CN104052529A - Antenna array and communication method used for full duplex communication - Google Patents

Abstract

In order to solve the disadvantages of the existing full duplex communication, the invention provides an antenna array and a communication method used for the full duplex communication. The antenna array comprises a first sub array with a plurality of first antenna elements and a second sub array with a plurality of second antenna elements. At the same time, one in the two sub arrays is used for sending a signal, the other one is used for receiving a signal, and thus the full duplex communication is carried out, wherein the first sub array and the second sub array are separated, and the polarization direction of the first antenna elements of the first sub array is orthogonal to the polarization direction of the second antenna elements of the second sub array. The antenna array has a high isolation degree between full duplex receiving and sending, and the self interference of full duplex receiving and sending is reduced.

Description

A kind of aerial array and a kind of communication means for full-duplex communication

Technical field

The present invention relates to radio communication, relate in particular to full-duplex communication.

Background technology

In traditional intercommunication system, half-duplex technology (for example TDD time division duplex and FDD Frequency Division Duplexing (FDD)) is generally used for separating transmitted signal and receives signal.

The present invention is intended to propose for realizing the technology of full-duplex communication.A long time ago, full-duplex communication has just caused widely to be paid close attention to, but until full-duplex communication today also lacks technical operability.Full duplex means that radio node can implement sending and receiving at same time by identical carrier frequency.Because full duplex has the potential of the spectrum efficiency of the system of doubling, therefore it is very attractive.Full duplex can also solve the problem of the concealed nodes in such as, competitive environment based on wireless network (WiFi, bluetooth etc.), be that competition under half-duplex situation can cause certain user's equipment cannot obtain all the time service, thereby become the node of " hiding ".

It is unpractical that full duplex is considered to conventionally, and this is because transmitting-receiving is conventionally by the strong self-interference causing between transmitter and the receiver of himself simultaneously, and has reduced spectrum efficiency.Therefore, the ultimate challenge that designs actual full duplex radio network is to remove the impact of strong self-interference, and strengthens the detectability of the signal to expecting.

Existing solution can classify as two kinds: based on the full duplex of simulation arrester (analog cancellator), and the full duplex separating based on space.

For the first kind, Fig. 1 shows typical existing scheme (referring to Full-Duplex Wireless Communications Using Off-The-Shelf Radios Feasibility and First Results, VTC, 2010).The major defect of this scheme is to need two transmitters at each transmitter channel, and one for normal transmission, and another is for disturbing inhibition.This and existing wireless system are incompatible.In addition, for cancellation self-interference, based on interference signal (x_1*h_ab) and the thering is special power and the signal of phase place (c_1) and need to be injected into and eliminate in branch road of estimation of eliminating the response (h_z) of branch road (cancellation brand) channel.In practice, determine that the phase place of c1 and amplitude are difficult, therefore h_ab be change and due to the reflection of antenna building around, h_ab will be difficult to tracking.Therefore, this class full duplex scheme lacks operability.

For Equations of The Second Kind, Fig. 2 shows typical existing scheme (referring to Self-Interference Suppression in Full-Duplex MIMO Relays, PIMRC2011).As shown in the figure, it is different that this system needs information source and the stay of two nights, is therefore conventionally applied in the environment of relaying.In this system, information source and the stay of two nights are not operated in full-duplex mode, and only have relaying to be operated in full duplex.And the reception antenna of relaying and transmitting antenna have very strong directive property, point to respectively information source and the stay of two nights, realize like this transmitting-receiving simultaneously.Visible, this class full duplex system has utilized the space of transmitter and its local receiver to separate to realize self-interference inhibition conventionally.But the application scenarios of this scheme is very restricted.In common base station, transmitting antenna cannot be carried out to this separation with reception antenna, be therefore difficult to use this technology.

Summary of the invention

As can be seen here, up to now, also not for realizing for example, full duplex transmission at conventional Cellular Networks system (LTE or LTE-A), and the feasible program of radio circuit not being changed.

In order to solve the problem in background technology, the present invention proposes new full duplex technical scheme.First aspect, has proposed a kind of new design of the aerial array that is suitable for full-duplex communication, and it has lower cost.The outer appearnce of the outward appearance of this new aerial array and conventional half-duplex aerial array seemingly.Difference is mainly in the following areas: 1. separate and orthogonal polarization sending and receiving antenna element; 2. use the antenna element of equipolarization to form vertical subarray, realized the vertical beam suppressing for self-interference and be shaped.

