CN106304363A - Federated resource distribution method based on real-time Communication for Power energy acquisition cellular network - Google Patents

Abstract

The invention discloses a kind of federated resource distribution method based on real-time Communication for Power energy acquisition cellular network, belong to mobile communication technology field.Including step: system scenarios analysis, problem describes；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, abandon and relay the high complexity and high network cost brought more, single relay selection algorithm and corresponding resource allocation algorithm is studied from the angle being prone to practice, consider single relay cooperative 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.

Description

Federated resource distribution method based on real-time Communication for Power energy acquisition cellular network

Technical field

The invention belongs to mobile communication technology field, more particularly, it relates to one is based on real-time Communication for Power energy acquisition honeybee The federated resource distribution method of nest network.

Background technology

Collaboration communication (Cooperative communication) basic thought can trace back to Cover and El Gamal The research about the theory of information characteristic of trunk channel in 1979.They analyze three classical meshed networks (source node, Destination node and via node) capacity, and assume that the working band of all nodes is identical, such system just can be decomposed into one Broadcast channel (from the point of view of source node) and an access channel (from the point of view of destination node).But, collaboration communication is at a lot of aspects There are differences with trunk channel.First, current research concentrates on and how to produce the diversity overcoming decline, Cover and El Gamal Main Analysis channel capacity under additive white Gaussian noise channels (AWGN)；Secondly, the relaying purpose of trunk channel is In order to help main channel, collaboration communication whole system resource is fixing, and user is information source and relaying person.It is obvious that Coordination mechanism will cause bit rate and launches the compromise of power.For power, on the one hand, when carrying out collaboration communication, due to Each user should launch its data, relays other user data again, so needing more power；On the other hand, due to Creating diversity gain, each user substantially launches power and can suitably reduce.Speed there is also same problem, at collaboration communication Time, although each user should send its data, relays other partner data again, but owing to creating collaboration diversity, often The spectrum efficiency of individual user is all improved, and channel code rate thus improves.This also form one and trades off.Therefore, there is people's handle Collaboration communication sees the joint game of the rate of doing work and bandwidth.

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) in collaboration communication key challenge is how distribution and supervisory relay node, although chooses more relaying Node can obtain better performance, but is as the increase of via node number, and the performance gain brought will reduce, and assist simultaneously The complexity making the design of scheme, signal detection and multiple access problem etc. can increase, and network cost also can improve, therefore from practice From the perspective of, single relay selection algorithm and corresponding resource allocation algorithm get growing concern for；

2) 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；

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. 201510593398.9, publication date on August 7th, 2009, discloses a entitled wireless Resource distribution in relay system, comprising: receive the first control information from the first base station, wherein said first controls information exists Occurring in very first time interval and first frequency subset, wherein the second base station is configured to be spaced and described in the described very first time First frequency subset is launched the second control information；Receive the 3rd control information from the 3rd base station, the wherein said 3rd controls letter Ceasing and occur in interval of the described very first time and second frequency subset, wherein the 4th base station was configured between the described very first time Every launching the 4th control information in described second frequency subset, wherein said 3rd base station and the 4th base station are low-power bases Stand；The first payload data is received from described first base station based on described first control information；Based on described 3rd control information from Described 3rd base station receives the second payload data.The method is avoided in the radio resource elements distributing to control data, from The signal of Base Transmitter and the interference between the signal that relay station is launched.

Chinese invention patent number 200880105998.4, publication date on July 7th, 2008, discloses a entitled wireless Resource distribution in hop relay network, it comprises the following steps: dispatches and is used for packet via one or more relay stations Sending the radio link to one or more movement stations, wherein, each radio link is provided which between two radio stations Wireless connections, described radio station includes base station, one or more relay station and one or more movement station, and, described One or more relay stations include the first relay station；Determining that resource is distributed based on described scheduling, the distribution of described resource at least distributes In order to send the bandwidth of one or more described packet in described radio link one；For distributing with described resource The one or more described packet being associated produces one or more mapping references, and wherein, each mapping is with reference to all by least One packet associates with the distribution of corresponding bandwidth；And by described resource distribute, the one or more map reference Sending to described first relay station with described packet, wherein, described first relay station is according to described resource distribution and described Individual or multiple mapping references forward described packet.This technology can be used to relevant with the distribution of corresponding resource for packet The information of connection.

Chinese invention patent number 200910215874.8, publication date December in 2009 30 days, disclose a entitled one Plant relay network system and descending resource allocation method thereof, comprising: base station is according to asking the terminal of service in the stage of submitting Channel conditions quantized value and the average of terminal on allowable resource have obtained resource quantized value, and computing terminal is at allowable resource On priority；The allowable resource in the stage of submitting is distributed to channel conditions quantized value and is not less than first pre-set by base station Threshold value and the highest terminal of calculated priority；Submitting after the allowable resource in stage is assigned, base station is distributed The allowable resource of delivery phase.Should be for these three links of the link that direct transfers, back haul link and the access link in junction network Between Radio Resource distribution carried out rational design, can improve cell throughout, increase spectrum efficiency.

Generally speaking, the open material of application number 201510593398.9 considers that the interference between base station and relay station is asked Topic, but do not set about from user perspective, do not account for the situation that user throughput is optimum.Application number 200880105998.4 Open material considers that packet distributes the federated resource distribution under being associated with corresponding resource, but it is complicated not account for algorithm Degree and the requirement of real-time operation.The open material of application number 200910215874.8 considers the relay selection under many relay scenes, But do not deeply consider algorithm list relay selection and resources optimization configuration thereof.

Summary of the invention

Energy acquisition factor band is not taken into full account for existing cooperation communication system list relaying and resource allocation method thereof The performance improvement come, the joint relay selection under energy cause and effect restrictive condition, requirement of real-time, the actual application of low complexity algorithm Etc. problem, the present invention proposes the federated resource distribution method of a kind of real-time Communication for Power energy acquisition cellular network, is considering energy Joint relay selection under amount cause and effect restrictive condition, the energy-saving scheme brought in conjunction with energy acquisition technology, auxiliary low complex degree is repeatedly For algorithm, maximize the network performance of user's real-time Communication for Power.

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

Federated resource distribution method based on real-time Communication for Power energy acquisition cellular network, including:

Step 1: system scenarios analysis, problem describes；

Scene has the cellulor S of a stable power-supplying, the relay station R of an energy acquisition and a destinations traffic honeycomb User is D, has direct path between cellulor S and phone user D, and relay station R selects decoding to forward working method, a transmission Process includes that N number of data block, cellulor S and relaying R take equal bandwidth B, it is considered to the cellulor S under this scene and relay station The federated resource assignment problem of R；

Assume that energy acquisition model uses Bernoulli process, sufficiently large for the battery capacity of energy acquisition, except being used for The energy that transmission consumes is ignored, and channel is block decline at a slow speed, and its response time is defined as T_P,Represent the i-th packet Channel coefficients between signal source S and relay station R during transmission,When representing the i-th packet transmission, signal source S saves with destinations traffic Channel coefficients between some D,Represent the channel coefficients between relay station R and destination communication node D when the i-th packet transmits；

Use half-duplex relay cooperative communication mode, comprise two time slots in the transmission time of each packet, in i-th First time slot of packet transmission time, signal source S transport symbol s uses constant power p_s, s meets s～CN (0,1), in The reception signal of R and purpose communication node D of continuing is expressed as:

$y_{s,r}^i=A_0^{\frac{1}{2}}{d}_0^{\frac{\mathrm{\α}}{2}}{d}_1^{-\frac{\mathrm{\α}}{2}}{h}_{s,r}^i\sqrt{p_s}s+n_r^i,\∀i=1,\mathrm{...},N$

$y_{s,d}^i=A_0^{\frac{1}{2}}{d}_0^{\frac{\mathrm{\α}}{2}}{d}^{-\frac{\mathrm{\α}}{2}}{h}_{s,d}^i\sqrt{p_s}s+n_d^i,\∀i=1,\mathrm{...},N,$

Wherein: d₀And A₀Represent reference distance and reference power, the d of large scale decline respectively₁Cellulor S is represented respectively with d And the distance between distance and cellulor S and destinations traffic user D between relay station R,WithRepresent i-th number respectively According to relay station R and the reception noise of destinations traffic user D of packet transmission time, α represents large scale fading factor,WithPoint Biao Shi the relay station R of i-th packet transmission time and the channel gain of destinations traffic user D；

The signal to noise ratio of the first time slot relay reception is expressed asWherein N₀Represent normalizing The noise power changed, W represents the bandwidth of distribution, second time slot, and destination communication node D uses high specific accepting method, receives The signal from relay station R be expressed asWherein d₂WithIn representing respectively Continue the distance between the R and destinations traffic user D of station and channel gain,Represent the second time slot of i-th packet transmission time Time destinations traffic user D reception noise, p_rRepresent the transmitting power of corresponding relay station R, receive from cellulor S and in Continue station R signal to noise ratio be expressed as

${\mathrm{SNR}}_{d_R}^i=A_0{d}_0^{\mathrm{\α}}\left(d^{-\mathrm{\α}}\right|h_{s,d}^i{|}^2{p}_s+d_2^{-\mathrm{\α}}\left|h_{r,d}^i{|}^2{p}_r^i\right){\left(N_{0}W\right)}^{-1};$

If selection tie link, the signal of the cellulor S that destinations traffic user D receives is expressed as

Accordingly, the letter of the cellulor S that destinations traffic user D receives Ratio of making an uproar is expressed as

Step 2: derivation average interrupt probability；

Described average interrupt probability isWherein OⁱRepresent the outage probability of i & lt transmission；

Step 3: optimization problem is summed up；

Under the conditions of above-mentioned supposed premise and constraint, sum up optimization problem as follows:

$P1:\underset{g}{\min }{E}_{D^i,E_H^i,\∀i}^g\[\frac{\underset{i=1}{\overset{N}{\Σ}}{O}^i\left(g\right(S^i,i\left)\right)}{N}\]$

$s.t.\underset{k=1}{\overset{i}{\Σ}}{p}_r^k{T}_p\≤\underset{n=1}{\overset{i}{\Σ}}{E}_H^n,$

$p_r^i\≥0,$

rⁱ∈{0,1},

$r^i\≤D^i,\∀i\∈\{1,2,\mathrm{...},N\}.$

WhereinRepresenting energy causality constraint, the object function of optimization problem is to minimizeThe variable optimized is g (Sⁱ, i), the constraints of optimization is

$\underset{k=1}{\overset{i}{\Σ}}{p}_r^k{T}_p\≤\underset{n=1}{\overset{i}{\Σ}}{E}_H^n,p_r^i\≥0,r^i\∈\{0,1\},r^i\≤D^i,\∀i\∈\{1,2,\mathrm{...},N\}\mathrm{..}$

Step 4: duty Optimization；

The optimized variable of optimization problem is model selection factor ri of transmission every time and corresponding power distribution SchemerⁱDiscrete variable span for 1,0},Continuous variable, span is more than or equal to 0, uses generation The method of valency function, defines Sⁱ=< Dⁱ,εⁱ> represent state set,Expression behavior collection, defines cost function simultaneouslyUtilize convex optimum theory about the alternative manner of cost function, can obtain :

$J^i\left(S^i\right)=\left\{\begin{array}{c}{\min }_{a^N\&Element;A\left(S^N\right)}{O}^N,i=N\\ {\min }_{a^N\&Element;A\left(S^i\right)}{O}^i+J^{i+1}\left(S^{i+1}\right),i<N\end{array}\right.,$

$A(S^i=<D^i,{\mathrm{\&epsiv;}}^i>)=\left\{\begin{array}{c}\{0,1\}\&times;\&lsqb;0,{\mathrm{\&epsiv;}}^i{T}_p^{-1}\&rsqb;,D^i=1\\ \left\{0\right\}\&times;\&lsqb;0,{\mathrm{\&epsiv;}}^i{T}_p^{-1}\&rsqb;,D^i=1\end{array}\right.,$

Finally, optimal solution can be expressed as

Further, described energy acquisition model uses Bernoulli process, including:

Wherein: E represents basic energy unit,Represent that the energy that the i-th time slot collects, each time slot averaged acquisition arrive Energy can be expressed as:Wherein ρ represents the parameter of Bernoulli process, 2T_pRepresent a time slot Length, it is assumed that the battery capacity of relaying is sufficiently large, corresponding energy cause and effect restrictive condition can be expressed as:

$\underset{i=1}{\overset{l}{\&Sigma;}}{p}_r^i{T}_p\&le;\underset{k=1}{\overset{l}{\&Sigma;}}{E}_H^k,\&ForAll;l=1,\mathrm{...},N.$

${\mathrm{\&epsiv;}}^{i+1}={({\mathrm{\&epsiv;}}^i-p_r^i{T}_p)}^{+}+E_H^{i+1},i-1,\mathrm{...},N-1,{\left(x\right)}^{+}=max\{x,0\},{\mathrm{\&epsiv;}}^1=E_H^1.$

Wherein: εⁱ⁺¹Represent after i+1 transmission time slot remaining energy, (x) in battery⁺Represent that x and 0 takes higher value Function, it is assumed that energy content of battery initial value is set as

Further, described step 2 includes:

Step 2.1: use Bernoulli process to judge whether i-th relaying is successfully decoded:

Wherein Dⁱ=1 represents the decoding of i & lt transmission success, Dⁱ=0 represents that i & lt transmission decodes unsuccessfully,

Step 2.2: redefine selection result r of link iⁱ∈ { 0,1}, wherein rⁱ=1 represents that the i-th time slot relaying participates in Cooperation forwards data, rⁱ=0 represents that the i-th time slot relaying is not involved in cooperation, cellulor S retransmission data, redefines i & lt transmission Outage probability OⁱForWherein

Step 2.3: the outage probability expression formula after derivation correction strategy；

$O_R\left(p_r^i\right)=\mathrm{Pr}\{{\mathrm{WT}}_p{\log }_2(1+{\mathrm{SNR}}_{d_R}^i)\&GreaterEqual;2T_{p}WR\}=\left\{\begin{array}{c}1-\frac{{\mathrm{\&mu;}}_2{e}^{-{\mathrm{\&mu;}}_{1}x}-{\mathrm{\&mu;}}_1{e}^{-{\mathrm{\&mu;}}_{2}x}}{{\mathrm{\&mu;}}_2-{\mathrm{\&mu;}}_1},{\mathrm{\&mu;}}_1\&NotEqual;{\mathrm{\&mu;}}_2.\\ 1-(1+{\mathrm{\&mu;}}_1)e^{-{\mathrm{\&mu;}}_{1}x},{\mathrm{\&mu;}}_1={\mathrm{\&mu;}}_2,\end{array}\right.$

$O_D=\mathrm{Pr}\{{\mathrm{WT}}_p{\log }_2(1+{\mathrm{SNR}}_{d_D}^i)<2T_{p}WR\}=1-e^{-\frac{{\mathrm{\&mu;}}_{1}x}{2}}.$

WhereinO_RRepresent cooperation probability, O_BRepresent recurrence probability.

Further, described step 3 also includes that the convex optimization of optimization problem P1 processes, and object function form is converted into:

$P2:\underset{p_r^i,r^i}{\min }\frac{O^i+E_{D^k,E_H^k,\&ForAll;k>i}\&lsqb;\underset{k=i+1}{\overset{N}{\&Sigma;}}{O}^k|p_r^i,r^i\&rsqb;}{N-i+1},\&ForAll;i.$

$s.t.\underset{k=1}{\overset{i}{\&Sigma;}}{p}_r^k{T}_p\&le;\underset{n=1}{\overset{i}{\&Sigma;}}{E}_H^n,$

$p_r^i\&GreaterEqual;0,$

rⁱ∈{0,1},

$r^i\&le;D^i,\&ForAll;i\&Element;\{1,2,\mathrm{...},N\}.$

WhereinRepresenting energy causality constraint, the variable of optimization isrⁱ, object function is to minimizeThe constraints optimized is Set:K ＞ i, thenIf FixedThe functional relationship of k ＞ i, optimization problem can obtain optimal solution.

Further, the through-put power anticipation function in the future such as described step 4 problem solving employing, arrange

Further, described step 4 problem solving uses reserved through-put power function, arranges:

$f_{i,j}\left(p_r^i\right)=\mathrm{\&beta;}\frac{E\&lsqb;{\mathrm{\&epsiv;}}^j\&rsqb;}{T_p},E\&lsqb;{\mathrm{\&epsiv;}}^j\&rsqb;=\left\{\begin{array}{c}{\mathrm{\&epsiv;}}^i-p_r^i{T}_p+\mathrm{\&rho;}E,j=i+1,\\ {\stackrel{\&OverBar;}{\mathrm{\&beta;}}}^{j-i-1}({\mathrm{\&epsiv;}}^i-p_r^i{T}_p)+\mathrm{\&rho;}E\left(\frac{1-{\stackrel{\&OverBar;}{\mathrm{\&beta;}}}^{j-i}}{\mathrm{\&beta;}}\right),j>i+1\end{array}\right.,\stackrel{\&OverBar;}{\mathrm{\&beta;}}=1-\mathrm{\&beta;}.$

Further, described step 4 problem solving uses prediction energy acquisition curvilinear function, arranges:

$p_r^i=\left\{\begin{array}{cc}\frac{{\mathrm{\&epsiv;}}^i}{T_p},& {\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}E\\ \frac{{\mathrm{\&epsiv;}}^i+(N-i)\mathrm{\&rho;}E}{(N-i+1)T_p},& {\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}E\end{array}\right.\mathrm{..}$

Further, described step 4 problem solving uses prediction energy acquisition curvilinear function, arranges:

$p_r^i=\left\{\begin{array}{c}\frac{{\mathrm{\&epsiv;}}^i}{T_p},{\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}Eor{O}_R\left(\frac{{\mathrm{\&epsiv;}}^i+(N-i)\mathrm{\&rho;}E}{(N-i+1)T_p}\right)\&GreaterEqual;{\mathrm{\&gamma;O}}_R\left(0\right)\\ \frac{{\mathrm{\&epsiv;}}^i+(N-i)\mathrm{\&rho;}E}{(N-i+1)T_p},{\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}E\end{array}\right.\mathrm{..}$

Further, described step 3 optimization problem solve employing method of Lagrange multipliers, including:

First the Lagrangian Form of optimization problem P2 is write out:

$L(r^i,p_r^i,{\mathrm{\&mu;}}_l,{\mathrm{\&lambda;}}_l)=\stackrel{\&OverBar;}{O}-{\mathrm{\&mu;}}_i(\underset{k=1}{\overset{i}{\&Sigma;}}{p}_r^k{T}_p-\underset{n=1}{\overset{i}{\&Sigma;}}{E}_H^n)-{\mathrm{\&lambda;}}_i(r^i-D^i),i\&Element;\{1,2,\mathrm{...},N\}$

Simultaneous againAnd with subgradient method iterative, wherein, l ∈ 1 ..., N}, μ_l,λ_lExpression is drawn The Ge Lang factor.

Further, the Lagrange factor μ in the Lagrangian Form of described optimization problem P2_l,λ_lIteration renewal side Method uses Subgradient Algorithm, and the iteration renewal equation of described Subgradient Algorithm is:

${\mathrm{\&mu;}}_l(n+1)={\&lsqb;{\mathrm{\&mu;}}_l\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&mu;}}_l}\left(n\right)(\underset{n=1}{\overset{i}{\&Sigma;}}{E}_H^n-\underset{k=1}{\overset{i}{\&Sigma;}}{p}_r^k{T}_p)\&rsqb;}^{+},i=1,2,\mathrm{...},N,l=1,2,\mathrm{...}N$

${\mathrm{\&lambda;}}_l(n+1)={\&lsqb;{\mathrm{\&lambda;}}_l\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&lambda;}}_l}\left(n\right)(D^i-r^i)\&rsqb;}^{+},i=1,2,\mathrm{...},N,l=1,2,\mathrm{...}N$

Wherein μ_l(n),λ_lN () represents the Lagrange factor of nth iteration respectively,Represent corresponding respectively Iteration step length, described iteration step length may be arranged such that

Beneficial effect:

Compared to prior art, the invention have the benefit that

(1) 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；

(2) present invention difference and conventional many relay selection method, abandon and relay the high complexity and high network brought more Cost, studies single relay selection algorithm and corresponding resource allocation algorithm from the angle being prone to practice, it is considered to Single relay cooperative communication based on energy acquisition, and derive the expression formula of handling capacity under this scene, maximize communication Throughput performance between node, carries out solving of optimization problem, has the directive significance of reality；

(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

Federated resource distribution method based on real-time Communication for Power energy acquisition cellular network, including:

Step 1: system scenarios analysis, problem describes；

The present invention is directed to special application scenarios, source reality application, scene setting is careful, reasonable, more has practical advice Meaning.Consider three node scenes in an energy acquisition small cell network, scene has the cellulor of a stable power-supplying Small Cell (is abbreviated as S), the relay station Relay (being abbreviated as R) of an energy acquisition and a destinations traffic phone user Destination (is abbreviated as D), it is considered to have direct path between cellulor S and phone user D, can be turned by relay station R Sending out, relay station R selects decoding to forward working method, and a transmitting procedure includes N number of data block, it is assumed that cellulor S and relaying R accounts for By equal bandwidth B, it is considered to the cellulor S under this scene and the federated resource assignment problem of relay station R.

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.Assume to be used for The battery capacity of energy acquisition is sufficiently large, ignores except being used for transmitting the energy of consumption.Channel is block decline at a slow speed, its Response time is defined as T_P, i other words the transmission time of a data block is T_P.Realistic meaning is had more, Wo Menkao in order to discuss The transmitting procedure considering one section of finite length is made up of N number of packet, then whole transmission cycle T meets: T=NT_P。Represent the I packet transmission time signal source S and relay station R between channel coefficients；Signal source S and mesh when representing the i-th packet transmission Channel coefficients between mark communication node D；Represent the letter between relay station R and destination communication node D when the i-th packet transmits Road coefficient；We use half-duplex relay cooperative communication mode, comprise two time slots in the transmission time of each packet.

At first time slot of i-th packet transmission time, if signal source S transport symbol s uses constant power p_s, S meets s～CN (0,1) simultaneously, then the reception signal of relaying R and purpose communication node D can be expressed as:

$y_{s,r}^i=A_0^{\frac{1}{2}}{d}_0^{\frac{\mathrm{\&alpha;}}{2}}{d}_1^{-\frac{\mathrm{\&alpha;}}{2}}{h}_{s,r}^i\sqrt{p_s}s+n_r^i,\&ForAll;i=1,\mathrm{...},N$

$y_{s,d}^i=A_0^{\frac{1}{2}}{d}_0^{\frac{\mathrm{\&alpha;}}{2}}{d}^{-\frac{\mathrm{\&alpha;}}{2}}{h}_{s,d}^i\sqrt{p_s}s+n_d^i,\&ForAll;i=1,\mathrm{...},N,$

Wherein: d₀And A₀Represent reference distance and reference power, the d of large scale decline respectively₁Cellulor S is represented respectively with d And the distance between distance and cellulor S and destinations traffic user D between relay station R,WithRepresent i-th number respectively According to relay station R and the reception noise of destinations traffic user D of packet transmission time, α represents large scale fading factor,WithPoint Biao Shi the relay station R of i-th packet transmission time and the channel gain of destinations traffic user D.

The signal to noise ratio of the first time slot relay reception can be expressed as

${\mathrm{SNR}}_r^i=A_0{d}_0^{\mathrm{\&alpha;}}{d}_1^{-\mathrm{\&alpha;}}|h_{s,r}^i{|}^2{p}_s{\left(N_{0}W\right)}^{-1}.$

Wherein: N₀Representing normalized noise power, W represents the bandwidth of distribution；In second time slot, destination communication node D uses high specific accepting method, and purpose communication node can be respectively received the signal that direct transfers of signal source S, and relay station R Forward signal.The signal from relay station R that purpose communication node D receives is expressed as

${\tilde{y}}_{r,d}^i=A_0^{\frac{1}{2}}{d}_0^{\frac{\mathrm{\&alpha;}}{0}}{d}_2^{-\frac{\mathrm{\&alpha;}}{2}}{h}_{r,d}^i\sqrt{p_r^i}s+{\tilde{n}}_d^i,\&ForAll;i=1,\mathrm{...},N$

Wherein: d₂WithRepresent the distance between relay station R and destinations traffic user D and channel gain respectively,Represent The reception noise of destinations traffic user D, p during the second time slot of i-th packet transmission time_rRepresent corresponding relay station R's Launch power.

Therefore, use high specific acceptance criterion destinations traffic user D receive from cellulor S and relay station R Signal to noise ratio is expressed as:

${\mathrm{SNR}}_{d_R}^i=A_0{d}_0^{\mathrm{\&alpha;}}\left(d^{-\mathrm{\&alpha;}}\right|h_{s,d}^i{|}^2{p}_s+d_2^{-\mathrm{\&alpha;}}\left|h_{r,d}^i{|}^2{p}_r^i\right){\left(N_{0}W\right)}^{-1};$

If selection tie link, the signal of the cellulor S that destinations traffic user D receives is expressed as:

${\tilde{y}}_{s,d}^i=A_0^{\frac{1}{2}}{d}_0^{\frac{\mathrm{\&alpha;}}{2}}{d}^{-\frac{\mathrm{\&alpha;}}{2}}{h}_{s,d}^i\sqrt{p_s}s+{\tilde{n}}_d^i,\&ForAll;i=1,\mathrm{...},N,$

Accordingly, the signal to noise ratio of the cellulor S that destinations traffic user D receives is expressed as:

${\mathrm{SNR}}_{d_D}^i=2A_0{d}_0^{\mathrm{\&alpha;}}{d}^{-\mathrm{\&alpha;}}|h_{s,d}^i{|}^2{p}_s{\left(N_{0}W\right)}^{-1}.$

Step 2: derivation average interrupt probability；

Described average interrupt probability isWherein OⁱRepresent the outage probability of i & lt transmission.

Step 3: optimization problem is summed up；

$P1:\underset{g}{\min }{E}_{D^i,E_H^i,\&ForAll;i}^g\&lsqb;\frac{\underset{i=1}{\overset{N}{\&Sigma;}}{O}^i\left(g\right(S^i,i\left)\right)}{N}\&rsqb;$

$s.t.\underset{k=1}{\overset{i}{\&Sigma;}}{p}_r^k{T}_p\&le;\underset{n=1}{\overset{i}{\&Sigma;}}{E}_H^n,$

$p_r^i\&GreaterEqual;0,$

rⁱ∈{0,1},

$r^i\&le;D^i,\&ForAll;i\&Element;\{1,2,\mathrm{...},N\}.$

We analyze the object function of optimization problem P1: optimization problem further is to minimizeExcellent The variable changed is g (Sⁱ, i), the constraints of optimization is WhereinRepresent energy causality constraint, present invention difference and conventional many relay selection method, abandon many relayings The high complexity brought and high network cost, study single relay selection algorithm and phase from the angle being prone to practice The resource allocation algorithm answered, it is considered to single relay cooperative communication based on energy acquisition, and derive handling up under this scene The expression formula of amount, maximizes the throughput performance between communication node, carries out solving of optimization problem, has the guidance of reality Meaning.

Step 4: duty Optimization；

The present invention utilizes cost function method to solve, effectively, the most real-time, is beneficial to reality application.

First analyze: optimized variable is model selection factor r of transmission every timeⁱAnd corresponding power allocation schemerⁱ Discrete variable span for 1,0},Continuous variable, span is more than or equal to 0.

Then, the method using cost function, define Sⁱ=< Dⁱ,εⁱ> represent state set,Expression behavior collection, Define cost function simultaneouslyUtilize convex optimum theory about cost function Alternative manner, is appreciated that:

$J^i\left(S^i\right)=\left\{\begin{array}{c}{\min }_{a^N\&Element;A\left(S^N\right)}{O}^N,i=N\\ {\min }_{a^N\&Element;A\left(S^i\right)}{O}^i+J^{i+1}\left(S^{i+1}\right),i<N\end{array}\right.,$

$A(S^i=<D^i,{\mathrm{\&epsiv;}}^i>)=\left\{\begin{array}{c}\{0,1\}\&times;\&lsqb;0,{\mathrm{\&epsiv;}}^i{T}_p^{-1}\&rsqb;,D^i=1\\ \left\{0\right\}\&times;\&lsqb;0,{\mathrm{\&epsiv;}}^i{T}_p^{-1}\&rsqb;,D^i=1\end{array}\right.,$

Finally, optimal solution can be expressed as

Embodiment two

So that algorithm is more nearly actual application, the situation of channel more can be embodied, closer to realistic channels.Aforementioned Can use on the basis of embodiment one.Specifically, including: energy acquisition model employing Bernoulli process, the most effectively.

Wherein: E represents basic energy unit,Represent the energy that the i-th time slot collects, then each time slot is averagely adopted Collect to energy can be expressed as:

$P_{ave}=E\&lsqb;E_H^i\&rsqb;=\frac{\mathrm{\&rho;}E}{2T_p}$

Wherein: ρ represents the parameter of Bernoulli process, 2T_pRepresent the length of a time slot；Accordingly, energy cause and effect limits Condition can be expressed as:

$\underset{k=1}{\overset{l}{\&Sigma;}}{p}_r^i{T}_p\&le;\underset{n=1}{\overset{l}{\&Sigma;}}{E}_H^k,\&ForAll;l\&Element;1,\mathrm{...},N.$

To put it more simply, we assume that the battery capacity of relaying is sufficiently large:

${\mathrm{\&epsiv;}}^{i+1}={({\mathrm{\&epsiv;}}^i-p_r^i{T}_p)}^{+}+E_H^{i+1},i-1,\mathrm{...},N-1,{\left(x\right)}^{+}=max\{x,0\},{\mathrm{\&epsiv;}}^1=E_H^1.$

Wherein, εⁱ⁺¹Represent after i+1 transmission time slot remaining energy, (x) in battery⁺Represent that x and 0 takes higher value Function, it is assumed that energy content of battery initial value is set as

Embodiment three

So that algorithm is more nearly actual application, we use the energy acquisition curve of a kind of simplification, in aforementioned reality Improve further on the basis of executing example one.Specifically,

Described step 2 includes:

Step 2.1: use Bernoulli process to judge whether i-th relaying is successfully decoded:

For reducing computational complexity, using whether Bernoulli process describes successfully decoded, wherein p is expressed as Bernoulli Jacob's mistake The parameter of journey, DⁱRepresent whether i & lt transmission is successfully decoded, Dⁱ=1 expression is successfully decoded, Dⁱ=0 expression decodes unsuccessfully.

The present invention uses Bernoulli Jacob's model simplification operand, more can embody the situation of channel, closer to realistic channels, innovation Point, Bernoulli Jacob's model parameter p is expressed as follows:

$p=\mathrm{Pr}\{{\mathrm{WT}}_p{\log }_2(1+{\mathrm{SNR}}_r^i)\&GreaterEqual;2T_{p}WR\}=\exp \{-\frac{(2^{2R}-1)N_0{\mathrm{Wd}}_1^{\mathrm{\&alpha;}}}{A_0{p}_{sr}{\mathrm{\&sigma;}}_{s,r}^2{d}_0^{\mathrm{\&alpha;}}}\}\mathrm{..}$

Step 2.2: we define link selection result r againⁱ∈{0,1},Wherein rⁱ=1 represents the i-th time slot relaying ginseng With the forwarding data that cooperate, rⁱ=0 represents that the i-th time slot relaying is not involved in cooperation, cellulor S retransmission data；Thus, we are again fixed The outage probability O of justice i & lt transmissionⁱAs follows:

Wherein

Step 2.3: the expression formula of final outage probability is as follows:

$O_R\left(p_r^i\right)=\mathrm{Pr}\{{\mathrm{WT}}_p{\log }_2<1+{\mathrm{SNR}}_{d_R}^i)\&GreaterEqual;2T_{p}WR\}=\left\{\begin{array}{c}1-\frac{{\mathrm{\&mu;}}_2{e}^{-{\mathrm{\&mu;}}_{1}x}-{\mathrm{\&mu;}}_1{e}^{-{\mathrm{\&mu;}}_{2}x}}{{\mathrm{\&mu;}}_2-{\mathrm{\&mu;}}_1},{\mathrm{\&mu;}}_1\&NotEqual;{\mathrm{\&mu;}}_2.\\ 1-(1+{\mathrm{\&mu;}}_1)e^{-{\mathrm{\&mu;}}_{1}x},{\mathrm{\&mu;}}_1={\mathrm{\&mu;}}_2,\end{array}\right.$

$O_D=\mathrm{Pr}\{{\mathrm{WT}}_p{\log }_2(1+{\mathrm{SNR}}_{d_D}^i)<2T_{p}WR\}=1-e^{-\frac{{\mathrm{\&mu;}}_{1}x}{2}}.$

WhereinRepresent cooperation probability, O_BRepresent recurrence probability.

Embodiment four

In order to improve improvement further, improving the operation efficiency of algorithm, the present invention uses convex optimization method to ask optimization Topic P1 converts, and the object function form after conversion solves and is more prone to, and computation complexity is low.Specifically, described step The 3 convex process also including optimization problem P1, have converted object function form, have solved and be more prone to, and computation complexity is low.

$P2:\underset{p_r^i,r^i}{\min }\frac{O^i+E_{D^k,E_H^k,\&ForAll;k>i}\&lsqb;\underset{k=i+1}{\overset{N}{\&Sigma;}}{O}^k|p_r^i,r^i\&rsqb;}{N-i+1},\&ForAll;i.$

$s.t.\underset{k=1}{\overset{i}{\&Sigma;}}{p}_r^k{T}_p\&le;\underset{n=1}{\overset{i}{\&Sigma;}}{E}_H^n,$

$p_r^i\&GreaterEqual;0,$

rⁱ∈{0,1},

$r^i\&le;D^i,\&ForAll;i\&Element;\{1,2,\mathrm{...},N\}.$

We analyze the variable of optimization problem P2: optimization furtherrⁱ, the object function of optimization problem is minimum ChangeThe constraints optimized is WhereinRepresent energy causality constraint.

Set:Then

$p_r^i=\underset{p_r^i\&Element;\&lsqb;0,\frac{{\mathrm{\&epsiv;}}^i}{T_p}\&rsqb;}{\arg \min }\frac{O_R\left(p_r^i\right)+E_{D^k,\&ForAll;k>i}\&lsqb;\underset{k=i+1}{\overset{N}{\&Sigma;}}{O}^k|p_r^i,r^i=1\&rsqb;}{N-i+1}=\underset{p_r^i\&Element;\&lsqb;0,\frac{{\mathrm{\&epsiv;}}^i}{T_p}\&rsqb;}{\arg \min }\frac{O_R\left(p_r^i\right)+p\underset{k=i+1}{\overset{N}{\&Sigma;}}{O}_R\left(f_{i,k}\right(p_r^i))}{N-i+1}.$

Thus it was found that once we are according to certain criterion, findThe functional relationship of k ＞ i, optimization is asked Topic will be drawn sword and solve.

Embodiment five

In order to improve improvement further, improving the operation efficiency of algorithm, the present invention, on the basis of embodiment four, proposes one The through-put power anticipation function in the future such as kind, for quickly realizing the resource distribution under this scene.Specifically, described step 4 is asked That inscribes solves through-put power anticipation function, the settings in the future such as employing

$f_{i,j}\left(p_r^i\right)=\frac{1}{T_p}(\frac{{\mathrm{\&epsiv;}}^i-p_r^i{T}_p}{N-i}+\mathrm{\&rho;}E).$

Embodiment six

With embodiment five, in order to improve improvement further, improving the operation efficiency of algorithm, the present invention is at the base of embodiment four On plinth, the reserved through-put power function of one is proposed, for quickly realizing the resource distribution under this scene.Specifically, described step Through-put power function is reserved in the employing that solves of rapid 4 problems, arranges

$f_{i,j}\left(p_r^i\right)=\mathrm{\&beta;}\frac{E\&lsqb;{\mathrm{\&epsiv;}}^j\&rsqb;}{T_p},$

$E\&lsqb;{\mathrm{\&epsiv;}}^j\&rsqb;=\left\{\begin{array}{c}{\mathrm{\&epsiv;}}^i-p_r^i{T}_p+\mathrm{\&rho;}E,j=i+1,\\ {\stackrel{\&OverBar;}{\mathrm{\&beta;}}}^{j-i-1}({\mathrm{\&epsiv;}}^i-p_r^i{T}_p)+\mathrm{\&rho;}E\left(\frac{1-{\stackrel{\&OverBar;}{\mathrm{\&beta;}}}^{j-i}}{\mathrm{\&beta;}}\right),j>i+1\end{array}\right.$

$\stackrel{\&OverBar;}{\mathrm{\&beta;}}=1-\mathrm{\&beta;}$

Wherein, β is as transmitting the percentage ratio used every time, and we, by making repeated attempts, arrange β=0.7 and have very Good convergence effect.

Embodiment seven

With embodiment five, in order to improve improvement further, improving the operation efficiency of algorithm, the present invention is at the base of embodiment four On plinth, one prediction energy acquisition curvilinear function is proposed, for quickly realizing the resource distribution under this scene.Specifically, institute That states step 4 problem solves employing prediction energy acquisition curvilinear function, arranges

$p_r^i=\left\{\begin{array}{cc}\frac{{\mathrm{\&epsiv;}}^i}{T_p},& {\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}E\\ \frac{{\mathrm{\&epsiv;}}^i+(N-i)\mathrm{\&rho;}E}{(N-i+1)T_p},& {\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}E\end{array}\right.\mathrm{..}$

For improving computational accuracy, optimizing systematic function, improve further, can arrange prediction energy acquisition curvilinear function is

$p_r^i=\left\{\begin{array}{c}\frac{{\mathrm{\&epsiv;}}^i}{T_p},{\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}Eor{O}_R\left(\frac{{\mathrm{\&epsiv;}}^i+(N-i)\mathrm{\&rho;}E}{(N-i+1)T_p}\right)\&GreaterEqual;{\mathrm{\&gamma;O}}_R\left(0\right)\\ \frac{{\mathrm{\&epsiv;}}^i+(N-i)\mathrm{\&rho;}E}{(N-i+1)T_p},{\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}E\end{array}\right.\mathrm{..}$

Embodiment eight

In order to improve improvement further, improving the operation efficiency of algorithm, the present invention proposes a kind of new optimization that solves and asks The thinking of topic P2, uses Lagrange multiplier method to go optimizing, and faster, algorithm complex is lower for speed.Specifically, described excellent Change problem P2 solve the Lagrange factor method that can use:

$L(r^i,p_r^i,{\mathrm{\&mu;}}_l,{\mathrm{\&lambda;}}_l)=\stackrel{\&OverBar;}{O}-{\mathrm{\&mu;}}_i(\underset{k=1}{\overset{i}{\&Sigma;}}{p}_r^k{T}_p-\underset{n=1}{\overset{i}{\&Sigma;}}{E}_H^n)-{\mathrm{\&lambda;}}_i(r^i-D^i),i\&Element;\{1,2,\mathrm{...},N\}$

Simultaneous againAnd with subgradient method iterative, wherein, l ∈ 1 ..., N}, μ_l,λ_lExpression is drawn The Ge Lang 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_l,λ_lIteration update method use Subgradient Algorithm, complexity is lower, more efficiently, described subgradient The iteration renewal equation of algorithm is

${\mathrm{\&mu;}}_l(n+1)={\&lsqb;{\mathrm{\&mu;}}_l\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&mu;}}_l}\left(n\right)(\underset{n=1}{\overset{i}{\&Sigma;}}{E}_H^n-\underset{k=1}{\overset{i}{\&Sigma;}}{p}_r^k{T}_p)\&rsqb;}^{+},i=1,2,\mathrm{...},N,l=1,2,\mathrm{...}N$

${\mathrm{\&lambda;}}_l(n+1)={\&lsqb;{\mathrm{\&lambda;}}_l\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&lambda;}}_l}\left(n\right)(D^i-r^i)\&rsqb;}^{+},i=1,2,\mathrm{...},N,l=1,2,\mathrm{...}N$

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

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

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

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. a federated resource distribution method based on real-time Communication for Power energy acquisition cellular network, it is characterised in that including:

Step 1: system scenarios analysis, problem describes；

Scene has the cellulor S of a stable power-supplying, the relay station R of an energy acquisition and a destinations traffic phone user For D, having direct path between cellulor S and phone user D, relay station R selects decoding to forward working method, a transmitting procedure Including N number of data block, cellulor S and relaying R takies equal bandwidth B, it is considered to cellulor S's under this scene and relay station R Federated resource assignment problem；

