CN102970256A - Zero-space-based interference elimination method for multi-antenna device-to-device (D2D) communication system - Google Patents

Abstract

The invention relates to the technical field of interference elimination for communication systems, in particular to a zero-space-based interference elimination method for a multi-antenna device-to-device (D2D) communication system. By the interference elimination method for the communication system, relatively accurate interference channel transient information is acquired by designing an optimal training sequence length by a channel estimation method, and a wave beam forming matrix is designed according to an interference channel zero space, so that the problem of same frequency interference caused by a reason that a D2D technology is introduced into a honeycomb system can be effectively solved in a manner of signal processing; and moreover, the interference elimination method is simple in theory and extremely high in operability.

Description

Many antennas D2D communication system interference elimination method based on kernel

Technical field

The present invention relates to communication system interference cancellation techniques field, be specifically related to a kind of many antennas D2D communication system interference elimination method based on kernel.

Background technology

Along with the raising of people to two-forty mobile communication business demand, the business throughput that how to improve cellular system becomes the important topic of communications field research.The technology such as Pico cell and Home eNodeB is introduced in the macrocellular network, promotes spectrum utilization efficiency and power system capacity in the whole macrocellular with the honeycomb of classification.But along with the expansion of type of traffic, become a kind of new traffic model between the inner short-range terminal communication in residential quarter, such as business such as video sharings closely.In this simultaneously, D2D(Device-to-Device, direct communication between equipment oriented/end-to-end) communication technology is introduced in the next generation communication system demand of communicating by letter to satisfy the inner short-range terminal in residential quarter.

In traditional cellular system, communication process between the terminal need to pass through BS(BaseStatioN, the base station) or the processing of core net internal element just can finish, and the D2D communication technology is a kind of under the control of cellular system BS, and the communication resource by other equipment in the multiplexing residential quarter when the terminal room communication link quality meets certain requirements carries out the direct point-to-point communication technology.Its technical characterstic is the power loss that reduces the equipment such as user terminal, BS and core net, alleviates the service processing pressure of BS and core net, improves the utilization ratio of frequency spectrum resource in the cellular cell.Because being the mandate frequency range at cellular system, D2D communication communicates, the multiplexing of the foundation of communication link and resource is to finish under the adjusting control of BS simultaneously, therefore the quality of D2D communication is operated in communication on the unauthorized frequency range with respect to those, be guaranteed to a certain extent, this also is the major reason that the professional D2D of employing of other local communications technology realizes.

Although the D2D technology can realize that the local user communicates by letter end to end and then improves the systematic function of cellular network, the inner potential problem of co-channel interference of cellular system of introducing D2D communication can not be ignored.Difference according to the multiplexing phone user of D2D (CUE) up-downgoing resource, interference between the transmitting terminal (DUE-T) of D2D user (DUE), D2D user's receiving terminal (DUE-R), CUE and BS can be divided into two kinds of situations: during when multiplexing ascending resource, have the interference of DUE-T to BS and CUE to DUE-R in the system; When multiplexing downlink resource, there is the interference of BS to DUE-R and DUE-T to CUE in the system.Be with the difference of other resource multiplex communication technologys (for example cognitive radio), the multiplexing resource of D2D user's signal post is to have assigned under the control of BS is coordinated, so owing to the internal system that resource multiplex causes is disturbed and can be controlled by cellular system, i.e. the interference of D2D can be controlled; This advantage is so that its lifting for systematic function has very wide application prospect.

Current research for the D2D technology mainly concentrates on the problem of eliminating of disturbing that solves.Large multi-scheme is the feasible algorithm of angle design from power control and running time-frequency resource dispatching distribution, to realize the elimination of the interference that communication produces for D2D.Also with good grounds different scene is studied co-allocation scheduling and the management of the resources such as time slot, channel, power, and for example: according to the geography information position of DUE and CUE, the CUE resource that space length is far away is distributed to DUE and carried out the modes such as multiplexing.

Yet for the communication system of many antennas, adopting beam forming technique to disturb and eliminating is a kind of feasible method, therefore beam forming technique is incorporated in the D2D system can well resolution system the mutual interference problem.In the D2D system, one comparatively the beam forming matrix design method of reality and easy operating be that kernel according to interference channel designs.Like this, after obtaining interference channel information, by to its decomposition technique, the kernel matrix that can obtain to want.But in the communication system of reality, it is very difficult obtaining perfect interference channel prompting message.

Summary of the invention

The technical problem that (one) will solve

The object of the present invention is to provide the acquisition algorithm of the best prompting message of a kind of interference channel, and be applied in many antennas D2D communication system interference cancellation techniques, be used for suppressing or eliminating the co-channel interference of many antennas D2D communication system.

(2) technical scheme

Technical solution of the present invention is as follows:

A kind of many antennas D2D communication system interference elimination method based on kernel comprises step:

S1. the statistical property in conjunction with communication system designs the first training sequence length;

S2. according to described the first training sequence length, estimate interference channel the first prompting message;

S3. in conjunction with described the first prompting message, design the optimum training sequence length;

S4. according to described optimum training sequence length, estimate the best prompting message of interference channel;

S5. decompose kernel matrix and beam forming matrix that described best prompting message obtains interference channel;

S6. based on described kernel matrix and beam forming matrix, modelled signal sends covariance matrix;

S7. send covariance matrix adjustment communication system parameter according to described beam forming matrix and signal, eliminate and disturb.

Preferably, described step S1 comprises step:

S101. system initialization is set transmitting power, the interference of every antenna of communication system and is born end interference power threshold value, transmission time;

S102. the parameter information of setting according to step S101 obtains the statistical property of system, take optimization system as purpose, designs the first training sequence length.

Preferably, among the described step S102, throughput-maximized or detect bit error rate and be minimised as purpose and design the first training sequence length with communication system link.

Preferably, among the described step S102, the constraints that designs the first training sequence length comprises that system signal transmitting terminal maximum transmit power and interference bear end interference power threshold value.

Preferably, among the described step S2, by utilizing training sequence training method of estimation, blind Channel Estimation Based or channel half blind estimating method estimation interference channel the first prompting message.

Preferably, among the described step S2, the estimation criterion of described method of estimation is minimum mean square error criterion or criterion of least squares.

Preferably, described step S3 comprises step:

S301. in conjunction with described the first prompting message, design the second training sequence length;

S302. compare the first training sequence length and the second training sequence length: if described the second training sequence length is not more than the first training sequence length, think that then the first training sequence length is the optimum training sequence length; If described the second training sequence length greater than the first training sequence length, is then estimated the interference channel prompting message in conjunction with the difference of the second training sequence length and the first training sequence length, calculate the optimum training sequence length.

Preferably, among the described step S5, decompose described best prompting message by singular value decomposition method or QR decomposition method.

Preferably, comprise among the described step S5:

When D2D user's transmitting terminal antenna number in the system greater than base station end antenna number and D2D user's receiving terminal antenna number during greater than cellular user terminal antenna number,, design based on the received beam shaping matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel at D2D user's receiving terminal based on the transmission beam forming matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of D2D user's transmitting terminal;

When D2D user's transmitting terminal antenna number in the system less than base station end antenna number and D2D user's receiving terminal antenna number during greater than cellular user terminal antenna number,, design based on the received beam shaping matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel at D2D user's receiving terminal based on the received beam shaping matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of base station end;

When D2D user's transmitting terminal antenna number in the system greater than base station end antenna number and D2D user's receiving terminal antenna number during less than cellular user terminal,, design based on the transmission beam forming matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel in phone user's transmitting terminal based on the transmission beam forming matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of D2D user's transmitting terminal;

When D2D user's transmitting terminal antenna number in the system less than base station end antenna number and D2D user's receiving terminal antenna number during less than cellular user terminal antenna number,, design based on the transmission beam forming matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel in phone user's transmitting terminal based on the received beam shaping matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of base station end.

Preferably, described step S7 comprises step:

S701. send the transmitted power of the antenna of each real work in the covariance matrix Adjustment System to keep total transmitting power constant according to signal;

S702. communicate in the interference channel kernel at corresponding terminal applies beam forming matrix, eliminate and disturb.

(3) beneficial effect

Communication system interference elimination method of the present invention is by design optimum training sequence length, utilize channel estimation methods to obtain relatively accurate interference channel prompting message and designed the beam forming matrix based on the interference channel kernel, the problem of co-channel interference that brings the angle that therefore can effectively process from signal suppresses or elimination D2D technology the induces one cellular system; Simultaneously, the interference elimination method of this invention realizes that principle is simple, has very strong operability.

Description of drawings

Fig. 1 is the flow chart of a kind of many antennas D2D communication system interference elimination method based on kernel of the present invention;

Fig. 2 be in many antennas D2D communication system during D2D user's multiplexed cellular systems ascending resource DUE-T to the interference model figure of BS;

Fig. 3 is the graph of a relation of interference elimination method among Fig. 1 D2D Lower Channel Bound capacity and training sequence length under different B S interference power thresholding.

Embodiment

Below in conjunction with drawings and Examples, the embodiment of inventing is described further.Following examples only are used for explanation the present invention, but are not used for limiting the scope of the invention.

In order better to set forth implementation detail of the present invention, referring now to Fig. 2 the D2D of the many antennas in present embodiment communication system configuration is illustrated: wherein DUE-T and DUE-R are a pair of D2D users, and its multiplexing CUE ascending resource communicates; The antenna number M of BS _BThe antenna number M of=2, CUE _CThe antenna number M of=1, DUE-T and DUE-R _T=M _R=4.

Flow chart a kind of many antennas D2D communication system interference elimination method based on kernel as shown in Figure 1 mainly comprises step:

S1. the statistical property in conjunction with communication system designs the first training sequence length, and informing user terminal; It mainly comprises step:

S101. system initialization, the transmitting power of setting the every antenna of CUE in the communication system is

The transmitting power of the every antenna of DUE is

DUE-T is that ζ, system are T in total transmission time to the interference power threshold value of BS; This step can also comprise the setting of other parameter.

S102. the parameter information of setting according to step S101 obtains the statistical property of system, take optimization system as purpose, such as throughput-maximized with communication system link or detect bit error rate and minimize etc. and to be optimized for purpose and to design the first training sequence length; Because optimizing link throughput uses wider in actual applications, preferably throughput-maximized as target design the first training sequence length take communication system link in the present embodiment, its constraints comprises that mainly system signal transmitting terminal maximum transmit power and interference bear end interference power threshold value etc.By optimizing the D2D link throughput, obtain DUE-R and CUE is sent training sequence length carry out channel estimating, and BS sends training sequence length to DUE-T and carry out channel estimating, obtain the first training sequence length N _lSpecifically be calculated as follows

$\underset{N_1}{\max }\frac{T-2N_1}{T}E\left\{\log \right|I+H{H}^H\frac{{\mathrm{\σ}}_d^2}{{\mathrm{\σ}}_z^2+{\mathrm{\ω}}_1}\left|\right\}$

s.t.?0<N ₁<T/2

N wherein ₁Be single interference channel training sequence length to be calculated; T is total call duration time that D2D user is once transmitted; H is the channel information of D2D self; Expectation E{ } be the statistical property to H, can on average try to achieve by limited multisample;

Be the average power on the D2D single antenna;

Be the noise variance on the single antenna; ω ₁For from the interference of CUE and be N ₁Function, its expression formula is different according to different channel estimation method, in the present embodiment

Finally, N ₁Can be determined by the one-dimensional discrete point search.

S2. according to the first training sequence length N ₁, estimation interference channel the first prompting message; Sending length to BS and DUE-R respectively by DUE-T and CUE in the present embodiment is N ₁Training sequence, receiving terminal adopt the channel estimation methods based on minimum mean square error criterion or criterion of least squares to estimate corresponding interference channel the first prompting message

With And this information fed back to associated terminal.Channel estimation methods also can be blind Channel Estimation Based or channel half blind estimating method or other any known methods of estimation.Because to channel information H _CRWith channel information H _TBThe channel estimating principle the same, so receive training sequence take the BS termination in the present embodiment and this process be described as example.N ₁After the individual time slot, the training sequence expression formula that the BS termination is received is:

Y _B=H _TBT ₁+Z

T wherein ₁M _T* N ₁The training matrix of dimension, Z is M _B* N ₁The noise matrix of dimension, the variance of its each element is In order to obtain H _TBEffective estimation, the first training sequence length N ₁Should satisfy N ₁〉=M _TH based on the estimation of LMMSE criterion _TBBe that interference channel the first prompting message can be expressed as:

${\hat{H}}_{\mathrm{TB}}=Y_B{(T_1^H{T}_1+{\mathrm{\&sigma;}}_z^2{M}_{B}I)}^{-1}{T}_1^H{R}_{H_{\mathrm{TB}}}$

In like manner can obtain H _CRChannel estimating

S3. in conjunction with interference channel the first prompting message of obtaining of estimation, further optimization system performance and obtain the length of optimum training sequence, the notice associated terminal upgrades channel estimating.

Present embodiment is specially: to interference channel the first prompting message of estimating to obtain With

Do singular value decomposition, from

Right singular matrix in obtain interference channel

The kernel matrix

And from Left singular matrix in obtain interference channel

The kernel matrix Be respectively

${\hat{H}}_{\mathrm{TB}}={\hat{U}}_{\mathrm{TBn}}\left[\begin{array}{cc}{\widehat{\mathrm{\&Sigma;}}}_{\mathrm{TB}}& 0\end{array}\right]\left[\begin{array}{c}{\hat{V}}_{\mathrm{TBs}}^H\\ {\hat{V}}_{\mathrm{TBn}}^H\end{array}\right]$

With

${\hat{H}}_{\mathrm{CR}}=[{\hat{U}}_{\mathrm{CRs}},{\hat{U}}_{\mathrm{CRn}}]\left[\begin{array}{c}{\widehat{\mathrm{\&Sigma;}}}_{\mathrm{CR}}\\ 0\end{array}\right]\left[{\hat{V}}_{\mathrm{CRs}}^H\right]$

With

Be the transmission precoding of DUE-T end, with For the reception precoding of DUE-R end, so that interfere information transmits in the interference channel kernel.Calculate in the following manner the second training sequence length N ₂:

$\underset{N_2,R_{\mathrm{dT}}}{\max }\frac{T-2N_2}{T}\log |I+\frac{\hat{A}{\hat{R}}_{\mathrm{dT}}{\hat{A}}^H}{{\mathrm{\&sigma;}}_z^2+{\mathrm{\&omega;}}_2}|,$

$s.t.\mathrm{tr}\left({\hat{R}}_{\mathrm{dT}}\right)\&le;J,$ ${\hat{R}}_{\mathrm{dT}}\&GreaterEqual;0,$ 0<N ₂<T/2,

Wherein

ω ₂For from the interference of CUE and be N ₂Function, its expression formula is different according to different channel estimation method, in the present embodiment $J=\min \{M_T{\mathrm{\&sigma;}}_d^2,\frac{M_B{\mathrm{\&sigma;}}_z^2+N_2{\mathrm{\&sigma;}}_d^2}{M_B^2{\mathrm{\&sigma;}}_z^2}\mathrm{\&zeta;}\}.$

Fig. 3 has provided at the graph of a relation between D2D link heap(ed) capacity and the training sequence length N under the different BS interference power thresholdings, the result shows when interference threshold ζ is very little, interference threshold becomes the main constraints of restriction D2D link capacity, needs more training sequence length to obtain more accurate kernel and then the interference threshold that satisfies BS requires so that the optimum capacity of D2D link; When interference threshold ζ is larger, then just can obtain optimum capacity in requisition for a small amount of training sequence length on the contrary.

Compare the first training sequence length N ₁With the second training sequence length N ₂If: the second training sequence length N ₂Be not more than the first training sequence length N ₁, then think the first training sequence length N ₂Be the optimum training sequence length; If the second training sequence length N ₂Greater than the first training sequence length N ₁, then in conjunction with the second training sequence length N ₂With the first training sequence length N ₁Difference N ₂-N ₁, repeat above-mentioned steps and calculate the optimum training sequence length.

S4. according to the optimum training sequence length, estimate the best prompting message of interference channel;

Adopt the LMMSE estimation criterion, obtain

Final estimation be that the best prompting message of interference channel can be expressed as:

${\tilde{H}}_{\mathrm{TB}}=Y_B{(T_2^H{T}_2+{\mathrm{\&sigma;}}_z^2{M}_{B}I)}^{-1}{T}_2^H{R}_{H_{\mathrm{TB}}}$

T wherein ₂M _T* N ₂All training sequence matrixes, and the covariance matrix of hypothesis channel satisfies

Because the imperfection of channel estimating, the error that definition is estimated is:

$\mathrm{\&Delta;}{H}_{\mathrm{TB}}=H_{\mathrm{TB}}-{\tilde{H}}_{\mathrm{TB}}$

Then by the covariance matrix that calculates channel errors be:

$R_{\mathrm{TB}}=E\left[\mathrm{\&Delta;}{H}_{\mathrm{TB}}\mathrm{\&Delta;}{H}_{\mathrm{TB}}^H\right]={(I+\frac{1}{{\mathrm{\&sigma;}}_z^2{M}_B}{T}_2{T}_2^H)}^{-1}$

The transmitted power of setting each antenna of DUE-T when system initialization is

The optimum training sequence satisfies The time, then finally calculate the channel errors covariance matrix and be

Adopt in a like fashion, DUE-R can obtain

Corresponding channel estimation errors is Δ H _CR, its covariance matrix is

S5. the channel information that estimation is obtained

With

Do the kernel matrix beam forming matrix that singular value decomposition obtains interference channel by singular value decomposition method or QR decomposition method; When D2D user's transmitting terminal antenna number in the system greater than base station end antenna number and D2D user's receiving terminal antenna number during greater than cellular user terminal antenna number,, design based on the received beam shaping matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel at D2D user's receiving terminal based on the transmission beam forming matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of D2D user's transmitting terminal; When D2D user's transmitting terminal antenna number in the system less than base station end antenna number and D2D user's receiving terminal antenna number during greater than cellular user terminal antenna number,, design based on the received beam shaping matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel at D2D user's receiving terminal based on the received beam shaping matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of base station end; When D2D user's transmitting terminal antenna number in the system greater than base station end antenna number and D2D user's receiving terminal antenna number during less than cellular user terminal,, design based on the transmission beam forming matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel in phone user's transmitting terminal based on the transmission beam forming matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of D2D user's transmitting terminal; When D2D user's transmitting terminal antenna number in the system less than base station end antenna number and D2D user's receiving terminal antenna number during less than cellular user terminal antenna number,, design based on the transmission beam forming matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel in phone user's transmitting terminal based on the received beam shaping matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of base station end.

Adopt singular value decomposition method in the present embodiment: according to different number of antennas, from

Left or right singular matrix in obtain interference channel

The kernel matrix

From Right or left singular matrix in obtain interference channel

The kernel matrix

And with corresponding kernel matrix as sending precoding and receiving precoding, so that interfere information transmits in the interference channel kernel.Be specially:

DUE-T obtains the information of interference channel by feedback channel

It is done singular value decomposition decompose, obtain

${\tilde{H}}_{\mathrm{TB}}={\tilde{U}}_{\mathrm{TBn}}\left[\begin{array}{cc}{\widetilde{\mathrm{\&Sigma;}}}_{\mathrm{TB}}& 0\end{array}\right]\left[\begin{array}{c}{\tilde{V}}_{\mathrm{TBs}}^H\\ {\tilde{V}}_{\mathrm{TBn}}^H\end{array}\right]$

M wherein _T* (M _T-M _B) the dimension matrix

Be

Kernel, will

Beam forming matrix as the DUE-T end.But because the actual interference channel is H _TB, to estimated value

Be SVD and decompose the kernel that obtains

Can not with H _TBSo complete quadrature is the signal of DUE-T process Can not eliminate fully the interference of BS after doing precoding.

In like manner, the interference channel of DUE-R to estimating

Being SVD decomposes:

${\tilde{H}}_{\mathrm{CR}}=[{\tilde{U}}_{\mathrm{CRs}},{\tilde{U}}_{\mathrm{CRn}}]\left[\begin{array}{c}{\widetilde{\mathrm{\&Sigma;}}}_{\mathrm{CR}}\\ 0\end{array}\right]\left[{\tilde{V}}_{\mathrm{CRs}}^H\right]$

M wherein _R* (M _R-M _C) the dimension matrix

Be

Kernel, will

Received beam shaping matrix as the DUE-R end.Because the imperfection of channel estimating,

Can not with the interference channel H of reality _CRQuadrature is used so work as DUE-R fully

When doing filtering to received signal, the signal of CUE can not be eliminated fully.

Estimate kernel

With

Kernel V with actual channel _TBnAnd U _CRnError calculate by following:

H _TBΔV _TBn≈-ΔH _TBV _TBn

In like manner Error delta U _CRnCan be expressed as:

$H_{\mathrm{CR}}^H\mathrm{\&Delta;}{U}_{\mathrm{CRn}}=-\mathrm{\&Delta;}{H}_{\mathrm{CR}}^H{U}_{\mathrm{CRn}}.$

S6. based on the beam forming matrix, the signal of design DUE-T sends covariance matrix; Send beam forming when DUE-R uses the received beam shaping when using at DUE-T, the equivalent channel between D2D can be expressed as:

$\tilde{A}={\tilde{U}}_{\mathrm{CRn}}^{H}H{\tilde{V}}_{\mathrm{TBn}}$

Then the best covariance matrix that adopts of DUE-T should be obtained by following formula:

$\underset{{\tilde{R}}_{\mathrm{dT}}}{\max }\log |I+\frac{\tilde{A}{\tilde{R}}_{\mathrm{dT}}{\tilde{A}}^H}{{\mathrm{\&sigma;}}_z^2+{\mathrm{\&omega;}}_2}|,$

$s.t.\mathrm{tr}\left({\tilde{R}}_{\mathrm{dT}}\right)\&le;J,$ ${\tilde{R}}_{\mathrm{dT}}\&GreaterEqual;0,$

Wherein J is determined by the maximum interference that maximum power and the BS of DUE-T can bear,

By right

Eigenvalues Decomposition obtain Q Λ Q ^H, wherein Q is unitary matrice, simultaneously

It is a diagonal matrix.Definition Above-mentioned optimization problem can change into

$\underset{X}{\max }\log |I+\frac{\mathrm{X\&Lambda;}}{{\mathrm{\&sigma;}}_z^2+{\mathrm{\&omega;}}_2}|,$

$s.t.\mathrm{tr}\left({\tilde{R}}_{\mathrm{dT}}\right)\&le;J,$ ${\tilde{R}}_{\mathrm{dT}}\&GreaterEqual;0,$

Adopt standardized method, can obtain when X is diagonal matrix, i.e. X=Diag (x ₁..., x _k) the problems referred to above optimum.

$\underset{\left\{x_i\right\}}{\max }\underset{i=1}{\overset{k}{\mathrm{\&Sigma;}}}\log (1+\frac{x_i{\mathrm{\&lambda;}}_i^2}{{\mathrm{\&sigma;}}_z^2+{\mathrm{\&omega;}}_2})$

$s.t.\underset{i=1}{\overset{k}{\mathrm{\&Sigma;}}}{x}_i\&le;J,x_i\&GreaterEqual;0;$

x _iCan obtain x by the water-filling algorithm of standard _iSolution be

$x_i={(v-\frac{{\mathrm{\&sigma;}}_z^2+{\mathrm{\&omega;}}_2}{{\mathrm{\&lambda;}}_i^2})}^{+},$

+ representative is got

With 0 maximum among both, v represents water line, and it satisfies:

$\underset{i=1}{\overset{k}{\mathrm{\&Sigma;}}}{(v-\frac{{\mathrm{\&sigma;}}_z^2+{\mathrm{\&omega;}}_2}{{\mathrm{\&lambda;}}_i^2})}^{+}=J,$

Further, the maximum throughput of D2D link can be expressed as:

$C_{D2D}=\frac{T-2N_2}{T}\underset{I=1}{\overset{k}{\mathrm{\&Sigma;}}}{\left(\log \frac{v{\mathrm{\&lambda;}}_i^2}{{\mathrm{\&sigma;}}_z^2+{\mathrm{\&omega;}}_2}\right)}^{+}.$

S7. send covariance matrix adjustment communication system parameter according to beam forming matrix and signal, eliminate and disturb.Be specially: when adopting the transmission beam forming technique, owing to only use the part transmitting antenna, thus need according to actual needs, total transmitted power mean allocation is in the actual transmitting antenna of using, constant to guarantee total transmitted power.Send or received beam shaping matrix in described terminal applies, and adopt the transmitting antenna number to adjust each according to reality and adopt the DUE-T antenna transmitted power of beam forming to be

Constant to guarantee total transmitted power.Antenna number according to native system is set, and DUE-T utilizes transmission beam forming matrix that transmitted signal is pointed to H _TBKernel transmission; DUE-R utilizes received beam shaping matrix at interference channel H _CRKernel in receive the data that DUE-T sends.

A kind of many antennas of robustness D2D communication system interference elimination method based on the interference channel kernel provided by the invention, the resource of other portable terminals communicates and eliminates interference each other in its permission D2D user multiplexed cellular systems.This method based on the known statistical property of system take the optimization system performance as target design interference channel training sequence with the covariance matrix of D2D transmitted signal.Take this training sequence as benchmark, under the cooperation of D2D user, base station and the terminal that is re-used, estimate interference channel information, and find the solution the channel zero space, at this kernel basis design sending and receiving beam forming matrix, so that interfere information transmits, eliminate the interference between D2D user and the multiplexing phone user in the kernel of interference channel.Advantage of the present invention and characteristics are according to estimating channel information design kernel beam forming matrix, and considered that the imperfection of channel estimating is on the impact of kernel design, to reach the purpose of robust optimization system property indices, has very strong practicality.

Above execution mode only is used for explanation the present invention; and be not limitation of the present invention; the those of ordinary skill of correlative technology field; in the situation that does not break away from the spirit and scope of the present invention; can also make a variety of changes and modification, so all technical schemes that are equal to also belong to protection category of the present invention.

Claims (10)

1. the many antennas D2D communication system interference elimination method based on kernel is characterized in that, comprises step:

S1. the statistical property in conjunction with communication system designs the first training sequence length;

S2. according to described the first training sequence length, estimate interference channel the first prompting message;

S3. in conjunction with described the first prompting message, design the optimum training sequence length;

S4. according to described optimum training sequence length, estimate the best prompting message of interference channel;

S5. decompose kernel matrix and beam forming matrix that described best prompting message obtains interference channel;

S6. based on described kernel matrix and beam forming matrix, modelled signal sends covariance matrix;

S7. send covariance matrix adjustment communication system parameter according to described beam forming matrix and signal, eliminate and disturb.

2. interference elimination method according to claim 1 is characterized in that, described step S1 comprises step:

S101. system initialization is set transmitting power, the interference of every antenna of communication system and is born end interference power threshold value, transmission time;

S102. the parameter information of setting according to step S101 obtains the statistical property of system, take optimization system as purpose, designs the first training sequence length.

3. interference elimination method according to claim 2 is characterized in that, and is among the described step S102, throughput-maximized or detect bit error rate and be minimised as purpose and design the first training sequence length with communication system link.

4. interference elimination method according to claim 3 is characterized in that, among the described step S102, the constraints that designs the first training sequence length comprises that system signal transmitting terminal maximum transmit power and interference bear end interference power threshold value.

5. according to claim 1 to 4 described interference elimination methods, it is characterized in that, among the described step S2, by utilizing training sequence training method of estimation, blind Channel Estimation Based or channel half blind estimating method estimation interference channel the first prompting message.

6. interference elimination method according to claim 5 is characterized in that, among the described step S2, the estimation criterion of described method of estimation is minimum mean square error criterion or criterion of least squares.

7. interference elimination method according to claim 1 is characterized in that, described step S3 comprises step:

S301. in conjunction with described the first prompting message, design the second training sequence length;

S302. compare the first training sequence length and the second training sequence length: if described the second training sequence length is not more than the first training sequence length, think that then the first training sequence length is the optimum training sequence length; If described the second training sequence length greater than the first training sequence length, is then estimated the interference channel prompting message in conjunction with the difference of the second training sequence length and the first training sequence length, calculate the optimum training sequence length.

8. interference elimination method according to claim 1 is characterized in that, among the described step S5, decomposes described best prompting message by singular value decomposition method or QR decomposition method.

9. according to claim 1-4, the described interference elimination method of 5-8 any one, it is characterized in that, comprise among the described step S5:

When D2D user's transmitting terminal antenna number in the system greater than base station end antenna number and D2D user's receiving terminal antenna number during greater than cellular user terminal antenna number,, design based on the received beam shaping matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel at D2D user's receiving terminal based on the transmission beam forming matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of D2D user's transmitting terminal;

When D2D user's transmitting terminal antenna number in the system less than base station end antenna number and D2D user's receiving terminal antenna number during greater than cellular user terminal antenna number,, design based on the received beam shaping matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel at D2D user's receiving terminal based on the received beam shaping matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of base station end;

When D2D user's transmitting terminal antenna number in the system greater than base station end antenna number and D2D user's receiving terminal antenna number during less than cellular user terminal,, design based on the transmission beam forming matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel in phone user's transmitting terminal based on the transmission beam forming matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of D2D user's transmitting terminal;

When D2D user's transmitting terminal antenna number in the system less than base station end antenna number and D2D user's receiving terminal antenna number during less than cellular user terminal antenna number,, design based on the transmission beam forming matrix of cellular user terminal to D2D user's receiving terminal interference channel kernel in phone user's transmitting terminal based on the received beam shaping matrix of D2D user's transmitting terminal to base station end interference channel kernel in the design of base station end.

10. interference elimination method according to claim 9 is characterized in that, described step S7 comprises step:

S701. send the transmitted power of the antenna of each real work in the covariance matrix Adjustment System to keep total transmitting power constant according to signal;

S702. communicate in the interference channel kernel at corresponding terminal applies beam forming matrix, eliminate and disturb.

Images