Second aspect, has proposed a kind of full duplex sending and receiving scheme, further to reduce the impact of self-interference and the performance of increase system.

New Antenna Design and the full duplex sending and receiving scheme that the present invention proposes can together with use, also can use respectively.

Concrete, according to a first aspect of the invention, a kind of aerial array is proposed, it is characterized in that, comprising: the first subarray, has multiple first day line elements; The second subarray, has multiple the second antenna elements; At one time, one in described the first subarray and described the second subarray for transmitted signal, another is for receiving signal, to carry out full-duplex communication, wherein, described the first subarray separates with the second subarray, and the polarised direction of the polarised direction of this first day line element of described the first subarray and the second antenna element of described the second subarray is orthogonal.

Therefore, the present invention this on the one hand institute's aerial array that provides carry out respectively sending and receiving with the antenna of orthogonal polarization, there is very high isolation, make the transmission in full-duplex communication can not cause too large impact to reception.By new full duplex antenna array design, avoid the amendment to transceiver circuit.Therefore, do not need special circuit, saved cost.And do not need to design complicated simulation self-interference yet and eliminate circuit and signal.Inventor is to having carried out emulation according to the performance of this aspect execution mode, simulation results show, arbitrary port between the minimum isolation of observing be 45dB, this than the common isolation height between the traditional antenna of two co-polarizations 20dB.

According to one preferred embodiment, in this aerial array, different transmitting-receiving units, or receiver, or transmitter is connected respectively to each first day line element, and, different transmitting-receiving units, or receiver, or transmitter is connected respectively to each the second antenna element.

In prior art, all antenna elements are all connected to same transmitting-receiving unit, so the suffered self-interference of each antenna can superpose to this transmitting-receiving unit, thereby self-interference can rise, and thereby isolation had to requirements at the higher level.In this embodiment, each transmitting-receiving unit is processed the signal of individual antenna unit, makes transceiver only be subject to the impact of suffered self-interference on this single transceiver, so alleviated for the insulated degree requirement that is sent to reception.And transceiver can be controlled separately each antenna, thereby make the beam forming of antenna be called possibility.

According to a further preferred embodiment, described transmitting-receiving unit, or receiver, or the signal that transmitter receives/sends each first day line element is respectively weighted control, be shaped with the vertical beam of the sending/receiving that realizes the first subarray; Described transmitting-receiving unit, or receiver, or the signal that transmitter receives/sends each the second antenna element is respectively weighted control, be shaped with the vertical beam of the sending/receiving that realizes the second subarray.

Preferably, described vertical beam is shaped for being reduced to the Power leakage of adjacent antenna unit, and suppresses the self-interference between the first subarray and the second subarray.

This preferred embodiment in, because each antenna element is independently, phase place and gain that each transceiver can digitally be controlled (weighting) each antenna element form the vertical beam of subarray rank.Can, by selecting applicable vertical weight to adjust vertical beam, to maximize the isolation between sending and receiving subarray, contribute to be reduced to the power leakage of its adjacent antenna, and suppress self-interference.

According to a further preferred embodiment, described the first subarray and the second subarray are also configured to carry out horizontal beam shaping, realize the beam forming in three dimensions, to reduce the power leakage of adjacent antenna unit, suppress self-interference, improve user simultaneously and accept performance.

In this embodiment, can design together vertical beam and traditional horizontal beam, thereby use the aerial array of 2 dimensions to realize 3D beam forming, can realize and minimizing self-interference and maximizing uplink and downlink trading off between handling up.

Be additional to or be independent of above first aspect, a second aspect of the present invention provides a kind of communication means for full-duplex communication, it is characterized in that, comprise the steps: that i. estimates to send downlink information to receiving the self-interference channel of the self-interference that produces of uplink information; Ii. according to this estimated self-interference channel, estimation of communication channels; Iii. according to this estimated self-interference channel and this estimated communication channel, carry out corresponding upward signal reception or downstream signal and send.

In this second aspect, self-interference channel is estimated, communication channel is estimated, thereby can obtain relatively accurate communication channel, and, carry out full-duplex communication according to estimated self-interference channel and communication channel, can obtain reasonable performance.Inventor is to having carried out emulation according to the execution mode of this second aspect, and simulation result shows the throughput gain that can obtain 30%.

According to one preferred embodiment, described step I comprises: a. sends downlink reference signal, measures uplink receiving signal simultaneously, wherein, contains the self-interference that downlink reference signal causes in uplink receiving signal; B. based on described uplink receiving signal and an estimation filter, estimate described self-interference channel, wherein, described estimation filter is relevant to downlink reference signal.

