficient assignment is Be primary user on channel k to secondary user's produce cumulative interferenceEstimated value；

Step 4, according to the parameter that step one to three obtains, the descending efficiency f (P) that can multiple secondary user's be transmitted over multiple channels is expressed as

$f\left(P\right)=\frac{\underset{k=1}{\overset{3}{\mathrm{\&Sigma;}}}{B\log }_2(1+\frac{{\mathrm{\&Gamma;}}_k{P}_k{H}_s^k}{{\mathrm{\&sigma;}}^2+I_P^k})}{P_C+5\underset{k=1}{\overset{3}{\mathrm{\&Sigma;}}}{p}_k}$

Wherein, B is the bandwidth of channel, when adopting quadrature amplitude modulation MQAM when error rate BER=0.001 of secondary user's, and the now coefficient of variation Γ between channel speed and the channel capacity of actual transmissions system_k=0.1258, P_CIt is the constant power expense of cognitive base station CBS, efficiency 1/ η of the power amplifier of cognitive base station CBS₀=0.20, P=[P₁,P₂,P₃] represent cognitive base station CBS transmit power vector on all channels, P here_k, { 1,2,3} represents its transmit power on channel k to k ∈；

Now, according to robustness optimization method, the problem maximizing secondary user's efficiency is modeled as

$P1:\underset{P_x}{\max }\&CenterDot;\underset{G_{\mathrm{Sj}},H_S^k,I_P^k}{\min }\&CenterDot;f\left(P\right)$

$s.t.C1:\underset{k=1}{\overset{3}{\mathrm{\&Sigma;}}}{P}_k\&le;P_{\max },$ $C2:P\&CenterDot;G_{\mathrm{Sj}}^T\&le;{\mathrm{IT}}_j,\&ForAll;j\&Element;\left\{\mathrm{1,2}\right\},$

$C3:G_{\mathrm{Sj}}\&Element;Q_j,\&ForAll;j\&Element;\left\{\mathrm{1,2}\right\},$

$C4:H_S^k\&Element;F_k,\&ForAll;k\&Element;\left\{\mathrm{1,2,3}\right\},$

$C5:I_P^k\&Element;I_k,\&ForAll;k\&Element;\left\{\mathrm{1,2,3}\right\}\&CenterDot;$

Wherein, P_maxRepresent the maximum transmit power of cognitive base station CBS, IT_jIt it is the jth primary user receiving terminal PR interference threshold value that can tolerate；

Step 5, according to following formula, solves secondary user's channel gain on channel kWorst-case value

${\left(H_S^k\right)}^{*}=\arg \underset{(H_S^k\&Element;F_k)}{\min }\&CenterDot;(1+\frac{{\mathrm{\&Gamma;}}_k{P}_k{H}_S^k}{{\mathrm{BN}}_0+I_P^k})=\arg \underset{(H_S^k\&Element;F_k)}{\min }\&CenterDot;\left(H_S^k\right),\&ForAll;k\&Element;\left\{\mathrm{1,2,3}\right\}$

Wherein, worst-case value refers at given feasible transmit power vectorMake object function in problem P1Minimum

Step 6, after this, then solves the cumulative interference that secondary user's is caused by primary user on channel kWorst-case value

${\left(I_S^k\right)}^{*}=\arg \underset{(I_S^k\&Element;L_k)}{\min }\&CenterDot;(1+\frac{{\mathrm{\&Gamma;}}_k{\stackrel{\&OverBar;}{P}}_k{H}_S^k}{{\mathrm{BN}}_0+I_P^k})=\arg \underset{(I_S^k\&Element;L_k)}{\min }\&CenterDot;\left(I_S^k\right),\&ForAll;k\&Element;\left\{\mathrm{1,2,3}\right\}$

Wherein, worst-case value refers to when a given feasible transmit power vector so that object function in problem P1MinimumValue；

Step 7, the correlation criterion according to robustness optimization, following equivalent convex constraint C6 can be converted into by P1 retraining C2 and C3:

$C6:P\&CenterDot;G_{\mathrm{Sj}}^T+{\mathrm{\&epsiv;}}_0{\left|\right|P\left|\right|}_2\&le;{\mathrm{IT}}_j,\&ForAll;j\&Element;\left\{\mathrm{1,2}\right\}$

Wherein, weight matrix M_j=I_3×3, its inverse matrixAdditionally, l₂Norm | | | |₂Dual norm be still l₂Norm | | | |₂, namelyStep 8, substitutes into the result in step 5 to seven in problem P1, can be translated into

