CN105072621A - Time-sharing cooperative communication implementation method based on OFDM cognitive network - Google Patents

Abstract

The invention discloses a time-sharing cooperative communication implementation method based on an OFDM cognitive network. The method is for the OFDM cognitive network with an energy collecting function, and belongs to the technical field of wireless communications. According to a conventional spectrum sharing method, a secondary user uses a band unoccupied by a primary user for communication, or uses a band being occupied by the primary user for communication on the premise that the communication quality of the primary user is ensured. According to the method provided by the invention, a secondary user transmitter can decode and transmit a primary user signal to promote primary user information transmission to acquire more opportunities to use a spectrum of the primary user for communication; the secondary user transmitter can collect energy from a part of received primary user signals to provide energy for relay transmission of the primary user information and the information transmission of the secondary user; the idea of dual decomposition is used for solving; optimal time and power allocation in a cooperative e communication scheme can be quickly acquired; and on the premise that the service quality of the primary user is ensured, the system capacity of the secondary user is effectively improved.

Description

A kind of implementation method of the timesharing communication for coordination based on OFDM cognition network

Technical field

The present invention relates to a kind of implementation method of the timesharing communication for coordination based on OFDM cognition network, belong to wireless communication technology field.

Background technology

The evolution of mobile communication system and develop rapidly, being mobile subscriber provides ubiquitous wireless access and high-speed wideband wireless transmission to become possibility, but day by day increases energy demand.The energy resource consumption of rapid upgrading not only increases the operating cost of communication system, has also caused serious environmental problem simultaneously.Green communications technology, especially energy collection technology more and more receive the concern of academia and industrial quarters.The wherein energy collection technology of radio signal, makes the equipment in wireless network skyborne for dispersed radio signal can be converted to electric energy and is used, extend its life cycle; And radio signal has been widely used in radio communication.Therefore, be worth for mobile terminal provides wireless data access and wireless energy to have very large Theory and applications simultaneously.

In traditional frequency spectrum sharing method, secondary user or the unappropriated band communication of use primary user, or the band communication using primary user taking under the prerequisite guaranteeing primary user's communication quality.And time user's reflector not only can decode forwarding primary user signal to promote primary user's information transmission in the method, thus the acquisition primary user's frequency spectrum that uses carries out the chance communicated more, can also from the part primary user signal received harvest energy, be the relay transmission of primary user's information and the information transmission energy supply of time user itself.By said method, the present invention utilizes the thinking of Duality Decomposition to solve, and obtains the middle time of communication for coordination scheme and the optimum allocation of power more quickly, under the prerequisite guaranteeing primary user's service quality, effectively improves the power system capacity of time user.

Summary of the invention

The object of the invention is to solve above-mentioned the deficiencies in the prior art, for the OFDM cognitive radio networks with collection of energy function, propose the implementation method of the timesharing communication for coordination of a kind of OFDM (that is: orthogonal frequency division multiplexi) cognition network, energy between the method support primary and secondary user and the dual cooperation of information are the co-allocations maximizing the resource such as time of implementation and power based on system energy efficiency and transmission performance.

Cognitive radio technology is a kind of effective way improving the availability of frequency spectrum: secondary user monitors the frequency band service condition of primary user, selects the unappropriated space of mandate frequency spectrum of primary user to communicate; Or secondary user adopts undelay pattern, when not affecting primary user's proper communication, time user is allowed to use the mandate frequency spectrum of primary user, and now, in order to not produce remarkable harmful interference to primary user's receiving terminal, the transmitting power of secondary user's transmitting terminal is often restricted.Different from said method, secondary user's transmitting terminal can forward primary user's signal as cooperative relaying decoding, and promotion primary user information transmission also obtains the chance that more use primary user frequency spectrums carry out communicating; Simultaneously time user's transmitting terminal can also from the part primary user signal received harvest energy, be the information transmission energy supply of cooperative relaying transmission and time user itself.

The present invention solves the technical scheme that its technical problem takes: a kind of implementation method of the timesharing communication for coordination based on OFDM cognition network, the method comprises the steps:

Step 1: at energy cooperation transmit stage (that is: time slot τ _{1, k}), primary user's transmitting terminal PT passes through a kth subcarrier, with power p _{p1, k}to primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _kbroadcast message x _{p1, k}.Primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _kreceived signal strength be respectively $y_k^{PU1}=\sqrt{p_{p1,k}}{h}_{p,k}{x}_{p1,k}+z_{k1}$ With $y_k^{ST1}=\sqrt{p_{p1,k}}{h}_{ps,k}{x}_{p1,k}+w_{k1}.$ In this stage, primary user's receiving terminal PU _kreceived signal strength speed be meanwhile, the signal that secondary user's transmitting terminal is received is converted to electric energy, namely

Step 2: in Cooperative For Information transmit stage, secondary user's transmitting terminal, as relaying, forwards the signal of primary user's transmitting terminal with semiduplex working method decoding.Particularly, at time slot τ _{21, k}, primary user's transmitting terminal PT passes through a kth subcarrier, with power p _{p2, k}to primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _ktransmit x _{p2, k}.Primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _kreceived signal strength be respectively $y_k^{PU21}=\sqrt{p_{p2,k}}{h}_{p,k}{x}_{p2,k}+z_{k2}$ With $y_k^{ST2}=\sqrt{p_{p2,k}}{h}_{ps,k}{x}_{p2,k}+w_{k2},$ Corresponding signal rate is respectively

$R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k})={\mathrm{\τ}}_{21,k}{\log }_2(1+\frac{p_{p2,k}{\left|h_{p,k}\right|}^2}{{\mathrm{\σ}}_k^2})$ With $R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k})={\mathrm{\τ}}_{21,k}{\log }_2(1+\frac{p_{p2,k}{\left|h_{ps,k}\right|}^2}{{\mathrm{\σ}}_k^2}).$