In this embodiment, with downlink reference signal, self-interference channel is estimated.

According to a further preferred embodiment, in described step b, described estimation filter comprises single-ended filter, and the estimated matrix of this filter is

wherein, P _dLthe vector of downlink reference signal,

Estimated described self-interference channel is

wherein, Y _uLbe described uplink receiving signal, it comprises downlink reference signal.

This execution mode provides a kind of use single-ended (one-tap) filter to estimate the method for self-interference channel.The strength ratio of the signal obtaining through self-interference channel due to downlink reference signal is much larger via the intensity of the upward signal of up channel transmission, so can be the downlink reference signal after interference channel by uplink receiving signal approximation, and because downlink reference signal is known, so can use simplifiedly single-ended filter to solve self-interference channel.Be appreciated that other similar filters, for example MMSE filter also can be used.

According to one preferred embodiment, in described step I i, be to estimate up channel, this step I i comprises: ii-1. measures uplink receiving reference signal, and described uplink receiving reference signal comprises uplink reference signals, and comprises the self-interference that the downlink information that simultaneously sends causes; Ii-2. this downlink information based on known and estimated described self-interference channel are removed the self-interference that described downlink information causes from uplink receiving signal, obtain and remove the reception uplink reference signals disturbing; The reception uplink reference signals that ii-3 disturbs based on described removal and another estimation filter, estimate described up channel, and wherein, described another estimation filter is relevant to uplink reference signals.

This execution mode provides estimates the technical scheme that up channel carries out full duplex reception.Because the downlink information sending is that known, self-interference channel is estimated simultaneously, send so eNB can obtain full duplex the self-interference producing, and remove this self-interference from receive signal.And because the reception signal of removing after self-interference should equal by the uplink reference signals after up channel, and uplink reference signals is that eNB is known, so can easily estimate up channel.

In an execution mode being more preferably, in described step I i-3, described another estimation filter comprises single-ended filter, and the estimated matrix of this filter is

wherein, P _uLthe vector of uplink reference signals,

Estimated described up channel is

wherein, Y (P _uL) be to remove the described reception uplink reference signals disturbing.

This execution mode provides a kind of use single-ended (one-tap) filter to estimate the method for up channel, easier.Be appreciated that other similar filters, for example MMSE filter also can be used.

According to a further preferred embodiment, the method also comprise the steps: x. according in described step I i estimated go out described up channel and described step I ii in recover uplink information, adjust estimated described self-interference channel; Y. according to the estimated described self-interference channel after adjusting, again carry out described step I i and estimate that up channel and described step I ii recover uplink information; Repeat above step x and described step y, until estimated self-interference channel, up channel and uplink information convergence.

While estimating self-interference channel in step I, ignored the impact of the upward signal sending through up channel before.In this embodiment, further revise the estimation to self-interference channel with the upward signal and the up channel that recover, improved the accuracy of channel estimating, improved performance performance.

According to one preferred embodiment, in described step I i also using estimated described up channel as estimated down channel; Or, in described step I i, using estimated described up channel as estimated down channel, on this basis, improve the accuracy of estimated down channel with channel reciprocity calibration steps.

This execution mode provides estimates the technical scheme that down channel carries out full duplex transmission.Because the method is for full duplex system, so uplink/downlink channel has reciprocity: two channels can be thought identical, or can make down channel more accurate by known channel reciprocity calibration steps.

According to a further preferred embodiment, described step I ii comprises: the down channel based on estimated and estimated self-interference channel, determine the delivery plan that downlink information is sent; Use this delivery plan to send this downlink information; Wherein, this delivery plan meets following constraint: make to send the transmitted power maximum of downlink information, and make to send the self-interference minimum that downlink information produces receiving uplink information.

In this embodiment, consider the impact of self-interference, determine that a full duplex delivery plan obtains performance balance between full duplex sending and receiving, thereby optimize full duplex transmitting-receiving.

Further preferred embodiment described delivery plan comprises pre-coding matrix according to one, and described pre-coding matrix is W _fD=GEV (H ^h _dLh _dL, H ^h _{rT, eNB}h _{rT, eNB}), wherein GEV () is the maximum vague generalization Eigenvalues Decomposition of getting two matrixes, H _dLestimated down channel, H _{rT, eNB}described estimated self-interference channel.

This execution mode provides one more concrete full duplex to be sent to the technical scheme of carrying out precoding.Simulation result shows, compares not full duplex is sent and carried out any optimization, and this technical scheme can suppress self-interference and reach 18dB.

Various aspects of the present invention are by the explanation of the specific embodiment by hereinafter and more clear.

Brief description of the drawings

By reading the detailed description that non-limiting example is done of doing with reference to the following drawings, other features, objects and advantages of the present invention will become more apparent:

Fig. 1 is the schematic diagram of existing a kind of full duplex system;

Fig. 2 is the schematic diagram of existing another kind of full duplex system;

Fig. 3 is existing aerial array and according to the structural representation of the aerial array of an embodiment of the invention;

Fig. 4 shows the relation between the class of attenuation of two dipole antennas and orthogonal polarization isolation;

Fig. 5 a shows the direction pattern of an antenna element;

Fig. 5 b shows the direction pattern according to the vertical subarray of one embodiment of the present invention;

Fig. 6 schematically shows the full duplex aerial array according to one embodiment of the present invention;

Fig. 7 is the screenshot capture of the simulation result of the isolation between aerial array middle port 8 and the port 9,13 of Fig. 6;

Fig. 8 is the simulation result of the self-interference power when adopting self-interference according to the embodiment of the present invention to suppress and not adopting self-interference to suppress;

Fig. 9 is the simulation result of the throughput of half-duplex operation and full-duplex communication according to the embodiment of the present invention.

In the accompanying drawings, the same or similar Reference numeral that runs through different diagram represents identical or corresponding parts or feature.

Embodiment

full duplex active antenna array

First aspect of the present invention provides a kind of aerial array, it is characterized in that, comprising:

The-the first subarray, has multiple first day line elements;

The-the second subarray, has multiple the second antenna elements;

At one time, one in described the first subarray and described the second subarray for transmitted signal, and another is for receiving signal, to carry out full-duplex communication,

Wherein, described the first subarray separates with the second subarray, and the polarised direction of the polarised direction of this first day line element of described the first subarray and the second antenna element of described the second subarray is orthogonal.

Fig. 3 shows the aerial array (left side) of the prior art with transmitter, and new full duplex antenna array design (the right) according to the embodiment of the present invention.This new array comprises the individual active antenna of MxN (M=2, N=8) (AA) unit, and it forms planar array.Each AA antenna element comprises that transmitter and receiver are to (transceiver) and antenna element.Top subarray and the bottom subarray of AA array are orthogonal polarizations, and can in identical frequency band, carry out respectively sending/receiving simultaneously, for example, implement full-duplex communication.Be appreciated that this structure is only example, these two subarrays can be arranged in any way, for example staggered, even overlapping, as long as the antenna element of subarray is separated from each other.

New aerial array and being for the important difference of semiduplex traditional aerial array:

1) sending and receiving antenna be separate and orthogonal (intersections) polarize

According to list of references (Vogel, W.J., J.Goldhirsh, and Y.Hase, " Land-Mobile-Satellite Fade Measurements in Australia; " AIAA Journal of Spacecraft and Rockets, Vol.29, No.1, Jan-Feb, pp.123-128) analysis in, the isolation between orthogonal (intersection) poliarizing antenna is two other monotonic decreasing functions of the attenuation grade between antenna, as two cross-polarized dipole antennas in Fig. 4.Decay rank is the monotonically increasing function of separating distance.Therefore, the isolation between antenna is for two near the cross polarised antennas (distance be less than 10 wavelength) of placing and Yan Shigao, and polarization difference can be ignored for the effect of the transmission to remote user (being greater than 10 meters).

The new aerial array for full duplex on the right in Fig. 3, the top of array is polarized to-45 degree, and bottom is polarized to+45 degree.In order to verify the isolation of this design, in HFSS, set up simulation model, and carry out isolation measurement.Arbitrary port between the minimum isolation of observing be 45dB, this than the common isolation height between the traditional antenna of two co-polarizations 20dB.The result of emulation can be referring to description hereinafter.Being appreciated that this pair of polarizing angle degree is only example, as long as other polarization angles meet orthogonal property, is all operable.

2) each antenna element is associated with RF transmitter and/or RF receiver

In traditional aerial array, as shown in left side in Fig. 3, antenna element is passive therein, and antenna element is connected to same transceiver.Unlike this, one preferred embodiment in, the each antenna element in new design is connected to different transceivers.

Comprise for the advantage of full duplex transmission with this decision design of the present invention:

(a) avoided the rising by the caused self-interference of signal of the multiple reception antenna of merging unit, and alleviated for the insulated degree requirement that is sent to reception antenna.For example, every row of the array in Fig. 3 left side comprise 8 reception antenna units.So for this array, theoretic power merges gain for 10*log10 (8)=9dB, this means at this traditional aerial array under the situation for full duplex, if this antenna receives for full duplex, the self-interference that so each antenna is subject to will be superimposed on TRX, the self-interference of this stack will increase 9dB, and this also just needs antenna to increase the isolation of extra 9dB, is difficult to meet.

(b) be allowed for suppressing the sending and receiving beam forming of self-interference.This will describe in detail hereinafter.

It should be noted that according to application scenario, in Fig. 3, in the execution mode on right side, in each antenna element, connected transceiver can be replaced by single transmitter or receiver.Use the benefit of TDD transceiver to be full duplex array hardware and traditional half-duplex array compatibility.In addition, can not be subject to again the impact that the antenna gain in the time switching to semiduplex mode reduces.

3) used digital vertical sending and receiving beam forming to suppress self-interference

For any polarization, every hurdle of aerial array comprises N/2 antenna element, and has formed vertical subarray.Because each antenna element is independently, phase place and the gain that can digitally control (weighting) each antenna element form the vertical beam in subarray level.Can be by selecting applicable vertical weight to adjust vertical beam, to maximize the isolation between sending and receiving subarray.

With s _m, r _mrepresent respectively the signal being transmitted or received via the vertical subarray m that comprises N/2 antenna element.W _m.nrepresent to be applied in the complex weight (phase place on antenna element (m, n)

For example, Fig. 5 a shows the beam pattern of common dipole antenna elements, and Fig. 5 b shows and the aerial array shown in Fig. 3 carried out to the directly vertical (w of merging _{m, n}=1) the vertical beam figure of vertical subarray afterwards (element separates=0.5 wavelength).Can observe vertical beam figure and become very precipitous after merging.More precipitous wave beam can contribute to the power leakage to its adjacent antenna, and suppresses self-interference.

In practice, can design together vertical beam and traditional horizontal beam, use the 3D beam forming of 2 dimension aerial arrays, minimize self-interference and maximizing compromise between handling up of uplink and downlink thereby can realize.

full duplex sending and receiving scheme

Based on above aerial array, or based on other any antenna or aerial array, second aspect of the present invention provides a kind of communication means for full-duplex communication, comprises the steps:

The self-interference channel of the self-interference that i. estimation transmission downlink information produces reception uplink information;

Ii. according to this estimated self-interference channel, estimation of communication channels;

Iii. according to this estimated communication channel of this estimated self-interference channel, carry out corresponding upward signal reception or downstream signal and send.

First communication channel and self-interference channel are defined below.With H _uLrepresent up (full duplex reception) channel (response), H _dLrepresent descending (full duplex transmission) channel (response).With H _rTor H _{rT, eNB}represent for the self-interference channel (response) to the self-interference of reception antenna at full duplex transmitting antenna, it can be represented as:

$H_{\mathrm{RT}}=\left(\begin{array}{cccc}{h}_{r1,t1}& h_{r1,t2}& ...& h_{r1,\mathrm{tN}}\\ h_{r2,t1}& h_{r2,t2}& ...& h_{r2,\mathrm{tN}}\\ ...& ...& ...& ...\\ h_{\mathrm{rM},t1}& h_{\mathrm{rM},t2}& ...& h_{\mathrm{rM},\mathrm{tN}}\end{array}\right)---\left(1\right)$

Wherein, M represents the quantity of transmitting antenna unit, and N represents the quantity of reception antenna unit.

As long as self-interference can saturated uplink receiver, can measure H according to known reference signal so _uLand H _dL.The saturation power level of receiver by be positioned at receiver front end low noise amplifier (LNA) maximal input level determine.Strong input signal in appearance higher than this level, LNA will stop providing gain, and in fact to signal injection nonlinear distortion.Conventionally, depend on sensitivity requirement, the saturation power level of typical LNA is from-10dBm to-30dBm.When input signal is during higher than LNA saturation power level, some systems can be walked around LNA.This allowed input signal until+4dBm, but cost is low receiver sensitivity.

Suppose that LNA has the saturation power level of typically-10dBm, in conjunction with the isolation of the 45dBm being provided by new antenna array design of the present invention, the maximum transmit power allowing is 35dBm so.This level is higher than the rated power of the radio node of most of commercial.Therefore, new Antenna Design of the present invention ensured any sending and receiving antenna between isolation be enough to avoid the saturation of receiver, to can measure all channel condition informations according to known reference signal, and can use following digital sending and receiving scheme.