$\begin{array}{c}P2:\underset{P_k}{\max }\&CenterDot;f_1\left(P\right)=\frac{U_R\left(P\right)}{U_{\mathrm{TP}}\left(P\right)}=\frac{\underset{k=1}{\overset{3}{\mathrm{\&Sigma;}}}{B\log }_2(1+\frac{{\mathrm{\&Gamma;}}_k{P}_k{\left(H_S^k\right)}^{*}}{{\mathrm{\&sigma;}}^2+{\left(I_P^k\right)}^{*}})}{P_C+5\underset{k=1}{\overset{3}{\mathrm{\&Sigma;}}}{P}_k}\\ s.t.C1,C6.\end{array}$

Step 9, as shown in Figure 4, adopts following power distribution algorithm, Solve problems P2:

The first step, given maximum iteration time L_max=50 and error tolerance threshold value κ=0.0001；

Second step, initializes and currently can index assignment n=0 to current iteration by valid value q=0 simultaneously；

3rd step, as convergence identifier F==0 and current iteration index n≤L_maxTime, order performs following statement；Otherwise, the 7th step is jumped to；

4th step, when maximum efficiency q gives timing, solves following convex optimization problem P3, it is thus achieved that power allocation vector P ' now；

$\begin{array}{c}P3:\underset{P}{\max }.U_R\left(P\right)-q{U}_{\mathrm{TP}}\left(P\right)\\ s.t.C1,C6.\end{array}$

5th step, if inequality U_R(P′)-qU_TP(P ') < κ sets up, then assignment P^*=P ' andConvergence mark assignment F=1；Otherwise, assignmentAnd n=n+1, convergence mark assignment F=0；

6th step, returns to the 3rd step, and continuation order performs correlative；

7th step, returns optimum energy valid value q^*With optimal power allocation vector P^*=[P₁ ^*,P₂ ^*,P₃ ^*]；

Step 10, cognitive base station CBS adopts given vector power P^*=[P₁ ^*,P₂ ^*,P₃ ^*], on channel 1 with power P₁ ^*Send data to secondary user's S1, channel 2 and 3 is respectively adopted P₂ ^*And P₃ ^*Data are sent to secondary user's S2.

By measuring, secondary user's in the present invention determines that its channel information and primary user disturb the uncertain set of information to secondary user's, optimize the power of self, ensure that the maximization of secondary user's efficiency in the uncertain situation of channel information；Secondary user's utilizes the uncertain set of the channel information of primary user to determine transmit power, strictly controls primary user's receiving terminal jamming power, it is ensured that the service quality QoS of primary user；By considering the uncertainty of channel information and jamming power, adopt the power transmission method of robustness, effectively eliminate the information uncertainty deterioration to cognition network performance.

The robustness of the present invention maximizes energy efficiency transmission method and can be widely applied in various cognition wireless network, such as, the uplink of the downlink transfer (described in application example) of cognitive base station and single secondary user's in centralized cognition wireless network, transmission point-to-multipoint between the transmission of point-to-point or secondary user's between distributed cognition network secondary user's.

The present invention considers the uncertain negative effect to cognition network performance of channel information, have the advantage that compared with existing hypothesis channel information traditional method accurately the present invention can strictly ensure that (jamming power of primary user receiving terminal PR will be strictly less than given interference threshold by cognitive base station CBS in application example for the service quality of primary user, the transmission quality of primary user is ensured) with this, and traditional method is when channel information has uncertainty, the interference of primary user is probably exceeded given interference threshold by it, thus seriously worsening the transmission quality of primary user；The present invention can promote the efficiency (in application example, the descending efficiency of cognitive base station CBS can be maximized) of secondary user's transmitting terminal, and traditional method is when channel information has uncertainty, and the efficiency of secondary user's can significantly decrease.

In sum, compared with traditional method, the present invention is when strict guarantee primary user's service quality, and the efficiency of maximizing secondary user's transmitting terminal, is the cognition wireless network effective means that realizes green communications.

The foregoing is only presently preferred embodiments of the present invention, not in order to limit the present invention, all any amendment, equivalent replacement and improvement etc. made within the spirit and principles in the present invention, should be included within protection scope of the present invention.

Claims (4)

1. the method for robustness efficiency transmission in a cognition network, it is characterised in that in this cognition network, the method for robustness efficiency transmission comprises the following steps:

Step one, secondary user's receiving terminal SR when not receiving useful data or the interference signal that receives of the pre-test of reception, estimates the primary user of the transmission estimated value to its interference on channel kWhen primary user mourns in silence, secondary user's receiving terminal SR measures the white noise power σ on channel k², when receiving data, estimate that it arrives the channel gain of secondary user's transmitting terminal ST according to training sequenceThese numerical value are also sent to secondary user's transmitting terminal ST by feedback channel by secondary user's receiving terminal SR；

Step 2, secondary user's transmitting terminal ST, when it does not have data to send, demodulates the public transmission signal of primary user, as confirmed response ACK, negative response NAK etc., estimates that it arrives the channel gain of primary user receiving terminal PR accordingly