At ensuing time slot τ _{22, k}, secondary user's transmitting terminal ST _kby a kth subcarrier, with power p _{s1, k}to primary user's receiving terminal PU _kdecoding forwards x _{p2, k}, now primary user's transmitting terminal PT keeps idle condition.At this time slot, primary user's receiving terminal PU _kreceived signal strength be $y_k^{PU22}=\sqrt{p_{s1,k}}{h}_{sp,k}{x}_{p2,k}+z_{k3},$ Corresponding signal rate is $R_k^{PU22}({\mathrm{\τ}}_{22},p_{s1,k})={\mathrm{\τ}}_{22}{\log }_2(1+\frac{p_{s1,k}{\left|h_{sp,k}\right|}^2}{{\mathrm{\σ}}_k^2}).$ Therefore, in reached at the signal rate of this cooperation stage primary user's receiving terminal be $P_{p2,k}({\mathrm{\τ}}_{21},{\mathrm{\τ}}_{22},p_{p2,k}{p}_{s1,k})=\min \[R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k})+R_k^{PU22}({\mathrm{\τ}}_{22,k},p_{s1,k}),R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k})\].$

Step 3: at secondary user profile individual transmission stage τ ₃, that is: after ensureing that primary user's achievable rate requires, secondary user's transmitting terminal ST _kby a kth subcarrier, with power p _{s2, k}signal x is sent to secondary user's receiving terminal SU _s,k, now primary user's transmitting terminal PT keeps idle condition.In this stage, the Received signal strength of secondary user's receiving terminal SU is corresponding achievable rate is according to constraints, make time maximized optimal resource allocation strategy of user's achievable rate (τ ₃, p _{s2, k}), can obtain by solving following formula, that is:

$\underset{{\mathrm{\τ}}_k,p_k}{\max }\underset{k=1}{\overset{K}{\Σ}}{R}_{s,k}({\mathrm{\τ}}_{3,k},p_{s2,k})$

$s.t.C1:R_{p1,k}({\mathrm{\τ}}_{1,k},p_{p1,k})+R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k})\≥R_{pk},\∀k$

$C2:R_{p1,k}({\mathrm{\τ}}_{1,k},p_{p1,k})+R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k})+R_k^{PU22}({\mathrm{\τ}}_{22,k},p_{s1,k})\≥R_{pk},\∀k$

$C3:p_{s1,k}{\mathrm{\τ}}_{22,k}+p_{s2,k}{\mathrm{\τ}}_{3,k}\≤\underset{j=1}{\overset{K}{\Σ}}(p_{p1,k}{\left|h_{j,k}\right|}^2+{\mathrm{\σ}}_{j,k}){\mathrm{\τ}}_{1,j}\mathrm{\η}\mathrm{\ϵ},\∀k$

$C4:\underset{k=1}{\overset{K}{\Σ}}{p}_{p1,k}{\mathrm{\τ}}_{1,k}+\underset{k=1}{\overset{K}{\Σ}}{p}_{p2,k}{\mathrm{\τ}}_{21,k}\≤P_t$

$C5:0\≤p_{p1,k}\≤p_{max},\≤p_{p2,k}\≤p_{max},\∀k$

$C6:{\mathrm{\τ}}_{1,k}+{\mathrm{\τ}}_{21,k}+{\mathrm{\τ}}_{22,k}+\mathrm{\τ}{3}_{1,k}\≤1,\∀k$

$C7:{\mathrm{\τ}}_{1,k}={\mathrm{\τ}}_{1,1},\∀k\∈\[2,K\]$

$C8:0\≤{\mathrm{\τ}}_{i,k}\≤1,\∀i\∈\{1,21,22,3\},k$

Step 4: above-mentioned non-convex problem is changed, makes p ' _{p1, k}=p _{p1, k}τ _{1, k}, p ' _{p2, k}=p _{p2, k}τ _{21, k}, p ' _{s1, k}=p _{s1, k}τ _{22, k}with p ' _{s2, k}=p _{s2, k}τ _{3, k}.R _p1,k(τ _1,k,p _p1,k)， $R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k}),R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k}),R_k^{PU22}({\mathrm{\τ}}_{22,k},p_{s1,k}),$ R _s,k(τ _{3, k}, p _{s2, k}) can be expressed as about (τ _k, p ' _k) concave function, namely $R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k}^{\′})={\mathrm{\τ}}_{21,k}{\log }_2(1+\frac{p_{p2,k}^{\′}{\left|h_{p,k}\right|}^2}{{\mathrm{\τ}}_{21,k}{\mathrm{\σ}}_k^2}),R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k}^{\′})={\mathrm{\τ}}_{21,k}{\log }_2(1+\frac{p_{p2,k}^{\′}{\left|h_{ps,k}\right|}^2}{{\mathrm{\τ}}_{21,k}{\mathrm{\σ}}_k^2}),$ $R_k^{U22}({\mathrm{\τ}}_{22,k},p_{p2,k}^{\′})={\mathrm{\τ}}_{22,k}{\log }_2(1+\frac{p_{s1,k}^{\′}{\left|h_{ps,k}\right|}^2}{{\mathrm{\τ}}_{22,k}{\mathrm{\σ}}_k^2})$ With $R_{s,k}({\mathrm{\τ}}_{3,k},p_{s2,k}^{\′})={\mathrm{\τ}}_{3,k}{\log }_2(1+\frac{p_{s2,k}^{\′}{\left|h_{ps,k}\right|}^2}{{\mathrm{\τ}}_{3,k}{\mathrm{\σ}}_k^2}).$ Thus, non-convex problem is rewritten as convex problem.Its concrete derivation proved comprises:

By observing above-mentioned function, all there is following form, wherein (x, t) is independent variable, and α>=0 is constant.Its Hessian matrix is:

${\▿}^{2}f(x,t)=\frac{{\mathrm{\κ}}^2}{t+\mathrm{\κ}x}*\left[\begin{array}{cc}-1& \frac{x}{t}\\ \frac{x}{t}& -\frac{x^2}{t^2}\end{array}\right]$

To given arbitrary real vector z=[z ₁, z ₂], have

Namely for negative semidefinite matrix, f (x, t) is the concave function about (x, t).In like manner, R can be proved _{p1, k}(τ _{1, k}, p ' _{p1, k}), $R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k}^{\′}),R_k^{ST2}({\mathrm{\τ}}_{21,k},P_{p2,k}^{\′}),R_k^{PU22}({\mathrm{\τ}}_{22,k},p_{s1,k}^{\′})$ And R _s,k(τ _{3, k}, p ' _{s2, k}) be the convex function of corresponding independent variable.

Step 5: application dual decomposition method, introduces dual variable for each constraints in above-mentioned convex problem and Lagrangian L (s), whole problem can be disassembled into k subproblem L _k(s _k, p _k, τ _k), that is:

At internal layer, fixing dual variable s _k, to each Lagrangian L _k(s _k, p _k, τ _k) Parallel application KKT condition, obtain the optimal value of corresponding former independent variable at skin, each former independent variable is collected, optimize whole updating dual variable according to subgradient algorithm, namely for the step-length that value is less.Iterate, until make both reach optimum simultaneously.Particularly, for internal layer iteration, by former independent variable (τ _k, p ' _k) be divided into τ _kwith p ' _ktwo groups of alternative optimization.Wherein fix τ _k, obtain p ' by following various further optimization _k;

$p_{p1,k}^{\′}=\frac{{\mathrm{\τ}}_{1,k}}{{\mathrm{\α}}_{1,k}}{\[\frac{({\mathrm{\λ}}_k+{\mathrm{\μ}}_k){\mathrm{\α}}_{1,k}}{\log \left(2\right)(\mathrm{\γ}+{\mathrm{\ω}}_k-\underset{j=1}{\overset{K}{\Σ}}{\mathrm{\υ}}_k{\left|h_{j,k}\right|}^2\mathrm{\η}\mathrm{\ϵ})}-1\]}^{+},$

$p_{p2,k}^{\′}={\mathrm{\τ}}_{21,k}\frac{-b_k+\sqrt{b_k^2-4a_k{c}_k}}{2a_k},$ Wherein a _k=log (2) (γ+κ _k) α _kβ _k,

b _k＝log(2)(γ+κ _k)(α _k+β _k)-(γ+κ _k)α _kβ _k,c _k＝log(2)(γ+κ _k)-(λ _kβ _k+μ _kα _k),

$p_{s1,k}^{\′}=\frac{{\mathrm{\τ}}_{22,k}}{{\mathrm{\ψ}}_k}{\[\frac{{\mathrm{\μ}}_k{\mathrm{\ψ}}_k}{1og\left(2\right){\mathrm{\υ}}_k}-1\]}^{+},{\mathrm{\ψ}}_k=\frac{{\left|h_{sp,k}\right|}^2}{{\mathrm{\σ}}_k^2},$

Then p ' is fixed _k, obtain τ by following various further optimization _k, successively; Optimize until the two reaches optimum simultaneously, that is:

k＝2…K.

Step 4 of the present invention comprises: prove R _{p1, k}(τ _{1, k}, p ' _{p1, k}), and R _s,k(τ _{3, k}, p ' _{s2, k}) for the concrete derivation of convex function be:

By observing above-mentioned function, all there is following form,

$f(x,t)=tlog(1+\frac{\mathrm{\α}x}{t})$

Wherein (x, t) is independent variable, and α >=0 is constant, and its Hessian matrix is

${\▿}^{2}f(x,t)=\frac{{\mathrm{\κ}}^2}{t+\mathrm{\κ}x}*\left[\begin{array}{cc}-1& \frac{x}{t}\\ \frac{x}{t}& -\frac{x^2}{t^2}\end{array}\right]$

To given arbitrary real vector z=[z ₁, z ₂], have

$z^T{\▿}^{2}f(x,t)z=-\frac{{\mathrm{\κ}}^2{({\mathrm{tz}}_1-{\mathrm{xz}}_2)}^2}{t^2(t+\mathrm{\κ}x)}\≤0$

Namely for negative semidefinite matrix, f (x, t) is the concave function about (x, t); In like manner, R can be proved _{p1, k}(τ _{1, k}, p ' _{p1, k}), $R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k}^{\′}),R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k}^{\′}),R_k^{PU22}({\mathrm{\τ}}_{22,k},p_{s1,k}^{\′})$ And R _s,k(τ _{3, k}, p ' _{s2, k}) be the convex function of corresponding independent variable.

Comprise in step 5 of the present invention: for internal layer iteration, by former independent variable (τ _k, p ' _k) be divided into τ _kwith p ' _ktwo groups of alternative optimization, wherein fix τ _k, obtain p ' by following various further optimization _k, namely;

$p_{p1,k}^{\′}=\frac{{\mathrm{\τ}}_{1,k}}{{\mathrm{\α}}_{1,k}}{\[\frac{({\mathrm{\λ}}_k+{\mathrm{\μ}}_k){\mathrm{\α}}_{1,k}}{\log \left(2\right)(\mathrm{\γ}+{\mathrm{\ω}}_k-\underset{j=1}{\overset{K}{\Σ}}{\mathrm{\υ}}_k{\left|h_{j,k}\right|}^2\mathrm{\η}\mathrm{\ϵ})}-1\]}^{+},$

$p_{p2,k}^{\′}={\mathrm{\τ}}_{21,k}\frac{-b_k+\sqrt{b_k^2-4a_k{c}_k}}{2a_k},$ Wherein a _k=log (2) (γ+κ _k) α _kβ _k,

b _k＝log(2)(γ+κ _k)(α _k+β _k)-(γ+κ _k)α _kβ _k,c _k＝log(2)(γ+κ _k)-(λ _kβ _k+μ _kα _k),

$p_{s1,k}^{\′}=\frac{{\mathrm{\τ}}_{22,k}}{{\mathrm{\ψ}}_k}{\[\frac{{\mathrm{\μ}}_k{\mathrm{\ψ}}_k}{1og\left(2\right){\mathrm{\υ}}_k}-1\]}^{+},{\mathrm{\ψ}}_k=\frac{{\left|h_{sp,k}\right|}^2}{{\mathrm{\σ}}_k^2},$

Then p ' is fixed _k, obtain τ by following various further optimization _k, optimize until the two reaches optimum simultaneously successively, that is:

k＝2…K.

The present invention is applied in the cooperation transmission mode between the primary and secondary user under wireless messages and synchronous energy transmission sight, effectively improves the throughput of communication system and the efficiency of system.

Beneficial effect:

1, the present invention is by introducing time user as cooperating relay, effectively improves system transfer rate and expands area coverage, in the transmission performance improving wireless communication system, having bright prospects.

2, the present invention is directed to wireless messages and synchronous energy transmission sight, by introducing the dual cooperation transmission mode of wireless messages and energy between primary and secondary user, effectively improve the energy efficiency of wireless communication system.

Accompanying drawing explanation

Fig. 1 is the system block diagram of the OFDM cognition network in the embodiment of the present invention with collection of energy.

Fig. 2 is the schematic diagram of timesharing communication for coordination scheme between multiple primary and secondary user in OFDM cognition network.

Fig. 3 be in OFDM cognition network in primary and secondary synergic user communication scheme based on convex optimization Resourse Distribute flow chart.

Embodiment

Below in conjunction with Figure of description, the invention is described in further detail.

As shown in Figure 1, network comprises primary user's transmitting terminal PT, K primary user's receiving terminal PU, K time user's transmitting terminal ST, one customer objective receiving terminal SU.All primary and secondary users are single antenna.Wherein primary user's transmitting terminal PT and time customer objective receiving terminal SU sends and Received signal strength K sub-carrier channels respectively; And primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _kbe operated in a kth subcarrier, and secondary user's transmitting terminal ST _kcan to primary user's receiving terminal PU _kinformation is carried out decoding and is forwarded and harvest energy from acknowledge(ment) signal.Secondary user sends ST _kafter assisting primary user's transmitting terminal PT to complete communication, take each subcarrier and transmit time user's self information to secondary user's receiving terminal SU.

Step 1: at energy cooperation transmit stage (that is: time slot τ _{1, k}), primary user PT transmitting terminal passes through a kth subcarrier, with power p _{p1, k}to primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _kbroadcast message x _{p1, k}.Primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _kreceived signal strength be respectively $y_k^{PU1}=\sqrt{p_{p1,k}}{h}_{p,k}{x}_{p1,k}+z_{k1}$ With $y_k^{ST1}=\sqrt{p_{p1,k}}{h}_{ps,k}{x}_{p1,k}+w_{k1}.$ In this stage, the Received signal strength speed of primary user's receiving terminal is meanwhile, the signal that secondary user's transmitting terminal is received is converted to electric energy, namely

Step 2: in Cooperative For Information transmit stage, secondary user's transmitting terminal, as relaying, forwards the signal of primary user's transmitting terminal with semiduplex working method decoding.Particularly, at time slot τ _{21, k}, primary user's transmitting terminal passes through a kth subcarrier, with power p _{p2, k}to primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _ktransmit x _{p2, k}.Primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _kreceived signal strength be respectively $y_k^{PU21}=\sqrt{p_{p2,k}}{h}_{p,k}{x}_{p2,k}+z_{k2}$ With $y_k^{ST2}=\sqrt{p_{p2,k}}{h}_{ps,k}{x}_{p2,k}+w_{k2},$ Corresponding signal rate is respectively

$R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k})={\mathrm{\τ}}_{21,k}{\log }_2(1+\frac{p_{p2,k}{\left|h_{p,k}\right|}^2}{{\mathrm{\σ}}_k^2})$ With $R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k})={\mathrm{\τ}}_{21,k}{\log }_2(1+\frac{p_{p2,k}{\left|h_{ps,k}\right|}^2}{{\mathrm{\σ}}_k^2}).$

At ensuing time slot τ _{22, k}, secondary user's transmitting terminal ST _kby a kth subcarrier, with power p _{s1, k}to primary user's receiving terminal PU _kdecoding forwards x _{p2, k}, now primary user's transmitting terminal keeps idle condition.At this time slot, primary user's receiving terminal PU _kreceived signal strength be $y_k^{PU22}=\sqrt{p_{s1,k}}{h}_{sp,k}{x}_{p2,k}+z_{k3},$ Corresponding signal rate is $R_k^{PU22}({\mathrm{\τ}}_{22},p_{s1,k})={\mathrm{\τ}}_{22}{\log }_2(1+\frac{p_{s1,k}{\left|h_{sp,k}\right|}^2}{{\mathrm{\σ}}_k^2}).$ Therefore, in reached at the signal rate of this cooperation stage primary user's receiving terminal be $P_{p2,k}({\mathrm{\τ}}_{21},{\mathrm{\τ}}_{22},p_{p2,k}{p}_{s1,k})=\min \[R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k})+R_k^{PU22}({\mathrm{\τ}}_{22,k},p_{s1,k}),R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k})\].$

Step 3: at secondary user profile individual transmission stage τ ₃, that is: after ensureing that primary user's achievable rate requires, secondary user's transmitting terminal ST _kby a kth subcarrier, with power p _{s2, k}signal x is sent to secondary user's receiving terminal SU _s,k, now primary user's transmitting terminal keeps idle condition.In this stage, the Received signal strength of secondary user's receiving terminal SU is corresponding achievable rate is according to constraints, make time maximized optimal resource allocation strategy of user's achievable rate (τ ₃, p _{s2, k}), can obtain by solving following formula, that is:

$\underset{{\mathrm{\τ}}_k,p_k}{\max }\underset{k=1}{\overset{K}{\Σ}}{R}_{s,k}({\mathrm{\τ}}_{3,k},p_{s2,k})$

$s.t.C1:R_{p1,k}({\mathrm{\τ}}_{1,k},p_{p1,k})+R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k})\≥R_{pk},\∀k$

$C2:R_{p1,k}({\mathrm{\τ}}_{1,k},p_{p1,k})+R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k})+R_k^{PU22}({\mathrm{\τ}}_{22,k},p_{s1,k})\≥R_{pk},\∀k$

$C3:p_{s1,k}{\mathrm{\τ}}_{22,k}+p_{s2,k}{\mathrm{\τ}}_{3,k}\≤\underset{j=1}{\overset{K}{\Σ}}(p_{p1,k}{\left|h_{j,k}\right|}^2+{\mathrm{\σ}}_{j,k}){\mathrm{\τ}}_{1,j}\mathrm{\η}\mathrm{\ϵ},\∀k$

$C4:\underset{k=1}{\overset{K}{\Σ}}{p}_{p1,k}{\mathrm{\τ}}_{1,k}+\underset{k=1}{\overset{K}{\Σ}}{p}_{p2,k}{\mathrm{\τ}}_{21,k}\≤P_t$

$C5:0\≤p_{p1,k}\≤p_{max},\≤p_{p2,k}\≤p_{max},\∀k$

$C6:{\mathrm{\τ}}_{1,k}+{\mathrm{\τ}}_{21,k}+{\mathrm{\τ}}_{22,k}+\mathrm{\τ}{3}_{1,k}\≤1,\∀k$

$C7:{\mathrm{\τ}}_{1,k}={\mathrm{\τ}}_{1,1},\∀k\∈\[2,K\]$

$C8:0\≤{\mathrm{\τ}}_{i,k}\≤1,\∀i\∈\{1,21,22,3\},k$

Step 4: above-mentioned non-convex problem is changed, makes p ' _{p1, k}=p _{p1, k}τ _{1, k}, p ' _{p2, k}=p _{p2, k}τ _{21, k}, p ' _{s1, k}=p _{s1, k}τ _{22, k}with p ' _{s2, k}=p _{s2, k}τ _{3, k}.R _p1,k(τ _1,k,p _p1,k)， $R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k}),R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k}),R_k^{PU22}({\mathrm{\τ}}_{22,k},p_{s1,k}),$ R _s,k(τ _{3, k}, p _{s2, k}) can be expressed as about (τ _k, p ' _k) concave function, namely $R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k}^{\′})={\mathrm{\τ}}_{21,k}{\log }_2(1+\frac{p_{p2,k}^{\′}|h_{p,k}{|}^2}{{\mathrm{\τ}}_{21,k}{\mathrm{\σ}}_k^2}),R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k}^{\′})={\mathrm{\τ}}_{21,k}{\log }_2(1+\frac{p_{p2,k}^{\′}{\left|h_{ps,k}\right|}^2}{{\mathrm{\τ}}_{21,k}{\mathrm{\σ}}_k^2}),$ $R_k^{U22}({\mathrm{\τ}}_{22,k},p_{p2,k}^{\′})={\mathrm{\τ}}_{22,k}{\log }_2(1+\frac{p_{s1,k}^{\′}{\left|h_{ps,k}\right|}^2}{{\mathrm{\τ}}_{22,k}{\mathrm{\σ}}_k^2})$ With $R_{s,k}({\mathrm{\τ}}_{3,k},p_{s2,k}^{\′})={\mathrm{\τ}}_{3,k}{\log }_2(1+\frac{p_{s2,k}^{\′}{\left|h_{ps,k}\right|}^2}{{\mathrm{\τ}}_{3,k}{\mathrm{\σ}}_k^2}).$ Thus, non-convex problem is rewritten as convex problem.

Step 5: application dual decomposition method, introduces dual variable for each constraints in above-mentioned convex problem and Lagrangian L (s), whole problem can be disassembled into k subproblem L _k(s _k, p _k, τ _k), as shown in the formula.At internal layer, fixing dual variable s _k, to each Lagrangian L _k(s _k, p _k, τ _k) Parallel application KKT condition, obtain the optimal value of corresponding former independent variable at skin, each former independent variable is collected, optimize whole updating dual variable according to subgradient algorithm, namely for the step-length that value is less.Iterate, until make both reach optimum simultaneously.

Particularly, for internal layer iteration, by former independent variable (τ _k, p ' _k) be divided into τ _kwith p ' _ktwo groups of alternative optimization.Wherein fix τ _k, obtain p ' by following various further optimization _k;

$p_{p1,k}^{\′}=\frac{{\mathrm{\τ}}_{1,k}}{{\mathrm{\α}}_{1,k}}{\[\frac{({\mathrm{\λ}}_k+{\mathrm{\μ}}_k){\mathrm{\α}}_{1,k}}{log\left(2\right)(\mathrm{\γ}+{\mathrm{\ω}}_k-\underset{j=1}{\overset{K}{\Σ}}{\mathrm{\υ}}_k{\left|h_{j,k}\right|}^2\mathrm{\η}\mathrm{\ϵ})}-1\]}^{+},$

$p_{p2,k}^{\′}={\mathrm{\τ}}_{21,k}\frac{-b_k+\sqrt{b_k^2-4a_k{c}_k}}{2a_k},$ Wherein a _k=log (2) (γ+κ _k) α _kβ _k,

b _k＝log(2)(γ+κ _k)(α _k+β _k)-(γ+κ _k)α _kβ _k,c _k＝log(2)(γ+κ _k)-(λ _kβ _k+μ _kα _k),

$p_{s1,k}^{\′}=\frac{{\mathrm{\τ}}_{22,k}}{{\mathrm{\ψ}}_k}{\[\frac{{\mathrm{\μ}}_k{\mathrm{\ψ}}_k}{1og\left(2\right){\mathrm{\υ}}_k}-1\]}^{+},{\mathrm{\ψ}}_k=\frac{|h_{sp,k}{|}^2}{{\mathrm{\σ}}_k^2},$

Then p ' is fixed _k, obtain τ by following various further optimization _k, successively; Optimize until the two reaches optimum simultaneously.

k＝2…K.

Finally, former independent variable (τ, p ') and dual variable s reach optimum simultaneously, and calculate reaching and speed of time user thus.

Further, in step 4 of the present invention, R is proved _{p1, k}(τ _{1, k}, p ' _{p1, k}), and R _s,k(τ _{3, k}, p ' _{s2, k}) for the concrete derivation of convex function be:

By observing above-mentioned function, all there is following form,

$f(x,t)=tlog(1+\frac{\mathrm{\α}x}{t})$

Wherein (x, t) is independent variable, and α >=0 is constant, and its Hessian matrix is

${\▿}^{2}f(x,t)=\frac{{\mathrm{\κ}}^2}{t+\mathrm{\κ}x}*\left[\begin{array}{cc}-1& \frac{x}{t}\\ \frac{x}{t}& -\frac{x^2}{t^2}\end{array}\right]$

To given arbitrary real vector z=[z ₁, z ₂], have

$z^T{\▿}^{2}f(x,t)z=-\frac{{\mathrm{\κ}}^2{({\mathrm{tz}}_1-{\mathrm{xz}}_2)}^2}{t^2(t+\mathrm{\κ}x)}\≤0$

Namely for negative semidefinite matrix, f (x, t) is the concave function about (x, t); In like manner, R can be proved _{p1, k}(τ _{1, k}, p ' _{p1, k}), $R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k}^{\′}),R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k}^{\′}),R_k^{PU22}({\mathrm{\τ}}_{22,k},p_{s1,k}^{\′})$ And R _s,k(τ _{3, k}, p ' _{s2, k}) be the convex function of corresponding independent variable.

Further, in step 5 of the present invention, for internal layer iteration, by former independent variable (τ _k, p ' _k) be divided into τ _kwith p ' _ktwo groups of alternative optimization, wherein fix τ _k, obtain p ' by following various further optimization _k, namely;

$p_{p1,k}^{\′}=\frac{{\mathrm{\τ}}_{1,k}}{{\mathrm{\α}}_{1,k}}{\[\frac{({\mathrm{\λ}}_k+{\mathrm{\μ}}_k){\mathrm{\α}}_{1,k}}{log\left(2\right)(\mathrm{\γ}+{\mathrm{\ω}}_k-\underset{j=1}{\overset{K}{\Σ}}{\mathrm{\υ}}_k{\left|h_{j,k}\right|}^2\mathrm{\η}\mathrm{\ϵ})}-1\]}^{+},$

$p_{p2,k}^{\′}={\mathrm{\τ}}_{21,k}\frac{-b_k+\sqrt{b_k^2-4a_k{c}_k}}{2a_k},$ Wherein a _k=log (2) (γ+κ _k) α _kβ _k,

b _k＝log(2)(γ+κ _k)(α _k+β _k)-(γ+κ _k)α _kβ _k,c _k＝log(2)(γ+κ _k)-(λ _kβ _k+μ _kα _k),

$p_{s1,k}^{\′}=\frac{{\mathrm{\τ}}_{22,k}}{{\mathrm{\ψ}}_k}{\[\frac{{\mathrm{\μ}}_k{\mathrm{\ψ}}_k}{\log \left(2\right){\mathrm{\υ}}_k}-1\]}^{+},{\mathrm{\ψ}}_k=\frac{{\left|h_{sp,k}\right|}^2}{{\mathrm{\σ}}_k^2},$

Then p ' is fixed _k, obtain τ by following various further optimization _k, optimize until the two reaches optimum simultaneously successively;

k＝2…K.

Embodiments described herein are only preferred implementation; and be not the restriction of scope; any based on the present invention's improvement of doing of spirit or equivalently to replace, only otherwise depart from the spirit and scope of the present invention, all should be encompassed within scope.

Claims (4)

1., based on an implementation method for the timesharing communication for coordination of OFDM cognition network, it is characterized in that, said method comprising the steps of:

Step 1: at energy cooperation transmit stage, that is: time slot τ _{1, k}, primary user's transmitting terminal PT passes through a kth subcarrier, with power p _{p1, k}to primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _kbroadcast message x _{p1, k}; Primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _kreceived signal strength be respectively $y_k^{PU1}=\sqrt{p_{p1,k}}{h}_{p,k}{x}_{p1,k}+z_{k1}$ With $y_k^{ST1}=\sqrt{p_{p1,k}}{h}_{ps,k}{x}_{p1,k}+w_{k1};$ In this stage, the Received signal strength speed of primary user's receiving terminal is meanwhile, the signal that secondary user's transmitting terminal is received is converted to electric energy, namely

Step 2: in Cooperative For Information transmit stage, secondary user's transmitting terminal, as relaying, forwards the signal of primary user's transmitting terminal with semiduplex working method decoding; At time slot τ _{21, k}, primary user's transmitting terminal PT passes through a kth subcarrier, with power p _{p2, k}to primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _ktransmit x _{p2, k}; Primary user's receiving terminal PU _kwith secondary user's transmitting terminal ST _kreceived signal strength be respectively $y_k^{PU21}=\sqrt{p_{p2,k}}{h}_{p,k}{x}_{p2,k}+z_{k2}$ With $y_k^{ST2}=\sqrt{p_{p2,k}}{h}_{ps,k}{x}_{p2,k}+w_{k2},$ Corresponding signal rate is respectively:

$R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k})={\mathrm{\τ}}_{21,k}{\log }_2(1+\frac{p_{p2,k}{\left|h_{p,k}\right|}^2}{{\mathrm{\σ}}_k^2})$ With $R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k})={\mathrm{\τ}}_{21,k}{\log }_2(1+\frac{p_{p2,k}{\left|h_{ps,k}\right|}^2}{{\mathrm{\σ}}_k^2});$

At ensuing time slot τ _{22, k}, secondary user's transmitting terminal ST _kby a kth subcarrier, with power p _{s1, k}to primary user's receiving terminal PU _kdecoding forwards x _{p2, k}, now primary user's transmitting terminal PT keeps idle condition, at this time slot, and primary user's receiving terminal PU _kreceived signal strength be $y_k^{PU22}=\sqrt{p_{s1,k}}{h}_{sp,k}{x}_{p2,k}+z_{k3},$ Corresponding signal rate is $R_k^{PU22}({\mathrm{\τ}}_{22},p_{s1,k})={\mathrm{\τ}}_{22}{\log }_2(1+\frac{p_{s1,k}{\left|h_{sp,k}\right|}^2}{{\mathrm{\σ}}_k^2});$ In this cooperation stage, reached at the signal rate of primary user's receiving terminal is $P_{p2,k}({\mathrm{\τ}}_{21},{\mathrm{\τ}}_{22},p_{p2,k}{p}_{s1,k})=\min \[R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k})+R_k^{PU22}({\mathrm{\τ}}_{22,k},p_{s1,k}),R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k})\];$

Step 3: at secondary user profile individual transmission stage τ _{3, k}, that is: after ensureing that primary user's achievable rate requires, secondary user's transmitting terminal ST _kby a kth subcarrier, with power p _{s2, k}signal x is sent to secondary user's receiving terminal SU _s,k, now primary user's transmitting terminal PT keeps idle condition; In this stage, the Received signal strength of secondary user's receiving terminal SU is corresponding achievable rate is according to constraints, make time maximized optimal resource allocation strategy of user's achievable rate (τ ₃, p _{s2, k}), obtain by solving following formula:

$\underset{{\mathrm{\τ}}_k,p_k}{max}\underset{k=1}{\overset{K}{\Σ}}{R}_{s,k}({\mathrm{\τ}}_{3,k},p_{s2,k})$

$s.t.C1:R_{p1,k}({\mathrm{\τ}}_{1,k},p_{p1,k})+R_k^{ST2}({\mathrm{\τ}}_{21,k},p_{p2,k})\≥R_{pk},\∀k$

$C2:R_{p1,k}({\mathrm{\τ}}_{1,k},p_{p1,k})+R_k^{PU21}({\mathrm{\τ}}_{21,k},p_{p2,k})+R_k^{PU22}({\mathrm{\τ}}_{22,k},p_{s1,k})\≥R_{pk},\∀k$

$C3:p_{s1,k}{\mathrm{\τ}}_{22,k}+p_{s2,k}{\mathrm{\τ}}_{3,k}\≤\underset{j=1}{\overset{K}{\Σ}}(p_{p1,k}{\left|h_{j,k}\right|}^2+{\mathrm{\σ}}_{j,k}){\mathrm{\τ}}_{1,j}\mathrm{\η}\mathrm{\ϵ},\∀k$

$C4:\underset{k=1}{\overset{K}{\Σ}}{p}_{p1,k}{\mathrm{\τ}}_{1,k}+\underset{k=1}{\overset{K}{\Σ}}{p}_{p2,k}{\mathrm{\τ}}_{21,k}\≤P_t$

$C5:0\≤p_{p1,k}\≤p_{max},\≤p_{p2,k}\≤p_{max},\∀k$

$C6:{\mathrm{\τ}}_{1,k}+{\mathrm{\τ}}_{21,k}+{\mathrm{\τ}}_{22,k}+\mathrm{\τ}{3}_{1,k}\≤1,\∀k$

$C7:{\mathrm{\τ}}_{1,k}={\mathrm{\τ}}_{1,1},\∀k\∈\[2,K\]$

$C8:0\≤{\mathrm{\τ}}_{i,k}\≤1,\∀i\∈\{1,21,22,3\},k$

Step 4: above-mentioned non-convex problem is changed, makes p ' _{p1, k}=p _{p1, k}τ _{1, k}, p ' _{p2, k}=p _{p2, k}τ _{21, k}, p ' _{s1, k}=p _{s1, k}τ _{22, k}with p ' _{s2, k}=p _{s2, k}τ _{3, k}, R _{p1, k}(τ _{1, k}, p _{p1, k}), r _s,k(τ _{3, k}, p _{s2, k}) be expressed as about (τ _k, p ' _k) concave function, that is: $R_{p1,k}({\mathrm{\τ}}_{1,k},p_{p1,k}^{\′})={\mathrm{\τ}}_{1,k}{\log }_2(1+\frac{p_{p1,k}^{\′}{\left|h_{p,k}\right|}^2}{{\mathrm{\τ}}_{1,k}{\mathrm{\σ}}_k^2}),$ $R_k^{PU21}\left({\mathrm{\τ}}_{21,k},p_{p2,k}^{\′}\right)={\mathrm{\τ}}_{21,k}{\log }_2\left(1+\frac{p_{p2,k}^{\′}{\left|h_{p,k}\right|}^2}{{\mathrm{\τ}}_{21,k}{\mathrm{\σ}}_k^2}\right),R_k^{ST2}\left({\mathrm{\τ}}_{21,k},p_{p2,k}^{\′}\right)={\mathrm{\τ}}_{21,k}{\log }_2\left(1+\frac{p_{p2,k}^{\′}{\left|h_{ps,k}\right|}^2}{{\mathrm{\τ}}_{21,k}{\mathrm{\σ}}_k^2}\right),$ $R_k^{U22}({\mathrm{\τ}}_{22,k},p_{p2,k}^{\′})={\mathrm{\τ}}_{22,k}{\log }_2(1+\frac{p_{s1,k}^{\′}{\left|h_{ps,k}\right|}^2}{{\mathrm{\τ}}_{22,k}{\mathrm{\σ}}_k^2})$ With $R_{s,k}({\mathrm{\τ}}_{3,k},p_{s2,k}^{\′})={\mathrm{\τ}}_{3,k}{\log }_2(1+\frac{p_{s2,k}^{\′}{\left|h_{ps,k}\right|}^2}{{\mathrm{\τ}}_{3,k}{\mathrm{\σ}}_k^2}),$ Thus, non-convex problem is rewritten as convex problem;

Step 5: application dual decomposition method, introduces dual variable for each constraints in above-mentioned convex problem and Lagrangian L (s), whole problem is disassembled into k subproblem L _k(s _k, p _k, τ _k), that is:

At internal layer, fixing dual variable s _k, to each Lagrangian L _k(s _k, p _k, τ _k) Parallel application KKT condition, obtain the optimal value of corresponding former independent variable at skin, each former independent variable is collected, optimize whole updating dual variable according to subgradient algorithm, namely for the step-length that value is less, iterate, until make both reach optimum simultaneously.

2. the implementation method of a kind of timesharing communication for coordination based on OFDM cognition network according to claim 1, it is characterized in that, described step 4 comprises: prove R _{p1, k}(τ _{1, k}, p ' _{p1, k}), and R _s,k(τ _{3, k}, p ' _{s2, k}) for the concrete derivation of convex function be:

By observing above-mentioned function, all there is following form,

$f(x,t)=tlog(1+\frac{\mathrm{\α}x}{t})$

Wherein (x, t) is independent variable, and α >=0 is constant, and its Hessian matrix is

${\▿}^{2}f(x,t)=\frac{{\mathrm{\κ}}^2}{t+\mathrm{\κ}x}*\left[\begin{array}{cc}-1& \frac{x}{t}\\ \frac{x}{t}& -\frac{x^2}{t^2}\end{array}\right]$

To given arbitrary real vector z=[z ₁, z ₂], have

$z^T{\▿}^{2}f(x,t)z=-\frac{{\mathrm{\κ}}^2{({\mathrm{tz}}_1-{\mathrm{xz}}_2)}^2}{t^2(t+\mathrm{\κ}x)}\≤0$

I.e. ▽ ²f (x, t) is negative semidefinite matrix, and f (x, t) is the concave function about (x, t); In like manner, R can be proved _{p1, k}(τ _{1, k}, p ' _{p1, k}), with for the convex function of corresponding independent variable.

3. the implementation method of a kind of timesharing communication for coordination based on OFDM cognition network according to claim 1, it is characterized in that, described step 5 comprises: for internal layer iteration, by former independent variable (τ _k, p ' _k) be divided into τ _kwith p ' _ktwo groups of alternative optimization, wherein fix τ _k, obtain p ' by following various further optimization _k, namely;

$p_{p1,k}^{\′}=\frac{{\mathrm{\τ}}_{1,k}}{{\mathrm{\α}}_{1,k}}{\[\frac{({\mathrm{\λ}}_k+{\mathrm{\μ}}_k){\mathrm{\α}}_{1,k}}{log\left(2\right)(\mathrm{\γ}+{\mathrm{\ω}}_k-\underset{j=1}{\overset{K}{\Σ}}{\mathrm{\υ}}_k{\left|h_{j,k}\right|}^2\mathrm{\η}\mathrm{\ϵ})}-1\]}^{+},$

$p_{p2,k}^{\′}={\mathrm{\τ}}_{21,k}\frac{-b_k+\sqrt{b_k^2-4a_k{c}_k}}{2a_k},$ Wherein a _k=log (2) (γ+κ _k) α _kβ _k,

b _k＝log(2)(γ+κ _k)(α _k+β _k)-(γ+κ _k)α _kβ _k,c _k＝log(2)(γ+κ _k)-(λ _kβ _k+μ _kα _k),

$p_{s1,k}^{\′}=\frac{{\mathrm{\τ}}_{22,k}}{{\mathrm{\ψ}}_k}{\[\frac{{\mathrm{\μ}}_k{\mathrm{\ψ}}_k}{\log \left(2\right){\mathrm{\υ}}_k}-1\]}^{+},{\mathrm{\ψ}}_k=\frac{{\left|h_{sp,k}\right|}^2}{{\mathrm{\σ}}_k^2},$

Then p ' is fixed _k, obtain τ by following various further optimization _k, optimize until the two reaches optimum simultaneously successively, that is:

k＝2…K.

4. the implementation method of a kind of timesharing communication for coordination based on OFDM cognition network according to claim 1, is characterized in that: described method is applied in the cooperation transmission mode between the primary and secondary user under wireless messages and synchronous energy transmission sight.