CN105704721A - D2D-P multiplexing cellular network communication method capable of increasing frequency spectrum utilization rate - Google Patents

Abstract

The invention relates to a D2D-P multiplexing cellular network communication method capable of increasing a frequency spectrum utilization rate. The method comprises the following steps of S1, establishing a cellular network; S2, according to Hk,m, using an interference alignment method to acquire a solution set of an N row and d column precoding matrix Fm of each D2D-Pm; S3, calculating an interference matrix Hk,mFm of the D2D-Pm to CUk and designing an N row and d column cellular user postposition coding matrix Wk orthogonal to the Hk,mFm; S4, using the interference alignment method based on a minimum mean square error to acquire transmitter precoding matrixes Fm and receiver postposition coding matrixes Gm of all the D2D-Pm and carrying out communication; S5, setting a SINR threshold omegath of the D2D-Pm and determining whether the D2D-Pm is connected to the cellular network; S6, using an improved water injection power distribution method to send power to all the D2D-Pm distribution; and S7, according to results of the step S2 to the step S6, carrying out communication. Compared to the prior art, by using the method, interferences of the D2D-Ps on a cellular user are aligned to a cellular user position; a communication priority of the cellular user is guaranteed; mutual interference among the D2D-Ps is controlled and a D2D-P sum rate is increased.

Description

A kind of D2D-P multiplexing cellular network communication method improving the availability of frequency spectrum

Technical field

The present invention relates to a kind of cellular network communication method, especially relate to a kind of D2D-P multiplexing cellular network communication method improving the availability of frequency spectrum。

Background technology

Along with the fast development of radio communication, how when limited frequency spectrum resource, maximized utilize frequency spectrum resource, be always up the focus of research。In cellular networks, (D2D-P) is joined its frequency spectrum resource of multiplexing in cellular network by equipment user by equipment, the availability of frequency spectrum can be improved, but phone user can be brought inevitable interference, affect the performance indications such as phone user's transfer rate。

And D2D-P not high for cellular network frequency spectrum resource utilization rate is to problems such as cellular network interfere, DaquanFeng et al. (DaquanFeng, LuLu, YiYuan-Wu, GeoffreyYeLi, GangFeng, ShaoqianLi.Device-to-DeviceCommunicationsUnderlayingCell ularNetworks, IEEETRANSACTIONSONCOMMUNICATIONS, VOL.61, NO.8, pp.3541-3551, AUGUST2013) following method is proposed: by setting threshold value in advance, select satisfactory D2D-P, then adopt and distribute power to potential phone user and satisfactory D2D-P, finally, potential phone user selects suitable user coordinate with satisfactory D2D-P, by selecting the D2D-P of OK range, cellular network coverage, active phone user and the number of D2D-P, the throughput gain of maximization system and access rate。This kind of method is limited in cellular network up-link, only ensure that the communication requirement of partial cell user, it does not have the concrete D2D-P that eliminates adds the interference etc. that cellular network brings。

LuYang et al. (LuYang, WeiZhang, ShiJin, " InterferenceAlignmentinDevice-to-DeviceLANUnderlayingCel lularNetworks ", IEEETRANSACTIONSONWIRELESSCOMMUICATIONS, VOL.14, NO.7, PP.3715-3723, JULY, 2015.) adopt interference alignment algorithm, solve the interference that D2D-P Reusespectrum resource causes, method is as follows: (1) is not when the up-link of base station is occupied by phone user completely, the interference of phone user is snapped in these idle links by D2D-P, so that phone user avoids interference；(2) when base station uplink is fully occupied, D2D-P can occupy part up-link, now arranges interference threshold, controls under threshold value by the link interference occupied, and controls the interference to phone user with this。Although the method eliminates the D2D-P interference to cellular subscriber communications link, but when phone user is more, the interference that D2D-P causes can not be completely eliminated, and D2D-P quantity and the performance indications such as transfer rate thereof after accessing do not account for, and only considered D2D-P and access the outage probability of cellular network。

JiamoJiang et al. (JiamoJiang, MugenPeng, WenboWang, KechengZhang, " Energyefficiencyoptimizationbasedoninterferencealignment fordevice-todeviceMIMOdownlinkunderlayingcellularnetwork ", IEEEGlobecom2013Workshop-InternationalWorkshoponDevice-t o-Device (D2D) CommnicationWithandWithoutInfrastructure, PP.585-590, 2013.) use interference alignment algorithm to eliminate cellular network dl interference, method is as follows: first have employed coding techniques and eliminates the interference that D2D-P addition cellular network brings。Then use interference alignment algorithm that the interference between phone user is snapped to specific signal dimension, it is ensured that interference-free between phone user；Finally, optimize rearmounted coding and linear predictive coding, maximize the energy efficiency of D2D-P。Take into account D2D-P and add the interference between the cellular network interference and the phone user that bring, but D2D-P singal reporting code such as handling capacity, the performance indications such as bit error rate do not account for。

Summary of the invention

Defect that the purpose of the present invention is contemplated to overcome above-mentioned prior art to exist and a kind of D2D-P multiplexing cellular network communication method disturbing low, D2D-P and the high raising availability of frequency spectrum of speed is provided。

The purpose of the present invention can be achieved through the following technical solutions:

A kind of D2D-P multiplexing cellular network communication method improving the availability of frequency spectrum, it is characterised in that comprise the following steps:

S1, sets up cellular network, and described cellular network includes cellular basestation, K phone user CU_kAnd M to equipment user to D2D-P_m, k=1,2 ..., K, m=1,2 ..., M, described D2D-P_mIncluding transmitter and receiver, described phone user, transmitter and receiver are provided with N number of transmitting antenna number and N number of reception antenna number, the signal vector Z that transmitter sends to receiver respectively_mFor d row, 1 row, D2D-P_mTo CU_kChannel matrix H_k,mFor N row, N row；

S2, according to H_k,m, utilize interference alignment method, try to achieve each D2D-P_mN row, d row pre-coding matrix F_mDisaggregation；

S3, calculates each D2D-P_mTo CU_kInterference matrix H_k,mF_m, design is orthogonal to H_k,mF_mN row, d row the rearmounted encoder matrix W of phone user_k, k=1,2 ..., K；

S4, utilizes based on the interference alignment method minimizing mean square error, tries to achieve all D2D-P_mTransmitter pre-coding matrix F_mEncoder matrix G rearmounted with receiver_m, and carry out cellular network communication, to obtain each phone user CU_kSignal to Interference plus Noise Ratio (SINR)；

S5, arranges D2D-P_mSINR threshold value ω_thIf, D2D-P_mSINR less than ω_th, then by this D2D-P_mAccess cellular network and β is set_mValue is 1, otherwise arranges β_mValue is 0；

S6, according to transmitter pre-coding matrix F_mAnd β_m, utilize the water injection power improved distribution method to all D2D-P_mDistribution transmit power P_m；

S7, according to the matrix calculation result of step S2～S6 and power distribution result, carries out cellular network communication。

Described step S2 is particularly as follows: by all of D2D-P_mTo CU_kInterference matrix H_k,mF_mSnap to this CU_kOn the same signal subspace of receiving terminal, formula specific as follows:

span(H_k,1F₁)=...=span (H_k,mF_m)=...=span (H_k,MF_M)

Wherein, k=1,2 ..., K, the subspace that span (A) representing matrix A column vector is opened。

Described step S4 specifically includes following steps:

S41, calculates the emitted machine pre-coding matrix F that receiver receives_mThe signal level S processed_m, and S_mThe rearmounted encoder matrix G of the machine that is received_mThe signal level obtained after process

S42, defines all D2D-P_mSignal level mean square error and ε_MSEAs follows:

${\mathrm{\ϵ}}_{MSE}=\underset{m=1}{\overset{M}{\Σ}}E\left\{\right||\stackrel{\‾}{S_m}-S_m|{|}^2\}=\underset{m=1}{\overset{M}{\Σ}}E\left\{\right||{G_m}^H{S}_m-S_m|{|}^2\}$

Lagrange's method of multipliers is utilized to solve optimization problem as follows:

$\begin{array}{cc}\underset{G_m,F_m}{\min }& {\mathrm{\ϵ}}_{MSE}\\ s.t.& \left|\right|F_m|{|}_F^2=P_m\end{array}$

Wherein, E{ } represent mathematic expectaion, | | X | |²The norm squared of representing matrix X,The F norm of representing matrix X, P_mFor D2D-P_mTransmit power,P is the total transmit power of all of D2D-P。

In described step S41, emitted machine pre-coding matrix F_mThe signal level S processed_mSpecifically it is calculated as follows formula:

$S_m=\sqrt{P_m}{H}_{m,m}{F}_m{Z}_m+\underset{j=1,j\≠m}{\overset{M}{\Σ}}\sqrt{P_j}{H}_{m,j}{F}_j{Z}_j+n_m$

Wherein, P_mIt is D2D-P_mThe transmit power of transmitter, H_m,mFor D2D-P_mThe N row of transmitted from transmitter to receiver, N row channel matrix, H_m,jFor D2D-P_jTransmitter is to D2D-P_mThe N row of receiver, N row channel matrix, n_mFor D2D-P_mThe environment noise that receiver receives。

Described step S42 specifically includes following steps:

S420, introduces lagrangian multiplier_m, obtain Lagrangian:

$L(F_m,G_m,{\mathrm{\λ}}_m)={\mathrm{\ϵ}}_{MSE}+{\mathrm{\λ}}_m\left(\right|\left|F_m\right|{|}_F^2-P_m)$

Wherein, ε_MSESee the definition of S42。

S421, respectively to G in Lagrangian_m、F_mSeek partial derivative, and to make partial derivative be zero, obtain equation (1) (2):

$F_m={(\underset{j=1}{\overset{M}{\Σ}}{H}_{j,m}{G}_j{G}_j^H{H}_{j,m}^H+{\mathrm{\λ}}_{m}I)}^{-1}{H}_{m,m}^H{G_m}^H---\left(1\right)$

$G_m={F_m}^H{H_{m,m}}^H{(\underset{j=1}{\overset{M}{\Σ}}{H}_{m,j}{F}_j{F}_j^H{\left(H_{m,j}\right)}^H+{\mathrm{\σ}}^{2}I)}^{-1}---\left(2\right)$

Wherein, m=1,2 ..., M, σ²For environment noise n_mVariance, and E{n_mn_m ^H}=σ²I, I are N rank unit matrixs；

S422, initializes pre-coding matrix F_mFor random matrix, obtain the random matrix of N row, d row；

S423, through type (2) calculates and obtains rearmounted encoder matrix G_m；

S424, substitutes into power limitation condition by formula (1)Try to achieve λ_m(λ_m>=0), λ_mSubstitution formula (1), updates F_m；

S425, calculates mean square error and ε_MSE；

S426, repeats step S423～S425, until ε_MSEConvergence, obtains G_mAnd F_m。

In described S5, D2D-P_mCommunication SINR threshold value ω_thFor: the phone user participating in communication measures the Signal to Interference plus Noise Ratio SINR of local reception signal, and issues cellular basestation by feedback channel, cellular basestation wherein minimum SINR is set to D2D-P communication SINR threshold value ω_th。Local reception signal includes all D2D-P interference to phone user, also has base station to be sent to the signal of all phone users；Certain phone user can receive base station and be sent to the signal of all phone users, wherein comprises desired signal and unwanted signal, and unwanted signal is considered as interference。

In described S6, the water injection power distribution method of described improvement solves D2D-P_mTransmit power P_mComputational methods be:

P_m=β_m(μ/γ_m-1)⁺

Wherein, μ is the level of water filling, meetsγ_mFor D2D-P_mChannel gain,P is the total transmit power of all of D2D-P, function (x)⁺=max (x, 0) represents the higher value in treating excess syndrome number x and 0。

Compared with prior art, the invention have the advantages that

(1) interference of phone user is snapped to by D2D-P phone user place, adopts rearmounted coding techniques to eliminate the D2D-P interference to phone user, it is ensured that the communication priority of phone user。

(2) adopt based on minimizing mean square error interference alignment techniques, process the interference between D2D-P, calculate the encoder matrix minimizing D2D-P mean square error, control interfering between D2D-P。

(3) judge whether D2D-P accesses cellular network by hard decision function, utilize the water injection power distribution method improved, to the D2D-P adding network_mDistribution transmit power P_m, improve D2D-P and speed。

Accompanying drawing explanation

Fig. 1 is phone user and equipment user's mutual interference schematic diagram under cellular network of the present invention；

Fig. 2 is overall flow figure of the present invention；

Fig. 3 is the inventive method detail flowchart。

Detailed description of the invention

Below in conjunction with the drawings and specific embodiments, the present invention is described in detail。The present embodiment is carried out premised on technical solution of the present invention, gives detailed embodiment and concrete operating process, but protection scope of the present invention is not limited to following embodiment。

Embodiment

Fig. 1 is the schematic diagram that in cellular network, equipment user couple and phone user are coexisted by equipment, this network packet is containing a cellular basestation, base station is LTE-A (LongTermEvolution-Advanced, long-term evolution upgrading version) evolved base station (evolvedNodeBs) of standard, and cover phone user and equipment to equipment user couple comprehensively。In its coverage, random distribution K phone user and M to D2D-P, and wherein, D2D-P is divided into transmitter and receiver。Setting phone user, D2D-P transmitting antenna number, reception antenna number are 2, have 2 phone users and 2 couples of D2D-P, D2D-P can obtain complete channel condition information with phone user in cellular basestation coverage。

Fig. 2 is a kind of D2D-P multiplexing cellular network communication method flow diagram improving the availability of frequency spectrum, comprises the following steps:

S1, sets up cellular network, and described cellular network includes cellular basestation, K phone user CU_kAnd M to equipment user to D2D-P_m, k=1,2 ..., K, m=1,2 ..., M, described D2D-P_mIncluding transmitter and receiver, phone user, transmitter and receiver transmitting antenna number be N, reception antenna number is N, transmitter is to the Z that the signal vector that receiver sends is d row, 1 row_m, D2D-P_mTo CU_kChannel matrix be N row, N row H_k,m；

S2, according to H_k,m, utilize interference alignment method, try to achieve each D2D-P_mN row, d row pre-coding matrix F_mDisaggregation；

S3, calculates D2D-P_mTo CU_kInterference matrix H_k,mF_m, design is orthogonal to H_k,mF_mN row, d row the rearmounted encoder matrix W of phone user_k；

S4, utilizes based on the interference alignment method minimizing mean square error, tries to achieve all D2D-P_mTransmitter pre-coding matrix F_mEncoder matrix G rearmounted with receiver_m, start cellular network and communicate；

S5, arranges D2D-P_mSINR threshold value ω_thIf, D2D-P_mSINR less than ω_th, then by this D2D-P_mAccess cellular network；β is set_mRepresent D2D-P_mWhether access cellular network, access then β_mValue is 1, otherwise β_mValue is 0；

S6, according to transmitter pre-coding matrix F_mAnd β_m, utilize the water injection power improved distribution method to all D2D-P_mDistribution transmit power P_m；

S7, according to the matrix calculation result of step S2～S6 and power distribution result, carries out cellular network communication。

Described step S2 is particularly as follows: by all of D2D-P_mTo CU_kInterference matrix H_k,mF_mSnap to this CU_kOn the same signal subspace of receiving terminal, formula specific as follows:

span(H_k,1F₁)=...=span (H_k,mF_m)=...=span (H_k,MF_M)

Wherein, k=1,2 ..., K, the subspace that span (A) representing matrix A column vector is opened。

Described step S4 specifically includes following steps:

S41, calculates the emitted machine pre-coding matrix F that receiver receives_mThe signal level S processed_mAnd the rearmounted encoder matrix G of the machine that is received_mThe signal level processed

S42, defines all D2D-P_mSignal level mean square error and ε_MSEAs follows:

${\mathrm{\ϵ}}_{MSE}=\underset{m=1}{\overset{M}{\Σ}}E\left\{\right||\stackrel{\‾}{S_m}-S_m|{|}^2\}=\underset{m=1}{\overset{M}{\Σ}}E\left\{\right||{G_m}^H{S}_m-S_m|{|}^2\}$

Lagrange's method of multipliers is utilized to solve optimization problem as follows:

$\begin{array}{cc}\underset{G_m,F_m}{\min }& {\mathrm{\ϵ}}_{MSE}\\ s.t.& \left|\right|F_m|{|}_F^2=P_m\end{array}$

Wherein, E represents expectation,The F norm of representing matrix X, P_mFor D2D-P_mTransmit power,P is the total transmit power of all of D2D-P。

In described step S41, emitted machine pre-coding matrix F_mThe signal level S processed_mSpecifically it is calculated as follows formula:

$S_m=\sqrt{P_m}{H}_{m,m}{F}_m{Z}_m+\underset{j=1,j\≠m}{\overset{M}{\Σ}}\sqrt{P_j}{H}_{m,j}{F}_j{Z}_j+n_m$

Wherein, P_mIt is D2D-P_mThe transmit power of transmitter, H_m,mFor D2D-P_mThe N row of transmitted from transmitter to receiver, N row channel matrix, H_m,jFor D2D-P_jTransmitter is to D2D-P_mThe N row of receiver, N row channel matrix, n_mFor D2D-P_mThe environment noise that receiver receives。

Described step S42 specifically includes following steps:

S420, introduces lagrangian multiplier_m, obtain Lagrangian:

$L(F_m,G_m,{\mathrm{\λ}}_m)={\mathrm{\ϵ}}_{MSE}+{\mathrm{\λ}}_m\left(\right|\left|F_m\right|{|}_F^2-P_m)$

Wherein, ε_MSESee the definition of S42。

S421, respectively to G in Lagrangian_m、F_mSeek partial derivative, and to make partial derivative be zero, obtain equation (1) (2):

$F_m={(\underset{j=1}{\overset{M}{\Σ}}{H}_{j,m}{G}_j{G}_j^H{H}_{j,m}^H+{\mathrm{\λ}}_{m}I)}^{-1}{H}_{m,m}^H{G_m}^H---\left(1\right)$

$G_m={F_m}^H{H_{m,m}}^H{(\underset{j=1}{\overset{M}{\Σ}}{H}_{m,j}{F}_j{F}_j^H{\left(H_{m,j}\right)}^H+{\mathrm{\σ}}^{2}I)}^{-1}---\left(2\right)$

Wherein, m=1,2 ..., M, σ²For environment noise n_mVariance, and E{n_mn_m ^H}=σ²I, I are N rank unit matrixs；

S422, initializes pre-coding matrix F_mFor random matrix, obtain the random matrix of N row, d row；

S423, through type (2) calculates and obtains rearmounted encoder matrix G_m；

S424, substitutes into power limitation condition by formula (1)Try to achieve λ_m, λ_m>=0, λ_mSubstitution formula (1), updates F_m；

S425, calculates mean square error and ε_MSE；

S426, repeats step S423～S425, until ε_MSEConvergence, obtains G_mAnd F_m。

In described S5, D2D-P_mCommunication SINR threshold value ω_thFor: the phone user participating in communication measures the Signal to Interference plus Noise Ratio SINR of local reception signal, and issues cellular basestation by feedback channel, cellular basestation wherein minimum SINR is set to D2D-P communication threshold ω_th, local reception signal includes all D2D-P interference to phone user, also has base station to be sent to the signal of all phone users；Certain phone user can receive base station and be sent to the signal of all phone users, wherein comprises desired signal and unwanted signal, and unwanted signal is considered as interference。

In described S6, P_mCalculating formula is:

P_m=β_m(μ/γ_m-1)⁺

Wherein, μ is the level of water filling, meetsγ_mFor D2D-P_mChannel gain,P is the total transmit power of all of D2D-P, function (x)⁺=max (x, 0) represents the higher value in treating excess syndrome number x and 0。

This communication means is applied to the cellular network shown in Fig. 1, and step is as follows:

(1) setting communication scenes there is a cellular basestation (BS₀), 2 phone user (CU₁、CU₂) and 2 couples of D2D-P；

Phone user, D2D-P all adopt multiple antennas, and phone user, the transmitting of D2D-P, reception antenna number are all set to 2；

D2D-P_mMiddle transmitter is sent to the matrix Z that the signal vector of receiver is 2 row, 1 row_m, m=1 or 2；

(2) note D2D-P_mTo CU_kThe channel matrix of (k=1 or 2) is the H of 2 row, 2 row_k,m, D2D-P_mOn pre-coding matrix be 2 row, 2 row F_m。By D2D-P_m(m=1,2) to CU_kInterference H_k,mF_mSnap to CU_kOn the same signal subspace of receiving terminal, as follows:

span(H_1,1F₁)=span (H_1,2F₂)

span(H_2,1F₁)=span (H_2,2F₂)

Wherein, the subspace that span (A) representing matrix A column vector expands；F can be tried to achieve by two above-mentioned constraintss_mDisaggregation；

(3) note CU_k(k=1 or 2) receives the W that the rearmounted encoder matrix of signal is 2 row, 2 row_k。Design rearmounted encoder matrix W_kIt is orthogonal to D2D-P_mTo CU_kInterference channel matrix H_k,mF_m(m=1 or 2), as follows:

W_k=Null (H_k,mF_m), m=1,2

Wherein, B=Null (A), representing matrix B is orthogonal to matrix A；

(4) note D2D-P_mThe rearmounted encoder matrix of receiving terminal is the G of 2 row, 2 row_m。Adopt based on minimizing mean square error interference alignment algorithm, obtain minimizing D2D-P_mThe rearmounted encoder matrix G of (m=1 or 2) mean square error₁, G₂With pre-coding matrix F₁, F₂, step is as follows:

A, respectively to G in Lagrangian_m、F_m(m=1 or 2) seeks partial derivative, and to make their partial derivative be zero, obtains equation (1) (2):

$F_m={(\underset{j=1}{\overset{M}{\Σ}}{H}_{j,m}{G}_j{G}_j^H{H}_{j,m}^H+{\mathrm{\λ}}_{m}I)}^{-1}{H}_{m,m}^H{G_m}^H---\left(1\right)$

$G_m={F_m}^H{H_{m,m}}^H{(\underset{j=1}{\overset{M}{\Σ}}{H}_{m,j}{F}_j{F}_j^H{\left(H_{m,j}\right)}^H+{\mathrm{\σ}}^{2}I)}^{-1}---\left(2\right)$

Wherein, m=1 or 2, σ²For environment noise n_mVariance；

B, initialization pre-coding matrix F_mIt it is the random matrix of 2 row, 2 row；

C, by formula (2) calculate obtain rearmounted encoder matrix G_m；

D, by (1) formula substitute into power limitation conditionTry to achieve λ_m(λ_m>=0), λ_mSubstitute into formula (1), update F_m；

E, calculating mean square error ε_MSE；

F, repetition step c, d, e, until ε_MSEConvergence, obtains G_m, F_mIt is and is solved；

(5) according to pre-coding matrix F_m(m=1 or 2), calculates the channel gain γ of D2D-Pm_m, as follows:

${\mathrm{\γ}}_m=\frac{1}{\left|\right|F_m|{|}^2},m=1,2$

(6) phone user participating in communication measures the Signal to Interference plus Noise Ratio (SINR) of local reception signal, and issues BS by feedback channel₀, by BS₀Wherein minimum SINR is set to D2D-P communication threshold ω_th。Then hard decision is adopted to judge D2D-P_mWhether access cellular network, use symbol beta_mRepresent, as follows:

${\mathrm{\&beta;}}_m=\left\{\begin{array}{c}1,{\mathrm{SINR}}_{D2Dm}<{\mathrm{\&omega;}}_{th}\\ 0,{\mathrm{SINR}}_{D2Dm}\&GreaterEqual;{\mathrm{\&omega;}}_{th}\end{array}\right.,m=1,2$

Wherein, SINR_D2DmRepresent D2D-P_mSINR value；β_mValue sets to 0, and represents D2D-P_mCellular network can not be accessed；β_mValue puts 1, represents D2D-P_mCellular network can be accessed；

(7) according to γ_mAnd β_m, according to the water injection power allocation algorithm distribution D2D-P improved_mTransmit power P_m, as follows:

P_m=β_m(μ/γ_m-1)⁺, m=1,2

Wherein, μ is the level of water filling, meetsFunction (x)⁺=max (x, 0) represents the maximum operation for the treatment of excess syndrome number x and 0。

Claims (7)

1. the D2D-P multiplexing cellular network communication method improving the availability of frequency spectrum, it is characterised in that comprise the following steps:

S1, sets up cellular network, and described cellular network includes cellular basestation, K phone user CU_kAnd M to equipment to equipment user to D2D-P_m, k=1,2 ..., K, m=1,2 ..., M, described D2D-P_mIncluding transmitter and receiver, described phone user, transmitter and receiver are provided with N number of transmitting antenna number and N number of reception antenna number, the signal vector Z that transmitter sends to receiver respectively_mFor d row, 1 row, D2D-P_mTo CU_kChannel matrix H_k,mFor N row, N row；

S2, according to H_k,m, utilize interference alignment method, try to achieve each D2D-P_mN row, d row pre-coding matrix F_mDisaggregation；

S3, calculates each D2D-P_mTo each CU_kInterference matrix H_k,mF_m, design is orthogonal to H_k,mF_mN row, d row the rearmounted encoder matrix W of each phone user_k；

S4, utilizes based on the interference alignment method minimizing mean square error, tries to achieve all D2D-P_mTransmitter pre-coding matrix F_mEncoder matrix G rearmounted with receiver_m, and carry out cellular network communication, to obtain each phone user CU_kSignal to Interference plus Noise Ratio SINR；

S5, arranges D2D-P_mSINR threshold value ω_thIf, D2D-P_mSINR less than ω_th, then by this D2D-P_mAccess cellular network, and β is set_mValue is 1, otherwise arranges β_mValue is 0；

S6, according to transmitter pre-coding matrix F_mAnd β_mValue, utilizes the water injection power improved distribution method to all D2D-P_mDistribution transmit power P_m；

S7, according to the matrix calculation result of step S2～S6 and power distribution result, carries out cellular network communication。

2. a kind of D2D-P multiplexing cellular network communication method improving the availability of frequency spectrum according to claim 1, it is characterised in that described step S2 is particularly as follows: by all of D2D-P_mTo CU_kInterference matrix H_k,mF_mSnap to this CU_kOn the same signal subspace of receiving terminal, formula specific as follows:

span(H_k,1F₁)=...=span (H_k,mF_m)=...=span (H_k,MF_M)

Wherein, k=1,2 ..., K, the subspace that span (A) representing matrix A column vector is opened。

3. a kind of D2D-P multiplexing cellular network communication method improving the availability of frequency spectrum according to claim 1, it is characterised in that described step S4 specifically includes following steps:

S41, calculates the emitted machine pre-coding matrix F that receiver receives_mThe signal level S processed_m, and S_mThe rearmounted encoder matrix G of the machine that is received_mThe signal level obtained after process

S42, defines all D2D-P_mSignal level mean square error and ε_MSEAs follows:

${\mathrm{\&epsiv;}}_{MSE}=\underset{m=1}{\overset{M}{\&Sigma;}}E\left\{\right||\stackrel{\&OverBar;}{S_m}-S_m|{|}^2\}=\underset{m=1}{\overset{M}{\&Sigma;}}E\left\{\right||{G_m}^H{S}_m-S_m|{|}^2\}$

Lagrange's method of multipliers is utilized to solve optimization problem as follows:

$\underset{G_m,F_m}{min}{\mathrm{\&epsiv;}}_{MSE}$

$s.t.\left|\right|F_m|{|}_{Fm}^2=P_m$

Wherein, E{ } represent mathematic expectaion, | | X | |²The norm squared of representing matrix X,The F norm squared of representing matrix X, P_mFor D2D-P_mTransmit power, and meetP is the total transmit power of all of D2D-P。

4. a kind of D2D-P multiplexing cellular network communication method improving the availability of frequency spectrum according to claim 3, it is characterised in that in described step S41, emitted machine pre-coding matrix F_mThe signal level S processed_mSpecifically it is calculated as follows formula:

$S_m=\sqrt{P_m}{H}_{m,m}{F}_m{Z}_m+\underset{j=1,j\&NotEqual;m}{\overset{M}{\&Sigma;}}\sqrt{P_j}{H}_{m,j}{F}_j{Z}_j+n_m$

Wherein, P_mIt is D2D-P_mThe transmit power of transmitter, H_m,mFor D2D-P_mThe N row of transmitted from transmitter to receiver, N row channel matrix, H_m,jFor D2D-P_jTransmitter is to D2D-P_mThe N row of receiver, N row channel matrix, n_mFor D2D-P_mThe environment noise that receiver receives。

5. a kind of D2D-P multiplexing cellular network communication method improving the availability of frequency spectrum according to claim 3, it is characterised in that described step S42 specifically includes following steps:

S420, introduces lagrangian multiplier_m, obtain Lagrangian:

$L(F_m,G_m,{\mathrm{\&lambda;}}_m)={\mathrm{\&epsiv;}}_{MSE}+{\mathrm{\&lambda;}}_m\left(\right|\left|F_m\right|{|}_F^2-P_m)$

Wherein, ε_MSEFor signal level mean square error and；

S421, respectively to G in Lagrangian_m、F_mSeek partial derivative, and to make partial derivative be zero, obtain equation (1), (2):

$F_m={(\underset{j=1}{\overset{M}{\&Sigma;}}{H}_{j,m}{G}_j{G}_j^H{H}_{j,m}^H+{\mathrm{\&lambda;}}_{m}I)}^{-1}{H}_{m,m}^H{G_m}^H---\left(1\right)$

$G_m={F_m}^H{H_{m,m}}^H{(\underset{j=1}{\overset{M}{\&Sigma;}}{H}_{m,j}{F}_j{F}_j^H{\left(H_{m,j}\right)}^H+{\mathrm{\&sigma;}}^{2}I)}^{-1}---\left(2\right)$

Wherein, m=1,2 ..., M, σ²For environment noise n_mVariance, and E{n_mn_m ^H}=σ²I, I are N rank unit matrixs；

S422, initializes pre-coding matrix F_mFor random matrix；

S423, through type (2) calculates and obtains rearmounted encoder matrix G_m；

S424, substitutes into power limitation condition by formula (1)Try to achieve λ_m, λ_mSubstitution formula (1), updates F_m；

S425, calculates all D2D-P_mSignal level mean square error and ε_MSE；

S426, repeats step S423～S425, until ε_MSEConvergence, obtains G_mAnd F_m。

6. a kind of D2D-P multiplexing cellular network communication method improving the availability of frequency spectrum according to claim 1, it is characterised in that in described S5, D2D-P_mCommunication SINR threshold value ω_thFor: the phone user participating in communication measures the Signal to Interference plus Noise Ratio SINR of local reception signal, and issues cellular basestation by feedback channel, cellular basestation wherein minimum SINR is set to D2D-P communication threshold ω_th。

7. a kind of D2D-P multiplexing cellular network communication method improving the availability of frequency spectrum according to claim 1, it is characterised in that in described S6, the water injection power distribution method of described improvement solves D2D-P_mTransmit power P_mComputational methods be:

P_m=β_m(μ/γ_m-1)⁺

Wherein, μ is the level of water filling, meetsγ_mFor D2D-P_mChannel gain,P is the total transmit power of all of D2D-P, function (x)⁺=max (x, 0) represents the higher value in treating excess syndrome number x and 0。