Step 3, before communication starts, by repeatedly repeating step one to three, it is determined that the uncertain set of relevant parameter；

Step 4, according to the parameter that step one to three obtains, the descending efficiency f (P) that can multiple secondary user's be transmitted over multiple channels is expressed as:

Wherein, B is the bandwidth of channel, when adopting quadrature amplitude modulation MQAM when error rate BER=0.001 of secondary user's, and now coefficient of variation Γ k=0.1258, the P between channel speed and the channel capacity of actual transmissions system_CIt is the constant power expense of the cognitive base station CBS secondary user's transmitting terminal ST being namely mentioned above, efficiency 1/ η of the power amplifier of cognitive base station CBS₀=0.20, P=[P₁,P₂,P₃] represent cognitive base station CBS transmit power vector on all channels, P here_k, { 1,2,3} represents the transmit power on channel k to k ∈；

Now, according to robustness optimization method, the problem maximizing secondary user's efficiency is modeled as:

Wherein, P_maxRepresent the maximum transmit power of cognitive base station CBS, IT_jIt it is the jth primary user receiving terminal PR interference threshold value that can tolerate；

Step 5, first solves the worst-case value of secondary user's channel gain on channel kWherein, worst-case value refers at given feasible transmit power vectorMake object function in problem P1Minimum value；

Step 6, then solve the cumulative interference that secondary user's is caused by primary user on channel kWorst-case valueWherein, worst-case value refers at given feasible transmit power vectorMake object function in problem P1MinimumValue；

Step 7, the correlation criterion according to robustness optimization, following equivalent convex constraint C6 can be converted into by P1 retraining C2 and C3:

Wherein, weight matrix M_j=I_3×3, inverse matrix is additionally, l₂Norm | | | | the dual norm of 2 is still l₂Norm | | | | 2, namely

Step 8, substitutes into the result in step 5 to seven in problem P1, can be converted into

Step 9, adopts following power distribution algorithm, Solve problems P2:

The first step, given maximum iteration time L_max=50 and error tolerance threshold value κ=0.0001；

Second step, initializes and currently can index assignment n=0 to current iteration by valid value q=0 simultaneously；

3rd step, as convergence identifier F==0 and current iteration index n≤L_maxTime, order performs following statement；Otherwise, the 7th step is jumped to；

4th step, when maximum efficiency q gives timing, solves following convex optimization problem P3, it is thus achieved that power allocation vector P ' now；

5th step, if inequality U_R(P′)-qU_TP(P ') < κ set up, then assignment P*=P ' andConvergence mark assignment F=1；Otherwise, assignmentAnd n=n+1, convergence mark assignment F=0；

6th step, returns to the 3rd step, and continuation order performs correlative；

7th step, returns optimum energy valid value q* and optimal power allocation vector

Step 10, cognitive base station CBS adopts given vector powerOn channel 1 with power P₁* send data to secondary user's S1, channel 2 and 3 is respectively adopted P₂And P *₃* data are sent to secondary user's S2.

2. the method for robustness efficiency transmission in cognition network as claimed in claim 1, it is characterised in that in step 3, it is determined that the uncertain set of relevant parameter is as follows:

Make ε₀=ε_j=τ_k=δ_k, { 1,2}, { 1,2, the 3} border representing unified uncertain set, generalized norm adopts l to k ∈ to j ∈ in addition₂Norm | | | | 2, for vector X ∈ R^1×n, l₂Norm | | | | 2 are represented by

The first step, for parameter G_Sj, for the receiving terminal of primary user P1, set up following uncertain set

Wherein,Namely when primary user P1 is not transmitted on channel 2, nowMaking weight matrix value is unit matrix It is G_S1Current estimated value；

For the receiving terminal of primary user P2, similar uncertain set can be obtained

Wherein,Namely when primary user P1 is not transmitted on channel 1 and 3, nowMaking weight matrix value is unit matrix It is G_S2Current estimated value；

Second step, sets up following uncertain set about parameter for secondary user's S1

Following uncertain set is set up about secondary user's S2

Wherein, F₁It it is secondary user's S1 channel gain on channel 1Uncertain set, F₂And F₃It is secondary user's S2 channel gain on channel 2 and 3 respectivelyWithUncertain set, now weight coefficient assignment isIt is secondary user's at secondary user's channel gain on channel kEstimated value；

3rd step, for parameterFollowing uncertain set can be obtained

Wherein, L_kIt it is primary user's cumulative interference to secondary user's on channel kUncertain set, Z_kBeing weight coefficient assignment isIt it is primary user's cumulative interference to secondary user's on channel kEstimated value.

3. the method for robustness efficiency transmission in cognition network as claimed in claim 1, it is characterised in that in step 5, the expression formula of worst-case value is:

4. the method for robustness efficiency transmission in cognition network as claimed in claim 1, it is characterised in that in step 6, the expression formula of worst-case value is: