CN106304239A - The relay selection method of energy acquisition multi-relay cooperation communication system - Google Patents

Abstract

The invention discloses the relay selection method of a kind of energy acquisition multi-relay cooperation communication system, belong to cooperative communication technology field.Including step: system scenarios analysis, problem is summed up；System mathematic model is set up；Then optimization method is utilized to obtain optimal solution.The present invention is directed to special application scenarios, source reality application, difference and conventional many relay selection method, consider multi-relay cooperation communication based on energy acquisition, and derive the expression formula of handling capacity under this scene, maximize the throughput performance between communication node, carry out solving of optimization problem, there is the directive significance of reality.The present invention is directed to solving of optimization problem, convex optimization is used to process, convert the object function of optimization problem, without approximate calculation, do not affect the computation complexity reduced greatly while the precision of problem, reduce the time delay that overhead produces, searching process uses Lagrange multiplier method, and speed of searching optimization is fast, uses subgradient method during algorithm iteration, and select progressive step-length, optimizing is more accurate.

Description

The relay selection method of energy acquisition multi-relay cooperation communication system

Technical field

The invention belongs to cooperative communication technology field, more particularly, it relates to a kind of energy acquisition multi-relay cooperation communication The relay selection method of system.

Background technology

1979, Cover and Gamal, in the research to trunk channel, refer to collaboration communication first The model of (Cooperative communication).They have studied in the AWFN of active, relaying and purpose three node composition The channel capacity of relay communication network network model, all nodes therein are operated in identical frequency range, thus system are divided into one Individual broadcast channel and an access channel.To the research of collaboration communication the most based on this most then.Collaboration communication is with direct Communication is compared and is provided that space diversity gain, realizes the data transmission of targeted customer's high speed, high reliability.In collaboration communication one Individual distribution and the supervisory relay node of key challenge is how, single relay selection algorithm, although systematic function can be improved, but Due to only with a via node, it is impossible to obtain " emerging in large numbers " gain of multiple relay node cooperation communication.Although choosing more Via node can obtain better performance, but be as the increase of via node number, the performance gain brought will reduce, The complexity of the design of the scheme that simultaneously cooperates, signal detection and multiple access problem etc. can increase, and network cost also can improve.

End is got up, and cooperative communication technology will become one of key technology of the 5th third-generation mobile communication, it is possible to greatly improve logical The performance of communication network, but also drawn the problem that several letter is to be solved simultaneously:

1) the 5th third-generation mobile communication is more strict for the requirement of energy consumption, if the utilization of regenerative resource can be taken into full account, Wish to maximize the handling capacity between communication node under the conditions of energy causality constraint；

2) in order to reach more preferable network performance, need to give full play to the function of relay station, select optimal relaying, utilize Good idling-resource, maximization network performance；

3) in the urgent need to the algorithm about the distribution of radio sensing network resource of actual application can be put into, algorithm is emphasized Low complex degree, real-time operation ability and high convergence rate.

Chinese Patent Application No. 201510282859.0, publication date on May 28th, 2015, discloses a entitled one Cooperative relaying system of selection, it comprises the following steps: utilize purpose user and the geographic distance of mobile subscriber and described purpose The comprehensive distance metric with described mobile subscriber of the purpose user described in probability calculation that cooperates of user and described mobile subscriber；Meter Calculate the multiple described comprehensive distance metric of described purpose user, and select the institute that maximum described comprehensive distance metric is corresponding State mobile subscriber's via node as cooperation relation to be set up.The method can improve the probability of success of relay selection, in reduction Continue the expense selected.

Chinese Patent Application No. 201510164751.1, publication date on April 8th, 2015, discloses a entitled one Fair cooperating relay system of selection, comprising: first originate according to the power supply of via node each in cooperation communication system, will association It is divided into three kinds of situations as communication system, cooperation communication system i.e. only exists battery powered via node, cooperation communication system In only exist and the via node or cooperation communication system that electrical network powers exist battery and via node that electrical network is powered simultaneously；So Select according to different situations afterwards, use different formula to select via node that data are carried out relay forwarding.The method Take into account signal quality and via node load, while relay selection fairness is effectively ensured, promote the overall performance of system.

Chinese Patent Application No. 201510227055.0, publication date on May 6th, 2015, discloses a entitled one MIMO cooperation relay selection method, it comprises the steps: with mimo channel capacity, node normalized emission power, receives sky Line Signal-to-noise ratio balance degree and mimo channel degree of association are the input of fuzzy algorithmic approach；Employing two-stage fuzzy criterion, the first order obscures Criterion is for describing mimo channel capacity, relation between node normalized emission power and MIMO relay selection, second level mould Stick with paste criterion for describing the relation between reception antenna Signal-to-noise ratio balance degree, mimo channel degree of association and MIMO relay selection；Will The MIMO relay selection result exported by two-stage fuzzy criterion respectively is merged, so that it is determined that required MIMO relaying joint Point.The method further increases the performance of MIMO cooperation communication system, has considerable flexibility and low implementation complexity.

Generally speaking, the open material of application number 201510282859.0 considers relay selection based on ideal distance, fall Low overhead, but do not set about from user perspective, do not account for the situation that user throughput is optimum.Application number The open material of 201510164751.1 considers relay selection based on multi-user's fairness, but do not account for single user how in The selection situation continued.The open material of application number 201510227055.0 considers the relay selection under MIMO sight, but does not has Consider algorithm complex and the requirement of real-time operation.

Summary of the invention

Do not take into full account that the performance that energy acquisition factor is brought changes for existing cooperation communication system relay selection method The problems such as joint relay selection under kind, energy cause and effect restrictive condition, the actual application of low complexity algorithm, the present invention proposes one The relay selection method of energy acquisition multi-relay cooperation communication system, in the associating under considering energy cause and effect restrictive condition Continue selection, the energy-saving scheme brought in conjunction with energy acquisition technology, assists low complex degree iterative algorithm, maximizes net between communication user Network performance.

For solving the problems referred to above, the technical solution adopted in the present invention is as follows:

A kind of relay selection method of energy acquisition multi-relay cooperation communication system, including:

Step 1: system scenarios analysis, problem is summed up；

Step 1.1: channel model is set up；

Scene has signal source S, the relay station R of M energy acquisition_i, i=1,2 ..., M and a destinations traffic Terminal D, does not has direct path, relay station R between signal source S and target communications terminal D_iSelection amplification forwarding working method, one Individual transmitting procedure includes N number of data block, and the transmission time of a data block is T^C, signal source S and relaying R_iTake equal bandwidth W, it is considered to the select permeability of the relay station of M energy acquisition under this scene, it is assumed that enough for the battery capacity of energy acquisition Greatly, ignore except being used for transmitting the energy of consumption, relay R for i-th_i, transmit the channel gain during data block of jth It is defined asWhereinWhen representing transmission jth data block, signal source S is to relay station R_iChannel coefficients,Represent and forward Relay station R during jth data block_iTo the channel coefficients of target communications terminal D, if having selected when transmitting jth packet I-th relays, and SNR can be defined as end to end:

${\mathrm{\Λ}}_{i,j}=\frac{{\mathrm{\γ}}_{i,j}^s{\mathrm{\γ}}_{i,j}^d}{{\mathrm{\γ}}_{i,j}^s+{\mathrm{\γ}}_{i,j}^d+1},$

WhereinRepresent signal source S and the noise of target communications terminal D respectively Ratio, P^sAnd P^trRepresent signal source S and the transmitting power of relay station Relay, N respectively₀Represent normalized noise power, and then obtain Handling capacity the most end to end:

R_i,j=log₂(1+Λ_i,j)；

Step 1.2: energy acquisition model is set up；

It is defined as the gross energy that i-th relaying adds up to collect by the end of moment t,Represent that i-th relaying exists The energy work rate that jth packet transmission cycle collects, T^ERepresent the time interval of energy acquisition, N^CRepresent that an energy is adopted The time interval of how many packets, N is comprised in collection interval^E=N/N^CRepresent the number of times that in total transmitting procedure, energy arrives, whole Transmission cycle can be expressed as T=N^ET^E, instantaneous through-put power is defined as P (t), and energy cause effect relation restricted representation is:

$\underset{0}{\overset{t}{\∫}}P\left(\mathrm{\τ}\right)d\mathrm{\τ}\≤E_{i,\mathrm{\Σ}}^{EH}\left(t\right);$

Step 2: system mathematic model is set up；

When jth data block is transmitted, if we choose i-th, relaying participates in cooperation, x_i,j=1, otherwise x_i,j= 0, under known energy and channel status, optimization problem can be grouped into:

$\begin{array}{cc}P1:& \underset{x_{i,j}}{\max }\stackrel{\‾}{R}=\frac{1}{N}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{R}_{i,j}{x}_{i,j}\\ s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\mathrm{\Σ}}}{x}_{i,j}\≤E_{i,\mathrm{\Σ}}^{EH}\left({\mathrm{lT}}^c\right),\∀l\∈J,i\∈I,\\ & \underset{i=1}{\overset{M}{\mathrm{\Σ}}}{x}_{i,j}\≤1,\∀j\∈J,\\ & x_{i,j}\∈\{0,1\},J=\{1,...,N\},I=\{1,...,M\}.\end{array};$

Step 3: the convex optimization of optimization problem processes；

For any one transmission bag j, if the given corresponding relay selection result of k≤j≤N, the most accurately or The status information of channel, then handling capacity expression formula isIt was accordingly found that the relay selection factor x_i,jIt is crucial, then writes out the average throughput expression formula for jth packet WhereinRepresent the average throughput of i-th relaying；

Convert object function further as follows:

$\stackrel{\‾}{R}\left(k\right)=\frac{1}{N-k+1}(\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{R}_{i,k}{x}_{i,k}+\underset{j=k+1}{\overset{N}{\mathrm{\Σ}}}{\stackrel{\‾}{R}}_{ave,j})=\frac{1}{N-k+1}(\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{R}_{i,k}{x}_{i,k}+\underset{i=0}{\overset{M}{\mathrm{\Σ}}}{\stackrel{\‾}{R}}_i{r}_i)$

WhereinRepresenting and start from kth packet until the end of transmission, i-th relays quilt The average time chosen, thus optimization problem P1 can be converted into P2:

$\begin{array}{cc}P2:& \underset{x_{i,j}}{\max }\stackrel{\‾}{R}\left(k\right)=\frac{1}{N-k+1}(\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{R}_{i,k}{x}_{i,k}+\underset{i=0}{\overset{M}{\mathrm{\Σ}}}{\stackrel{\‾}{R}}_i{r}_i)\\ s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\mathrm{\Σ}}}{x}_{i,j}\≤E_{i,\mathrm{\Σ}}^{EH}\left({\mathrm{lT}}^c\right),\∀l\∈J,i\∈I,\\ & \underset{i=1}{\overset{M}{\mathrm{\Σ}}}{x}_{i,j}\≤1,\∀j\∈J,\\ & x_{i,j}\∈\{0,1\},J=\{1,...,N\},I=\{1,...,M\}.\end{array};$

Step 4: solve the average time r that i-th relaying is selected_i；

In the case of knowing ESI energy acquisition curve, based on the EH curve for all relayings, can obtain every respectively Individual relaying in a continuous print transmission cycle selected average time

Step 5: utilize relative throughput gain to carry out relay selection；

DefinitionWhereinRepresent relaying i₁Selected is flat All handling capacities,Represent relaying i₂Selected average throughput, Δ R_i,kRepresent the relative throughput gain of relaying i, thus obtain The selection strategy collection that must relay is as follows:

$K_k=\arg \underset{i}{max}\left\{{\mathrm{\ΔR}}_{i,k}\left|\begin{array}{c}\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\Σ}}{x}_{i,j}\≤E_{i,\mathrm{\Σ}}^{EH}\left({\mathrm{lT}}^c\right),\∀l\∈J,i\∈I,\\ \underset{i=1}{\overset{M}{\Σ}}{x}_{i.j}\≤1,\∀j\∈J,\\ x_{i,j}\∈\{0,1\},J=\{1,...,N\},I=\{1,...,M\}.\end{array}\right.\right\}$

And thus make optimal middle rank selection.

Further, described step 2 includes:

If in the case of can not obtaining the whole state of channel completely, optimization object function is rewritten into:

$\stackrel{\‾}{R}\left(k\right)=E\[\frac{1}{N-k+1}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=k}{\overset{N}{\mathrm{\Σ}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\]=\frac{\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\∀i,j}{{\mathrm{\γ}}_{i,j}^s,{\mathrm{\γ}}_{i,j}^d}}\[\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=k}{\overset{N}{\mathrm{\Σ}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\]}{N-k+1}$

Thus optimization problem P1 can be converted into P3:

$\begin{array}{cc}P3:& \underset{x_{i,j}}{\max }\stackrel{\‾}{R}\left(k\right)\end{array}=\frac{\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\∀i,j}{{\mathrm{\γ}}_{i,j}^s,{\mathrm{\γ}}_{i,j}^d}}\[\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=k+1}{\overset{N}{\mathrm{\Σ}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\]}{N-k+1}$

$\begin{array}{cc}s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\Σ}}{x}_{i,j}\≤E_{i,\mathrm{\Σ}}^{EH}\left({\mathrm{lT}}^c\right),\∀l\∈J,i\∈I\end{array},$

$\underset{i=1}{\overset{M}{\Σ}}{x}_{i,j}\≤1,\∀j\∈J,$

x_i,j∈ 0,1}, J={1 ..., N}, I={1 ..., M}.

WhereinRepresent the handling capacity of the whole system of prediction on the premise of known front k is according to bag status information；E [·|x_i,k] represent the expectation after the relaying giving selection i-th for kth packet when transmit,Expression can not be complete The expectation that the signal to noise ratio statistical property of channel is made is utilized on the premise of obtaining the whole state of channel.

Further, described step 2 can also include:

If channel status can not accurately be obtained, the state of energy acquisition can not be known in time and accurately, optimization problem It is rewritten into:

$\stackrel{\‾}{R}\left(k\right)=E\[\frac{1}{N-k+1}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=k}{\overset{N}{\mathrm{\Σ}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\]=\frac{\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\∀i,j}{{\mathrm{\γ}}_{i,j}^s,{\mathrm{\γ}}_{i,j}^d,P_{i,j}^{EH}}}\[\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=k+1}{\overset{N}{\mathrm{\Σ}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\]}{N-k+1}$

Thus optimization problem P1 can be converted into P4:

$\begin{array}{cc}P4:& \underset{x_{i,j}}{\max }\stackrel{\‾}{R}\left(k\right)\end{array}=\frac{\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\∀i,j}{{\mathrm{\γ}}_{i,j}^s,{\mathrm{\γ}}_{i,j}^d,P_{i,j}^{EH}}}\[\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=k+1}{\overset{N}{\mathrm{\Σ}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\]}{N-k+1}$

$\begin{array}{cc}s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\Σ}}{x}_{i,j}\≤E_{i,\mathrm{\Σ}}^{EH}\left({\mathrm{lT}}^c\right),\∀l\∈J,i\∈I\end{array},$

$\underset{i=1}{\overset{M}{\Σ}}{x}_{i,j}\≤1,\∀j\∈J,$

x_i,j∈ 0,1}, J={1 ..., N}, I={1 ..., M}.

WhereinRepresent the handling capacity of the whole system of prediction, E on the premise of known front k is according to bag status information [·|x_i,k] represent the expectation after the relaying giving selection i-th for kth packet when transmit,Expression can not be complete The expectation utilizing the signal to noise ratio statistical property of channel to make on the premise of the full acquisition whole state of channel and energy acquisition state；

Described step 4 includes: update the average time r that i-th relaying is selected_iAs follows

Further, the Lagrangian Form of described step 2 optimization problem P3 is:

$\begin{array}{c}L(x_{i,j}{\mathrm{\μ}}_{ml},{\mathrm{\λ}}_l)=\frac{\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\∀i,j}{{\mathrm{\γ}}_{i,j}^s,{\mathrm{\γ}}_{i,j}^d}}\[\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=k+1}{\overset{N}{\mathrm{\Σ}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\]}{N-k+1}\\ -{\mathrm{\μ}}_{ml}(\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\Σ}}{x}_{i,j}-E_{i,\mathrm{\Σ}}^{EH}({\mathrm{lT}}^c))-{\mathrm{\λ}}_l(\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{x}_{i,j}-1)\end{array}$

Simultaneous againAnd with subgradient method iterative, wherein, l ∈ 1 ..., N}, m ∈ 1 ..., M}, μ_ml, λ_lRepresent Lagrange factor.

Further, the Lagrange factor μ in the Lagrangian Form of described optimization problem P3_ml,λ_lIteration update Method uses Subgradient Algorithm, and the iteration renewal equation of described Subgradient Algorithm is:

${\mathrm{\μ}}_{ml}(n+1)={\[{\mathrm{\μ}}_{ml}\left(n\right)-{\mathrm{\α}}_{{\mathrm{\μ}}_{ml}}\left(n\right)\left(E_{i,\mathrm{\Σ}}^{EH}\right({\mathrm{lT}}^c)-\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\Σ}}{x}_{i,j})\]}^{+},m=1,2,...,M,l=1,2,...N$

${\mathrm{\λ}}_l(n+1)={\[{\mathrm{\λ}}_l\left(n\right)-{\mathrm{\α}}_{{\mathrm{\λ}}_l}\left(n\right)(1-\underset{i=1}{\overset{M}{\Σ}}{x}_{i,j})\]}^{+},l=1,2,...N$

Wherein μ_ml(n),λ_lN () represents the Lagrange factor of nth iteration, α respectively_μml(n),α_λlN () represents phase respectively The iteration step length answered.

Further, the iteration step length of described Subgradient Algorithm iteration renewal equation may be arranged such that

${\mathrm{\α}}_{{\mathrm{\μ}}_{ml}}\left(n\right)={\mathrm{\α}}_{{\mathrm{\λ}}_l}\left(n\right)=\frac{1}{n^2},l=1,2,...,N,m=1,2,...,M.$

Further, the Lagrangian Form of described optimization problem P4 is:

$\begin{array}{c}L(x_{i,j}{\mathrm{\μ}}_{ml},{\mathrm{\λ}}_l)=\frac{\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\∀i,j}{{\mathrm{\γ}}_{i,j}^s,{\mathrm{\γ}}_{i,j}^d,P_{i,j}^{EH}}}\[\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=k+1}{\overset{N}{\mathrm{\Σ}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\]}{N-k+1}\\ -{\mathrm{\μ}}_{ml}(\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\Σ}}{x}_{i,j}-E_{i,\mathrm{\Σ}}^{EH}({\mathrm{lT}}^c))-{\mathrm{\λ}}_l(\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{x}_{i,j}-1)\end{array}$

Simultaneous againAnd with subgradient method iterative, wherein, l ∈ 1 ..., N}, m ∈ 1 ..., M}, μ_ml, λ_lRepresent Lagrange factor.

Further, the Lagrange factor μ in the Lagrangian Form of described optimization problem P4_ml,λ_lIteration update Method uses Subgradient Algorithm, and the iteration renewal equation of described Subgradient Algorithm is

${\mathrm{\μ}}_{ml}(n+1)={\[{\mathrm{\μ}}_{ml}\left(n\right)-{\mathrm{\α}}_{{\mathrm{\μ}}_{ml}}\left(n\right)\left(E_{i,\mathrm{\Σ}}^{EH}\right({\mathrm{lT}}^c)-\frac{l}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\Σ}}{x}_{i,j})\]}^{+},m=1,2,...,M,l=1,2,...N$

${\mathrm{\λ}}_l(n+1)=\[{\mathrm{\λ}}_l\left(n\right)-{\mathrm{\α}}_{{\mathrm{\λ}}_l}\left(n\right){(1-\underset{i=1}{\overset{M}{\Σ}}{x}_{i,j})\]}^{+},l=1,2,...N$

Wherein μ_ml(n),λ_lN () represents the Lagrange factor of nth iteration respectively,Represent phase respectively The iteration step length answered.

Further, the iteration step length of described Subgradient Algorithm iteration renewal equation may be arranged such that

${\mathrm{\α}}_{{\mathrm{\μ}}_{ml}}\left(n\right)={\mathrm{\α}}_{{\mathrm{\λ}}_l}\left(n\right)=\frac{1}{n},l=1,2,...,N,m=1,2,...,M.$

Further, described step 1.2 also includes:

$P_{i,j}^{EH}=\left\{\begin{array}{cc}{P}_{i,ave}^{EH}(1+\mathrm{\ϵ}),& p\\ P_{i,ave}^{EH},& 1-2p\\ P_{i,ave}^{EH}(1-\mathrm{\ϵ}),& p\end{array}\right..$

Beneficial effect:

Compared to prior art, the invention have the benefit that

(1) present invention difference and conventional many relay selection method, it is considered to multi-relay cooperation communication based on energy acquisition, And derive the expression formula of handling capacity under this scene, maximize the throughput performance between communication node, carry out optimum Solving of change problem, has the directive significance of reality；

(2) the present invention is directed to special application scenarios, source reality application, scene setting is careful, reasonable, more has practice to refer to Lead meaning；

(3) present invention takes into full account the environmental protection scheme of regenerative resource, in conjunction with energy acquisition technology, increases and considers that energy is adopted The select permeability of collection relaying, on the premise of not affecting network performance, it is considered to the systematic function optimum under cause and effect restrictive condition is asked Topic, reaches the compromise of energy consumption and network rate, more rationally makes full use of regenerative resource, reduce the energy consumption of network；

(4) the present invention is directed to solving of optimization problem, use convex optimization to process, convert the object function of optimization problem, Without approximate calculation, do not affect the computation complexity reduced greatly while the precision of problem, reduce overhead and produce Time delay；

(5) optimizing of the present invention uses Lagrange multiplier method, and speed of searching optimization is fast, uses subgradient during algorithm iteration Method, and select progressive step-length, optimizing is more accurate；

(6) resource allocation methods of the present invention, algorithm is reasonable in design, it is easy to accomplish.

Accompanying drawing explanation

Fig. 1 is present system scene configuration diagram.

Detailed description of the invention

In order to make the purpose of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, right The present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, and It is not used in the restriction present invention.

Embodiment one

Present invention difference and conventional many relay selection method, it is considered to multi-relay cooperation communication based on energy acquisition, and Derive the expression formula of handling capacity under this scene, maximize the throughput performance between communication node, carry out optimization Solving of problem, has the directive significance of reality.Specifically, the relaying choosing of a kind of energy acquisition multi-relay cooperation communication system Selection method, including:

Step 1: system scenarios analysis, problem is summed up；

Step 1.1: channel model is set up；

The present invention is directed to special application scenarios, source reality application, scene setting is careful, reasonable, more has practical advice Meaning.As shown in Figure 1, it is considered to one, based on energy acquisition multi-relay cooperation communication scenes, has a signal source in scene Source (is abbreviated as S), and the relay station Relay of M energy acquisition (is abbreviated as R_i, i=1,2 ..., M) and a destinations traffic Terminal D estination (is abbreviated as D), it is considered to do not have direct path between signal source S and target communications terminal D, it is necessary to logical Cross the relay station R of energy acquisition_iCooperation forwards, relay station R_iSelecting amplification forwarding working method, a transmitting procedure includes N number of Data block, it is assumed that signal source S and relaying R_iTake equal bandwidth W, it is considered to the relay station of M energy acquisition under this scene Select permeability, assumes that the battery capacity for energy acquisition is sufficiently large simultaneously, ignores except being used for transmitting the energy of consumption.

Assuming that channel is block decline at a slow speed, its response time is defined as T^C, i other words the transmission time of a data block It is T^C.Realistic meaning is had more in order to discuss, it is contemplated that the transmitting procedure of one section of finite length is made up of N number of packet, that Whole transmission cycle T meets: T=NT^C.R is relayed for i-th_i, channel gain when transmitting the data block of jth is defined asWhereinWhen representing transmission jth data block, signal source S is to relay station R_iChannel coefficients,Represent and forward jth Relay station R during data block_iChannel coefficients to target communications terminal D.

Owing to relaying uses the half-duplex mode of two time slots, at first time slot signal source S to whole M energy acquisitions Relay station broadcast transmission signal, energy state and channel conditions according to M energy acquisition relaying select optimal cooperation Forward relay, and consider the transmitting procedure of whole N number of packet, distribute the transmitting power of currently selected relaying.If Have selected i-th relaying when transmitting jth packet, SNR can be defined as end to end:

${\mathrm{\Λ}}_{i,j}=\frac{{\mathrm{\γ}}_{i,j}^s{\mathrm{\γ}}_{i,j}^d}{{\mathrm{\γ}}_{i,j}^s+{\mathrm{\γ}}_{i,j}^d+1},$

WhereinRepresent signal source S and the noise of target communications terminal D respectively Ratio, P^sAnd P^trRepresent signal source S and the transmitting power of relay station Relay, N respectively₀Represent normalized noise power.

Thus we can obtain corresponding handling capacity the most end to end:

R_i,j=log₂(1+Λ_i,j)

Step 1.2: energy acquisition model is set up；

It is defined as the gross energy that i-th relaying adds up to collect by the end of moment t,Represent that i-th relaying exists The energy work rate that jth packet transmission cycle collects, T^ERepresent the time interval of energy acquisition, N^CRepresent that an energy is adopted The time interval of how many packets, N is comprised in collection interval^E=N/N^CRepresent the number of times that in total transmitting procedure, energy arrives.

Thus, whole transmission cycle can also be expressed as T=N^ET^EIf instantaneous through-put power is expressed as P (t), then Energy cause effect relation limits and can be expressed as

Step 2: system mathematic model is set up；

It is contemplated that the transmission mode of AF relaying, when jth data block is transmitted, if we choose i-th to relay, Then

The most also there will be one all select not in situation, we specify that this situation is x_0,j=1.If it is known that energy Amount and channel status, optimization problem can be grouped into:

$\begin{array}{cc}P1:& \underset{x_{i,j}}{max}\stackrel{\‾}{R}=\frac{1}{N}\underset{i=1}{\overset{M}{\Σ}}\underset{j=1}{\overset{N}{\Σ}}{R}_{i,j}{x}_{i,j}\end{array}$

$\begin{array}{cc}s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\Σ}}{x}_{i,j}\≤E_{i,\mathrm{\Σ}}^{EH}\left({\mathrm{lT}}^c\right),\∀l\∈J,i\∈I\end{array},$

$\underset{i=1}{\overset{M}{\Σ}}{x}_{i,j}\≤1,\∀j\∈J,$

x_i,j∈ 0,1}, J={1 ..., N}, I={1 ..., M}.

We analyze optimization problem P1 further:

Wherein, the object function of optimization problem is maximization system average throughput

Optimized variable is relay selection factor x_i,j,

Constraints is

The present invention takes into full account the environmental protection scheme of regenerative resource, in conjunction with energy acquisition technology, increases and considers energy acquisition The select permeability of relaying, on the premise of not affecting network performance, it is considered to the systematic function optimal problem under cause and effect restrictive condition, Reach the compromise of energy consumption and network rate, more rationally make full use of regenerative resource, reduce the energy consumption of network.

Step 3: the convex optimization of optimization problem processes；

For any one transmission bag j, if the given corresponding relay selection result of k≤j≤N, the most accurately know simultaneously The status information of channel, then handling capacity expression formula is as follows:

$\underset{i=1}{\overset{M}{\Σ}}{R}_{i,j}{x}_{i,j}=\underset{i=1}{\overset{M}{\Σ}}{\log }_2(1+{\mathrm{\Λ}}_{i,j})x_{i,j}.$

Thus it was found that relay selection factor x_i,jBeing crucial, it is average that we can write out for jth packet Handling capacity expression formula:

${\stackrel{\&OverBar;}{R}}_{ave,j}=\underset{i=0}{\overset{M}{\&Sigma;}}{\stackrel{\&OverBar;}{R}}_i\mathrm{Pr}ob(x_{i,j}=1|\left\{x_{i,k}\right\},k<j),$

WhereinRepresent the average throughput of i-th relaying.

Thus we to convert object function further as follows:

$\stackrel{\&OverBar;}{R}\left(k\right)=\frac{1}{N-k+1}(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+\underset{j=k+1}{\overset{N}{\mathrm{\&Sigma;}}}{\stackrel{\&OverBar;}{R}}_{ave,j})=\frac{1}{N-k+1}(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+\underset{i=0}{\overset{M}{\mathrm{\&Sigma;}}}{\stackrel{\&OverBar;}{R}}_i{r}_i)$

WhereinRepresenting and start from kth packet until the end of transmission, i-th relays quilt The average time chosen, thus optimization problem P1 can be converted into P2:

$\begin{array}{cc}P2:& \underset{x_{i,j}}{max}\stackrel{\&OverBar;}{R}\left(k\right)=\frac{1}{N-k+1}(\underset{i=1}{\overset{M}{\&Sigma;}}{R}_{i,k}{x}_{i,k}+\underset{i=0}{\overset{M}{\&Sigma;}}{\stackrel{\&OverBar;}{R}}_i{r}_i)\end{array}$

$\begin{array}{cc}s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j}\&le;E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right),\&ForAll;l\&Element;J,i\&Element;I\end{array},$

$\underset{i=1}{\overset{M}{\&Sigma;}}{x}_{i,j}\&le;1,\&ForAll;j\&Element;J,$

x_i,j∈ 0,1}, J={1 ..., N}, I={1 ..., M}.

The present invention is directed to solving of optimization problem, use convex optimization to process, convert the object function of optimization problem, without Cross approximate calculation, do not affect the computation complexity reduced greatly while the precision of problem, reduce that overhead produces time Prolong.

Analysis optimization problem P2 further:

The object function of optimization problem is the system average throughput after maximizing convex process

$\frac{1}{N-k+1}(\underset{i=1}{\overset{M}{\&Sigma;}}{R}_{i,k}{x}_{i,k}+\underset{i=0}{\overset{M}{\&Sigma;}}{\stackrel{\&OverBar;}{R}}_i{r}_i),$

Optimized variable is relay selection factor x_i,j, constraints is

Step 4: solve the average time r that i-th relaying is selected_i；

In the case of knowing ESI energy acquisition curve, based on the EH curve for all relayings, we can ask respectively Going out the average time that each relaying is selected in a continuous print transmission cycle, expression formula is as follows:

Step 5: utilize relative throughput gain to carry out relay selection；

DefinitionWhereinRepresent relaying i₁Selected average throughput,Represent relaying i₂Selected In average throughput, Δ R_i,kRepresent the relative throughput gain of relaying i, thus we obtain the selection strategy collection of relaying such as Under:

$K_k=\arg \underset{i}{max}\left\{{\mathrm{\&Delta;R}}_{i,k}\left|\begin{array}{c}\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j}\&le;E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right),\&ForAll;l\&Element;J,i\&Element;I,\\ \underset{i=1}{\overset{M}{\&Sigma;}}{x}_{i,j}\&le;1,\&ForAll;j\&Element;J,\\ x_{i,j}\&Element;\{0,1\},J=\{1,...,N\},I=\{1,...,M\}.\end{array}\right.\right\},$

And thus make optimal middle rank selection.

Embodiment two

In actual communication system, we sometimes tend not to know in time and accurately the state of whole channel, Even if wanting to know in time and accurately the state of whole channel, also to pay huge Resources Consumption.In order to change this present situation, Change further adaptability and the robustness of algorithm, it is ensured that relay selection algorithm of the present invention is can not to obtain channel completely whole Also can correctly perform in the case of state.On the basis of the embodiment of the present invention one, we have following improvement project.Concrete next Say:

Described step 2 includes, if in the case of the whole state of channel can not be obtained completely, and the target letter of optimization problem Number is rewritten into:

$\stackrel{\&OverBar;}{R}\left(k\right)=E\&lsqb;\frac{1}{N-k+1}\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;=\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\&ForAll;i,j}{{\mathrm{\&gamma;}}_{i,j}^s,{\mathrm{\&gamma;}}_{i,j}^d}}\&lsqb;\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;}{N-k+1}$

Thus optimization problem P1 can be converted into P3:

$\begin{array}{cc}P3:& \underset{x_{i,j}}{\max }\stackrel{\&OverBar;}{R}\left(k\right)\end{array}=\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\&ForAll;i,j}{{\mathrm{\&gamma;}}_{i,j}^s,{\mathrm{\&gamma;}}_{i,j}^d}}\&lsqb;\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k+1}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;}{N-k+1}$

$\begin{array}{cc}s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j}\&le;E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right),\&ForAll;l\&Element;J,i\&Element;I\end{array},$

$\underset{i=1}{\overset{M}{\&Sigma;}}{x}_{i,j}\&le;1,\&ForAll;j\&Element;J,$

x_i,j∈ 0,1}, J={1 ..., N}, I={1 ..., M}.

WhereinRepresent the handling capacity of the whole system of prediction on the premise of known front k is according to bag status information.

Analysis optimization problem P3 further:

Object function is to maximize revised system average throughputOptimize Variable is relay selection factor x_i,j, constraints isWherein E [| x_i,k] represent The expectation after the relaying selecting i-th is given for kth packet when transmitting,It is whole that expression can not obtain channel completely The expectation that the signal to noise ratio statistical property of channel is made is utilized on the premise of state.

Embodiment three

At preceding embodiment two, we discuss the situation of a kind of state that can not know whole channel in time and accurately, Here we consider a kind of more common situation further.If the state of energy acquisition can not be known in time and accurately, On the basis of the embodiment of the present invention one, we have following improvement project.Specifically:

Optimization problem object function is rewritten into:

$\stackrel{\&OverBar;}{R}\left(k\right)=E\&lsqb;\frac{1}{N-k+1}\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;=\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\&ForAll;i,j}{{\mathrm{\&gamma;}}_{i,j}^s,{\mathrm{\&gamma;}}_{i,j}^d,P_{i,j}^{EH}}}\&lsqb;\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;}{N-k+1}$

Thus optimization problem P1 can be converted into P4:

$\begin{array}{cc}P4:& \underset{x_{i,j}}{\max }\stackrel{\&OverBar;}{R}\left(k\right)\end{array}=\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\&ForAll;i,j}{{\mathrm{\&gamma;}}_{i,j}^s,{\mathrm{\&gamma;}}_{i,j}^d,P_{i,j}^{EH}}}\&lsqb;\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k+1}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;}{N-k+1}$

$\begin{array}{cc}s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j}\&le;E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right),\&ForAll;l\&Element;J,i\&Element;I\end{array},$

$\underset{i=1}{\overset{M}{\&Sigma;}}{x}_{i,j}\&le;1,\&ForAll;j\&Element;J,$

x_i,j∈ 0,1}, J={1 ..., N}, I={1 ..., M}.

WhereinRepresent the handling capacity of the whole system of prediction on the premise of known front k is according to bag status information.

Analysis optimization problem P4 further:

Object function is to maximize revised system average throughput Optimized variable is relay selection factor x_i,j, constraints is

Wherein E [| x_i,k] represent the expectation after the relaying giving selection i-th for kth packet when transmit, Expression utilizes the signal to noise ratio statistical property of channel to make on the premise of can not obtaining the whole state of channel and energy acquisition state completely The expectation gone out.

Described step 4 includes: update the average time r that i-th relaying is selected_iAs follows:

Embodiment four

In order to improve further improvement, improving the operation efficiency of algorithm, the present invention proposes a kind of new to solve subproblem P3 Thinking, use Lagrange multiplier method go optimizing, faster, algorithm complex is lower for speed.Specifically, described subproblem P3 solves the Lagrange factor method that can use:

$\begin{array}{c}L(x_{i,j}{\mathrm{\&mu;}}_{ml},{\mathrm{\&lambda;}}_l)=\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\&ForAll;i,j}{{\mathrm{\&gamma;}}_{i,j}^s,{\mathrm{\&gamma;}}_{i,j}^d}}\&lsqb;\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k+1}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;}{N-k+1}\\ -{\mathrm{\&mu;}}_{ml}(\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j}-E_{i,\mathrm{\&Sigma;}}^{EH}({\mathrm{lT}}^c))-{\mathrm{\&lambda;}}_l(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{x}_{i,j}-1)\end{array}$

Simultaneous againAnd with subgradient method iterative, wherein, l ∈ 1 ..., N}, m ∈ 1 ..., M}, μ_ml, λ_lRepresent Lagrange factor.

On the basis of using Lagrange multiplier algorithm, during loop iteration, we can use subgradient each time Method, and select progressive step-length, optimizing is more accurate.Specifically, in the Lagrangian Form of described optimization problem P3 Lagrange factor μ_ml,λ_lIteration update method use Subgradient Algorithm, complexity is lower, more efficiently, described time ladder The iteration renewal equation of degree algorithm is

${\mathrm{\&mu;}}_{ml}(n+1)={\&lsqb;{\mathrm{\&mu;}}_{ml}\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&mu;}}_{ml}}\left(n\right)\left(E_{i,\mathrm{\&Sigma;}}^{EH}\right({\mathrm{lT}}^c)-\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j})\&rsqb;}^{+},m=1,2,...,M,l=1,2,...N$

${\mathrm{\&lambda;}}_l(n+1)={\&lsqb;{\mathrm{\&lambda;}}_l\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&lambda;}}_l}\left(n\right)(1-\underset{i=1}{\overset{M}{\&Sigma;}}{x}_{i,j})\&rsqb;}^{+},l=1,2,...N$

Wherein μ_ml(n),λ_lN () represents the Lagrange factor of nth iteration respectively,Represent phase respectively The iteration step length answered.

So that iteration speed is faster, precision is higher, and we select the iteration step length reduced that goes forward one by one.Described iteration step length May be arranged such that

${\mathrm{\&alpha;}}_{{\mathrm{\&mu;}}_{ml}}\left(n\right)={\mathrm{\&alpha;}}_{{\mathrm{\&lambda;}}_l}\left(n\right)=\frac{1}{n^2},l=1,2,...,N,m=1,2,...,M.$

Embodiment five

In order to improve further improvement, improving the operation efficiency of algorithm, the present invention proposes a kind of new to solve subproblem P4 Thinking, use Lagrange multiplier method go optimizing, faster, algorithm complex is lower for speed.Specifically, described subproblem P4 solves the Lagrange factor method that can use:

$\begin{array}{c}L(x_{i,j}{\mathrm{\&mu;}}_{ml},{\mathrm{\&lambda;}}_l)=\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\&ForAll;i,j}{{\mathrm{\&gamma;}}_{i,j}^s,{\mathrm{\&gamma;}}_{i,j}^d,P_{i,j}^{EH}}}\&lsqb;\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k+1}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;}{N-k+1}\\ -{\mathrm{\&mu;}}_{ml}(\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j}-E_{i,\mathrm{\&Sigma;}}^{EH}({\mathrm{lT}}^c))-{\mathrm{\&lambda;}}_l(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{x}_{i,j}-1)\end{array}$

Simultaneous againAnd with subgradient method iterative, wherein, l ∈ 1 ..., N}, m ∈ 1 ..., M}, μ_ml, λ_lRepresent Lagrange factor.

On the basis of using Lagrange multiplier algorithm, during loop iteration, we can use subgradient each time Method, and select progressive step-length, optimizing is more accurate.Specifically, in the Lagrangian Form of described optimization problem P3 Lagrange factor μ_ml,λ_lIteration update method use Subgradient Algorithm, complexity is lower, more efficiently, described time ladder The iteration renewal equation of degree algorithm is

${\mathrm{\&mu;}}_{ml}(n+1)={\&lsqb;{\mathrm{\&mu;}}_{ml}\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&mu;}}_{ml}}\left(n\right)\left(E_{i,\mathrm{\&Sigma;}}^{EH}\right({\mathrm{lT}}^c)-\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j})\&rsqb;}^{+},m=1,2,...,M,l=1,2,...N$

${\mathrm{\&lambda;}}_l(n+1)={\&lsqb;{\mathrm{\&lambda;}}_l\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&lambda;}}_l}\left(n\right)(1-\underset{i=1}{\overset{M}{\&Sigma;}}{x}_{i,j})\&rsqb;}^{+},l=1,2,...N$

Wherein μ_ml(n),λ_lN () represents the Lagrange factor of nth iteration respectively,Represent phase respectively The iteration step length answered.

So that iteration speed is faster, precision is higher, and we select the iteration step length reduced that goes forward one by one.Described iteration step length May be arranged such that

${\mathrm{\&alpha;}}_{{\mathrm{\&mu;}}_{ml}}\left(n\right)={\mathrm{\&alpha;}}_{{\mathrm{\&lambda;}}_l}\left(n\right)=\frac{1}{n},l=1,2,...,N,m=1,2,...,M.$

Embodiment six

So that algorithm is more nearly actual application, we use the energy acquisition curve of a kind of simplification, aforementioned five All can use on the basis of individual embodiment.Specifically,

Described step 1.2 includes:

$P_{i,j}^{EH}=\left\{\begin{array}{cc}{P}_{i,ave}^{EH}(1+\mathrm{\&epsiv;}),& p\\ P_{i,ave}^{EH},& 1-2p\\ P_{i,ave}^{EH}(1-\mathrm{\&epsiv;}),& p\end{array}\right..$

The foregoing is only presently preferred embodiments of the present invention, not in order to limit the present invention, all essences in the present invention Any amendment, equivalent and the improvement etc. made within god and principle, should be included within the scope of the present invention.

Claims (10)

1. the relay selection method of an energy acquisition multi-relay cooperation communication system, it is characterised in that including:

Step 1: system scenarios analysis, problem is summed up；

Step 1.1: channel model is set up；

Scene has signal source S, the relay station R of M energy acquisition_i, i=1,2 ..., a M and target communications terminal D, Direct path, relay station R is not had between signal source S and target communications terminal D_iSelect amplification forwarding working method, a transmission Process includes N number of data block, and the transmission time of a data block is T^C, signal source S and relaying R_iTake equal bandwidth W, it is considered to The select permeability of the relay station of M energy acquisition under this scene, it is assumed that sufficiently large for the battery capacity of energy acquisition, except Ignore for transmitting the energy of consumption, relay R for i-th_i, channel gain when transmitting the data block of jth is defined asWhereinWhen representing transmission jth data block, signal source S is to relay station R_iChannel coefficients,Represent and forward jth Relay station R during data block_iTo the channel coefficients of target communications terminal D, if having selected i-th when transmitting jth packet Relaying, SNR can be defined as end to end:

${\mathrm{\&Lambda;}}_{i,j}=\frac{{\mathrm{\&gamma;}}_{i,j}^s{\mathrm{\&gamma;}}_{i,j}^d}{{\mathrm{\&gamma;}}_{i,j}^s+{\mathrm{\&gamma;}}_{i,j}^d+1},$

WhereinRepresent signal source S and the signal to noise ratio of target communications terminal D, P respectively^sWith P^trRepresent signal source S and the transmitting power of relay station Relay, N respectively₀Represent normalized noise power, and then obtain corresponding Handling capacity end to end:

R_i,j=log₂(1+Λ_i,j)；

Step 1.2: energy acquisition model is set up；

It is defined as the gross energy that i-th relaying adds up to collect by the end of moment t,Represent that i-th relays in jth The energy work rate that packet transmission cycle collects, T^ERepresent the time interval of energy acquisition, N^CRepresent an energy acquisition interval In comprise the time interval of how many packets, N^E=N/N^CRepresent the number of times that in total transmitting procedure, energy arrives, whole transmission week Phase can be expressed as T=N^ET^E, instantaneous through-put power is defined as P (t), and energy cause effect relation restricted representation is:

$\underset{0}{\overset{t}{\&Integral;}}P\left(\mathrm{\&tau;}\right)d\mathrm{\&tau;}\&le;E_{i,\mathrm{\&Sigma;}}^{EH}\left(t\right);$

Step 2: system mathematic model is set up；

When jth data block is transmitted, if we choose i-th, relaying participates in cooperation, x_i,j=1, otherwise x_i,j=0, Under known energy and channel status, optimization problem can be grouped into:

$\begin{array}{cc}P1:& \underset{x_{i,j}}{\max }\stackrel{\&OverBar;}{R}=\frac{1}{N}\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}\\ & \begin{array}{cc}s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\mathrm{\&Sigma;}}}{x}_{i,j}\&le;E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right),\&ForAll;l\&Element;J,i\&Element;I,\\ & \underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{x}_{i,j}\&le;1,\&ForAll;j\&Element;J,\\ & x_{i,j}\&Element;\left\{0,1\right\},J=\left\{1,...,N\right\},I=\left\{1,...,M\right\}.\end{array}\end{array};$

Step 3: the convex optimization of optimization problem processes；

For any one transmission bag j, if the given corresponding relay selection result of k≤j≤N, the most accurate or channel Status information, then handling capacity expression formula isIt was accordingly found that relay selection factor x_i,jIt is Key, then write out the average throughput expression formula for jth packetWherein Represent the average throughput of i-th relaying；

Convert object function further as follows:

$\stackrel{\&OverBar;}{R}\left(k\right)=\frac{1}{N-k+1}\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+\underset{j=k+1}{\overset{N}{\mathrm{\&Sigma;}}}{\stackrel{\&OverBar;}{R}}_{ave,j}\right)=\frac{1}{N-k+1}\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+\underset{i=0}{\overset{M}{\mathrm{\&Sigma;}}}{\stackrel{\&OverBar;}{R}}_i{r}_i\right)$

WhereinRepresenting and start from kth packet until the end of transmission, i-th relaying is selected Average time, thus optimization problem P1 can be converted into P2:

$\begin{array}{cc}P2:& \underset{x_{i,j}}{\max }\stackrel{\&OverBar;}{R}\left(k\right)=\frac{1}{N-k+1}\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+\underset{i=0}{\overset{M}{\mathrm{\&Sigma;}}}{\stackrel{\&OverBar;}{R}}_i{r}_i\right)\\ & \begin{array}{cc}s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\mathrm{\&Sigma;}}}{x}_{i,j}\&le;E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right),\&ForAll;l\&Element;J,i\&Element;I,\\ & \underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{x}_{i,j}\&le;1,\&ForAll;j\&Element;J,\\ & x_{i,j}\&Element;\left\{0,1\right\},J=\left\{1,...,N\right\},I=\left\{1,...,M\right\}.\end{array}\end{array};$

Step 4: solve the average time r that i-th relaying is selected_i；

In the case of knowing ESI energy acquisition curve, based on the EH curve for all relayings, can obtain respectively each in Continue selected in a continuous print transmission cycle average time

Step 5: utilize relative throughput gain to carry out relay selection；

DefinitionWhereinRepresent relaying i₁Selected averagely gulps down The amount of telling,Represent relaying i₂Selected average throughput, Δ R_i,kRepresent the relative throughput gain of relaying i, thus in obtaining The selection strategy collection continued is as follows:

$K_k=\arg \underset{i}{max}\left\{{\mathrm{\&Delta;R}}_{i,k}\right|\begin{array}{c}\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j}\&le;E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right),\&ForAll;l\&Element;J,i\&Element;I,\\ \underset{i=1}{\overset{M}{\&Sigma;}}{x}_{i,j}\&le;1,\&ForAll;j\&Element;J,\\ x_{i,j}\&Element;\{0,1\},J=\{1,...,N\},I=\{1,...,M\}.\end{array}\}$

And thus make optimal middle rank selection.

A kind of relay selection method the most according to claim 1, it is characterised in that described step 2 includes:

If in the case of can not obtaining the whole state of channel completely, optimization object function is rewritten into:

$\stackrel{\&OverBar;}{R}\left(k\right)=E\&lsqb;\frac{1}{N-k+1}\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;=\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\&ForAll;i,j}{{\mathrm{\&gamma;}}_{i,j}^s,{\mathrm{\&gamma;}}_{i,j}^d}}\&lsqb;\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k+1}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,j}\&rsqb;}{N-k+1}$

Thus optimization problem P1 can be converted into P3:

$\begin{array}{cc}s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j}\&le;E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right),\&ForAll;l\&Element;J,i\&Element;I,\end{array}$

$\underset{i=1}{\overset{M}{\&Sigma;}}{x}_{i,j}\&le;1,\&ForAll;j\&Element;J,$

x_i,j∈ 0,1}, J={1 ..., N}, I={1 ..., M}.

WhereinRepresent the handling capacity of the whole system of prediction on the premise of known front k is according to bag status information；E[· x_i,k] represent the expectation after the relaying giving selection i-th for kth packet when transmit,Expression can not obtain completely The expectation that the signal to noise ratio statistical property of channel is made is utilized on the premise of the whole state of channel.

A kind of relay selection method the most according to claim 1, it is characterised in that described step 2 includes:

If channel status can not accurately be obtained, the state of energy acquisition can not be known in time and accurately, and optimization problem is rewritten Become:

$\stackrel{\&OverBar;}{R}\left(k\right)=E\&lsqb;\frac{1}{N-k+1}\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;=\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\&ForAll;i,j}{{\mathrm{\&gamma;}}_{i,j}^s,{\mathrm{\&gamma;}}_{i,j}^d,P_{i,j}^{EH}}}\&lsqb;\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k+1}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;}{N-k+1}$

Thus optimization problem P1 can be converted into P4:

$\begin{array}{cc}s.t.& \frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j}\&le;E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right),\&ForAll;l\&Element;J,i\&Element;I,\end{array}$

$\underset{i=1}{\overset{M}{\&Sigma;}}{x}_{i,j}\&le;1,\&ForAll;j\&Element;J,$

x_i,j∈ 0,1}, J={1 ..., N}, I={1 ..., M}.

WhereinExpression is the handling capacity of the whole system of prediction on the premise of known front k is according to bag status information, E [| x_i,k] represent the expectation after the relaying giving selection i-th for kth packet when transmit,Expression can not obtain completely The expectation that the signal to noise ratio statistical property of channel is made is utilized on the premise of obtaining the whole state of channel and energy acquisition state；

Described step 4 includes: update the average time r that i-th relaying is selectedⁱAs follows

A kind of relay selection method the most according to claim 2, it is characterised in that described step 2 optimization problem P3's Lagrangian Form is:

$\begin{array}{c}L\left(x_{i,j},{\mathrm{\&mu;}}_{ml},{\mathrm{\&lambda;}}_l\right)=\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\&ForAll;i,j}{{\mathrm{\&gamma;}}_{i,j}^s,{\mathrm{\&gamma;}}_{i,j}^d}}\&lsqb;\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k+1}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;}{N-k+1}\\ -{\mathrm{\&mu;}}_{ml}\left(\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\mathrm{\&Sigma;}}}{x}_{i,j}-E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right)\right)-{\mathrm{\&lambda;}}_l\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{x}_{i,j}-1\right)\end{array}$

Simultaneous againAnd with subgradient method iterative, wherein, l ∈ 1 ..., N}, m ∈ 1 ..., M}, μ_ml,λ_lTable Show Lagrange factor.

A kind of relay selection method the most according to claim 4, it is characterised in that the Lagrange of described optimization problem P3 Lagrange factor μ in form_ml,λ_lIteration update method use Subgradient Algorithm, the iteration of described Subgradient Algorithm is more New equation is:

${\mathrm{\&mu;}}_{ml}(n+1)={\&lsqb;{\mathrm{\&mu;}}_{ml}\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&mu;}}_{ml}}\left(n\right)\left(E_{i,\mathrm{\&Sigma;}}^{EH}\right({\mathrm{lT}}^c)-\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\&Sigma;}}{x}_{i,j})\&rsqb;}^{+},m=1,2,...,M,l=1,2,...N$

${\mathrm{\&lambda;}}_l(n+1)={\&lsqb;{\mathrm{\&lambda;}}_l\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&lambda;}}_l}\left(n\right)(1-\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{x}_{i,j})\&rsqb;}^{+},l=1,2,...N$

Wherein μ_ml(n),λ_lN () represents the Lagrange factor of nth iteration respectively,Represent corresponding respectively Iteration step length.

A kind of relay selection method the most according to claim 5, it is characterised in that described Subgradient Algorithm iteration renewal side The iteration step length of journey may be arranged such that

${\mathrm{\&alpha;}}_{{\mathrm{\&mu;}}_{ml}}\left(n\right)={\mathrm{\&alpha;}}_{{\mathrm{\&lambda;}}_l}\left(n\right)=\frac{1}{n^2},l=1,2,...,N,m=1,2,...,M.$

A kind of relay selection method the most according to claim 3, it is characterised in that the glug of described optimization problem P4 is bright Day form is:

$\begin{array}{c}L\left(x_{i,j},{\mathrm{\&mu;}}_{ml},{\mathrm{\&lambda;}}_l\right)=\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{R}_{i,k}{x}_{i,k}+E_{\underset{\&ForAll;i,j}{{\mathrm{\&gamma;}}_{i,j}^s,{\mathrm{\&gamma;}}_{i,j}^d,P_{i,j}^{EH}}}\&lsqb;\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{j=k+1}{\overset{N}{\mathrm{\&Sigma;}}}{R}_{i,j}{x}_{i,j}|x_{i,k}\&rsqb;}{N-k+1}\\ -{\mathrm{\&mu;}}_{ml}\left(\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\mathrm{\&Sigma;}}}{x}_{i,j}-E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right)\right)-{\mathrm{\&lambda;}}_l\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{x}_{i,j}-1\right)\end{array}$

Simultaneous againAnd with subgradient method iterative, wherein, l ∈ 1 ..., N}, m ∈ 1 ..., M}, μ_ml,λ_lTable Show Lagrange factor.

A kind of relay selection method the most according to claim 7, it is characterised in that the Lagrange of described optimization problem P4 Lagrange factor μ in form_ml,λ_lIteration update method use Subgradient Algorithm, the iteration of described Subgradient Algorithm is more New equation is

${\mathrm{\&mu;}}_{ml}\left(n+1\right)={\&lsqb;{\mathrm{\&mu;}}_{ml}\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&mu;}}_{ml}}\left(n\right)\left(E_{i,\mathrm{\&Sigma;}}^{EH}\left({\mathrm{lT}}^c\right)-\frac{1}{2}{P}^{tr}{T}^C\underset{j=1}{\overset{l}{\mathrm{\&Sigma;}}}{x}_{i,j}\right)\&rsqb;}^{+},m=1,2,...,M,l=1,2,...N$

${\mathrm{\&lambda;}}_l(n+1)={\&lsqb;{\mathrm{\&lambda;}}_l\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&lambda;}}_l}\left(n\right)(1-\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{x}_{i,j})\&rsqb;}^{+},l=1,2,...N$

Wherein μ_ml(n),λ_lN () represents the Lagrange factor of nth iteration respectively,Represent corresponding respectively Iteration step length.

A kind of relay selection method the most according to claim 8, it is characterised in that described Subgradient Algorithm iteration renewal side The iteration step length of journey may be arranged such that

${\mathrm{\&alpha;}}_{{\mathrm{\&mu;}}_{ml}}\left(n\right)={\mathrm{\&alpha;}}_{{\mathrm{\&lambda;}}_l}\left(n\right)=\frac{1}{n},l=1,2,...,N,m=1,2,...,M.$

10. according to any one relay selection method described in claim 1-9, it is characterised in that described step 1.2 is also wrapped Include:

$P_{i,j}^{EH}=\left\{\begin{array}{cc}{P}_{i,ave}^{EH}(1+\mathrm{\&epsiv;}),& p\\ P_{i,ave}^{EH},& 1-2p\\ P_{i,ave}^{EH}(1-\mathrm{\&epsiv;}),& p\end{array}\right..$