Assume that energy acquisition model uses Bernoulli process, sufficiently large for the battery capacity of energy acquisition, except being used for transmitting The energy consumed is ignored, and channel is block decline at a slow speed, and its response time is defined as T_P,Represent the i-th packet transmission Time signal source S and relay station R between channel coefficients,Signal source S and destination communication node D when representing the i-th packet transmission Between channel coefficients,Represent the channel coefficients between relay station R and destination communication node D when the i-th packet transmits；

Use half-duplex relay cooperative communication mode, comprise two time slots in the transmission time of each packet, in i-th data First time slot of packet transmission time, signal source S transport symbol s uses constant power p_s, s meets s～CN (0,1), relaying R and The reception signal of purpose communication node D is expressed as:

$y_{s,r}^i=A_0^{\frac{1}{2}}{d}_0^{\frac{\mathrm{\&alpha;}}{2}}{d}_1^{-\frac{\mathrm{\&alpha;}}{2}}{h}_{s,r}^i\sqrt{p_s}s+n_r^i,\&ForAll;i=1,...,N$

$y_{s,d}^i=A_0^{\frac{1}{2}}{d}_0^{\frac{\mathrm{\&alpha;}}{2}}{d}^{-\frac{\mathrm{\&alpha;}}{2}}{h}_{s,d}^i\sqrt{p_s}s+n_d^i,\&ForAll;i=1,...,N,$

Wherein: d₀And A₀Represent reference distance and reference power, the d of large scale decline respectively₁With d represent respectively cellulor S and in Continue the distance between distance and cellulor S and the destinations traffic user D between the R of station,WithRepresent i-th packet respectively The relay station R of transmission time and the reception noise of destinations traffic user D, α represents large scale fading factor,WithTable respectively Show relay station R and the channel gain of destinations traffic user D of i-th packet transmission time；

The signal to noise ratio of the first time slot relay reception is expressed asWherein N₀Represent normalized Noise power, W represents the bandwidth of distribution, second time slot, and destination communication node D uses high specific accepting method, and receive comes It is expressed as from the signal of relay station RWherein d₂WithRepresent relay station R respectively And the distance between destinations traffic user D and channel gain,Target when representing the second time slot of i-th packet transmission time The reception noise of communication user D, p_rRepresent the transmitting power of corresponding relay station R, receive from cellulor S and relay station R Signal to noise ratio be expressed as

If selection tie link, the signal of the cellulor S that destinations traffic user D receives is expressed as Accordingly, the signal to noise ratio of the cellulor S that destinations traffic user D receives is expressed as

Step 2: derivation average interrupt probability；

Described average interrupt probability isWherein ΟⁱRepresent the outage probability of i & lt transmission；

Step 3: optimization problem is summed up；

Under the conditions of above-mentioned supposed premise and constraint, sum up optimization problem as follows:

$\begin{array}{cc}P1:& \underset{g}{\min }{E}_{D^i,E_H^i,\&ForAll;i}^g\&lsqb;\frac{\underset{i=1}{\overset{N}{\&Sigma;}}{O}^i\left(g\right(S^i,i\left)\right)}{N}\&rsqb;\\ & s.t.\underset{k=1}{\overset{i}{\&Sigma;}}{p}_r^k{T}_p\&le;\underset{n=1}{\overset{i}{\&Sigma;}}{E}_H^n,\\ & p_r^i\&GreaterEqual;0,\\ & r^i\&Element;\{0,1\},\\ & r^i\&le;D^i,\&ForAll;i\&Element;\{1,2,\mathrm{...},N\}.\end{array}$

WhereinRepresenting energy causality constraint, the object function of optimization problem is to minimize The variable optimized is g (Sⁱ, i), the constraints of optimization is

Step 4: duty Optimization；

The optimized variable of optimization problem is model selection factor r of transmission every timeⁱAnd corresponding power allocation scheme rⁱDiscrete variable span for 1,0},Continuous variable, span is more than or equal to 0, uses cost letter The method of number, defines Sⁱ=< Dⁱ,εⁱ> represent state set,Expression behavior collection, defines cost function simultaneouslyUtilize convex optimum theory about the alternative manner of cost function, can obtain :

$J^i\left(S^i\right)=\left\{\begin{array}{c}{\min }_{a^N\&Element;A\left(S^N\right)}{O}^N,i=N\\ {\min }_{a^N\&Element;A\left(S^i\right)}{O}^i+J^{i+1}\left(S^{i+1}\right),i<N\end{array}\right.,$

$A(S^i=<D^i,{\mathrm{\&epsiv;}}^i>)=\left\{\begin{array}{c}\{0,1\}\&times;\&lsqb;0,{\mathrm{\&epsiv;}}^i{T}_p^{-1}\&rsqb;,D^i=1\\ \left\{0\right\}\&times;\&lsqb;0,{\mathrm{\&epsiv;}}^i{T}_p^{-1}\&rsqb;,D^i=1\end{array}\right.,$

Finally, optimal solution can be expressed as

Federated resource distribution method the most according to claim 1, it is characterised in that described energy acquisition model uses uncle to exert Profit process, including:

Wherein: E represents basic energy unit,Represent the energy that the i-th time slot collects, the energy that each time slot averaged acquisition arrives Amount can be expressed as:Wherein ρ represents the parameter of Bernoulli process, 2T_pRepresent the length of a time slot, Assuming that the battery capacity of relaying is sufficiently large, corresponding energy cause and effect restrictive condition can be expressed as:

$\underset{i=1}{\overset{l}{\&Sigma;}}{p}_r^i{T}_p\&le;\underset{k=1}{\overset{l}{\&Sigma;}}{E}_H^k,\&ForAll;l=1,...,N.$

${\mathrm{\&epsiv;}}^{i+1}={({\mathrm{\&epsiv;}}^i-p_r^i{T}_p)}^{+}+E_H^{i+1},i-1,...,N-1,{\left(x\right)}^{+}=max\{x,0\},{\mathrm{\&epsiv;}}^1=E_H^1.$

Wherein: εⁱ⁺¹Represent after i+1 transmission time slot remaining energy, (x) in battery⁺Represent that x and 0 takes the letter of higher value Number, it is assumed that energy content of battery initial value is set as

Federated resource distribution method the most according to claim 1, it is characterised in that described step 2 includes:

Step 2.1: use Bernoulli process to judge whether i-th relaying is successfully decoded:

Wherein Dⁱ=1 represents the decoding of i & lt transmission success, Dⁱ=0 represents that i & lt transmission decodes unsuccessfully,

Step 2.2: redefine selection result r of link iⁱ∈ { 0,1}, wherein rⁱ=1 represents that the i-th time slot relaying participates in cooperation Forward data, rⁱ=0 represents that the i-th time slot relaying is not involved in cooperation, cellulor S retransmission data, redefines in i & lt transmission Disconnected probability ΟⁱForWherein

Step 2.3: the outage probability expression formula after derivation correction strategy；

$O_R\left(p_r^i\right)=\mathrm{Pr}\{{\mathrm{WT}}_P{\log }_2(1+{\mathrm{SNR}}_{d_R}^i)\&GreaterEqual;2T_{p}WR\}=\left\{\begin{array}{c}1-\frac{{\mathrm{\&mu;}}_2{e}^{-{\mathrm{\&mu;}}_{1}x}-{\mathrm{\&mu;}}_1{e}^{-{\mathrm{\&mu;}}_{2}x}}{{\mathrm{\&mu;}}_2-{\mathrm{\&mu;}}_1},{\mathrm{\&mu;}}_1\&NotEqual;{\mathrm{\&mu;}}_2\\ 1-(1+{\mathrm{\&mu;}}_1)e^{-{\mathrm{\&mu;}}_{1}x},{\mathrm{\&mu;}}_1={\mathrm{\&mu;}}_2,\end{array}\right..$

$O_D=\mathrm{Pr}\{{\mathrm{WT}}_p{\log }_2(1+{\mathrm{SNR}}_{d_D}^i)<2T_{p}WR\}=1-e^{-\frac{{\mathrm{\&mu;}}_{1}x}{2}}.$

WhereinΟ_RRepresent cooperation probability, Ο_BRepresent recurrence probability.

Federated resource distribution method the most according to claim 1, it is characterised in that described step 3 also includes that optimization is asked The convex optimization of topic P1 processes, and object function form is converted into:

$\begin{array}{cc}P2:& \underset{p_r^i,r^i}{\min }\frac{O^i+E_{D^k,E_H^k,\&ForAll;k>i}\&lsqb;\underset{k=i+1}{\overset{N}{\&Sigma;}}{O}^k|p_r^i,r^i\&rsqb;}{N-i+1},\&ForAll;i.\\ & s.t.\underset{k=1}{\overset{i}{\&Sigma;}}{p}_r^k{T}_p\&le;\underset{n=1}{\overset{i}{\&Sigma;}}{E}_H^n,\\ & p_r^i\&GreaterEqual;0,\\ & r^i\&Element;\{0,1\},\\ & r^i\&le;D^i,\&ForAll;i\&Element;\{1,2,\mathrm{...},N\}.\end{array}$

WhereinRepresenting energy causality constraint, the variable of optimization isObject function is to minimizeThe constraints optimized is Set:Then SetFunctional relationship, optimization problem can obtain optimal solution.

Federated resource distribution method the most according to claim 4, it is characterised in that described step 4 problem solving employing etc. Through-put power anticipation function in the future, is arranged

Federated resource distribution method the most according to claim 4, it is characterised in that described step 4 problem solving is adopted With reserved through-put power function, arrange:

$f_{i,j}\left(p_r^i\right)=\mathrm{\&beta;}\frac{E\&lsqb;{\mathrm{\&epsiv;}}^j\&rsqb;}{T_p},E\&lsqb;{\mathrm{\&epsiv;}}^j\&rsqb;=\left\{\begin{array}{c}{\mathrm{\&epsiv;}}^i-p_r^i{T}_p+\mathrm{\&rho;}E,j=i+1,\\ {\stackrel{\&OverBar;}{\mathrm{\&beta;}}}^{j-i-1}({\mathrm{\&epsiv;}}^i-p_r^i{T}_p)+\mathrm{\&rho;}E\left(\frac{1-{\stackrel{\&OverBar;}{\mathrm{\&beta;}}}^{j-i}}{\mathrm{\&beta;}}\right),j>i+1\end{array}\right.,\stackrel{\&OverBar;}{\mathrm{\&beta;}}=1-\mathrm{\&beta;}.$

Federated resource distribution method the most according to claim 4, it is characterised in that described step 4 problem solving is adopted With prediction energy acquisition curvilinear function, arrange:

$p_r^i=\left\{\begin{array}{cc}\frac{{\mathrm{\&epsiv;}}^i}{T_p},& {\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}E\\ \frac{{\mathrm{\&epsiv;}}^i+(N-i)\mathrm{\&rho;}E}{(N-i+1)T_p},& {\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}E\end{array}\right.\mathrm{..}$

Federated resource distribution method the most according to claim 7, it is characterised in that described step 4 problem solving is adopted With prediction energy acquisition curvilinear function, arrange:

$p_r^i=\left\{\begin{array}{c}\frac{{\mathrm{\&epsiv;}}^i}{T_p},{\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}Eor{O}_R\left(\frac{{\mathrm{\&epsiv;}}^i+(N-i)\mathrm{\&rho;}E}{(N-i+1)T_p}\right)\&GreaterEqual;{\mathrm{\&gamma;O}}_R\left(0\right)\\ \frac{{\mathrm{\&epsiv;}}^i+(N-i)\mathrm{\&rho;}E}{(N-i+1)T_p},{\mathrm{\&epsiv;}}^i\&le;\mathrm{\&rho;}E\end{array}\right.\mathrm{..}$

Federated resource distribution method the most according to claim 4, it is characterised in that asking of described step 3 optimization problem Solve and use method of Lagrange multipliers, including:

First the Lagrangian Form of optimization problem P2 is write out:

$L(r^i,p_r^i,{\mathrm{\&mu;}}_l,{\mathrm{\&lambda;}}_l)=\stackrel{\&OverBar;}{O}-{\mathrm{\&mu;}}_l(\underset{k=1}{\overset{i}{\&Sigma;}}{p}_r^k{T}_p-\underset{n=1}{\overset{i}{\&Sigma;}}{E}_H^n)-{\mathrm{\&lambda;}}_l(r^i-D^i),i\&Element;\{1,2,...,N\}$

Simultaneous againAnd with subgradient method iterative, wherein, l ∈ 1 ..., N}, μ_l,λ_lRepresent that glug is bright Day factor.

Federated resource distribution method the most according to claim 9, it is characterised in that the glug of described optimization problem P2 is bright Lagrange factor μ in day form_l,λ_lIteration update method use Subgradient Algorithm, the iteration of described Subgradient Algorithm is more New equation is:

${\mathrm{\&mu;}}_l(n+1)={\&lsqb;{\mathrm{\&mu;}}_l\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&mu;}}_l}\left(n\right)(\underset{n=1}{\overset{i}{\&Sigma;}}{E}_H^n-\underset{k=1}{\overset{i}{\&Sigma;}}{p}_r^k{T}_p)\&rsqb;}^{+},i=1,2,...,N,l=1,2,...N$

${\mathrm{\&lambda;}}_l(n+1)={\&lsqb;{\mathrm{\&lambda;}}_l\left(n\right)-{\mathrm{\&alpha;}}_{{\mathrm{\&lambda;}}_l}\left(n\right)(D^i-r^i)\&rsqb;}^{+},i=1,2,...,N,l=1,2,...N$

Wherein μ_l(n),λ_lN () represents the Lagrange factor of nth iteration respectively,Represent corresponding changing respectively Riding instead of walk length, described iteration step length may be arranged such that