First,, in step I, base station eNB should estimate to send self-interference (or claiming loopback (loop back) to disturb) channel of downlink information to the self-interference that receives uplink information and produce.

One preferred embodiment in, will be based on downlink reference signal P _dL=diag (p _dL) estimate self-interference channel, wherein p _dLit is downlink reference signal vector.Be appreciated that because reference signal is P _dLbase station eNB sends, and therefore base station eNB should be known this reference signal P _dLvalue.

More concrete, in step a, base station eNB sends downlink reference signal P _dL, measure uplink receiving signal Y simultaneously _uL, wherein, uplink receiving signal Y _uLin contain the self-interference that downlink reference signal causes.

Use channel model defined above to uplink receiving signal Y _uLcarry out modeling, should obtain

Y _UL＝H _RT·P _DL+H _UL·s _UL+n _UL (2)

Wherein, s _uLbe the uplink information sending, comprise upstream data information or uplink reference signals etc.N _uLit is the noise in up channel.

In step b, based on uplink receiving signal Y _uLand estimation filter, estimate self-interference channel, wherein, described estimation filter is relevant to downlink reference signal.

More specifically, use an estimation filter F _rT, it can be traditional single tap filter (one-tap filter), for example:

$F_{\mathrm{RT}}=\frac{P_{\mathrm{DL}}^H}{P_{\mathrm{DL}}{P}_{\mathrm{DL}}^H}$

By the expressed Y of equation (2) _uLwith F _rTmultiply each other and can obtain $H_{\mathrm{RT},\mathrm{eNB}}+(H_{\mathrm{UL}}{s}_{\mathrm{UL}}+n_{\mathrm{UL}})\frac{P_{\mathrm{DL}}^H}{P_{\mathrm{DL}}{P}_{\mathrm{DL}}^H}$

Observe this equation, consider the following fact, local transmitter is than the more close receiver of remote subscriber transmitter.Self-interference channel power || H _{rT, eNB}|| the upward signal of the expectation far above leached is added to noise power || (H _uLs _uL+ n _uL) F _rT||.Therefore, Y _uLwith F _rTmultiplying each other the value obtaining can be approx as H _{rT, eNB}estimated value , estimated self-interference channel can equal

${\hat{H}}_{\mathrm{RT},\mathrm{eNB}}=Y_{\mathrm{UL}}{F}_{\mathrm{RT}}---\left(3\right)$

Wherein, Y _uLand F _rTall known, so eNB just can calculate estimated self-interference channel

And, can further obtain estimated self-interference channel with self-interference channel H accurately _{rT, eNB}between relation:

${\hat{H}}_{\mathrm{RT},\mathrm{eNB}}=H_{\mathrm{RT},\mathrm{eNB}}+(H_{\mathrm{UL}}{s}_{\mathrm{UL}}+n_{\mathrm{UL}})\frac{P_{\mathrm{DL}}^H}{P_{\mathrm{DL}}{P}_{\mathrm{DL}}^H}---\left(4\right)$

This equation (4) will be used in the time that the self-interference channel to estimated is adjusted, and will be described in more detail below.

In the estimation that obtains self-interference channel afterwards, in described step I i, estimate up channel H _uL.

According to one preferred embodiment, base station eNB is to estimate up channel H according to uplink reference signals _uL.

Concrete, in step I i-1, base station eNB is measured uplink receiving reference signal Y _uL, uplink receiving reference signal Y _uLcomprise uplink reference signals P _uL, and comprise the downlink information x simultaneously sending _dLthe self-interference causing.Wherein, downlink information x _dLcan be downlink reference signal, can be also any downlink data signal.More concrete, the uplink receiving reference signal Y after channel model _uLexpressed as follows

Y _UL＝H _RT·x _DL+H _UL·P _UL+n _UL (5)

Use as H _rTestimation, and x _dLknown.So, this downlink information x of base station eNB based on known _dLwith estimated described self-interference channel from uplink receiving signal Y _uLthe self-interference that the described downlink information of middle removal causes obtain and remove the reception uplink reference signals Y (P disturbing _uL)

$Y\left(P_{\mathrm{UL}}\right)=Y_{\mathrm{UL}}-{\hat{H}}_{\mathrm{RT},\mathrm{eNB}}\·x_{\mathrm{DL}}---\left(6\right)$

Wherein, noise n _uLcan ignore.

In theory, Y (P _uL) the namely uplink reference signals after up channel, i.e. H _uLp _uL.Based on this, in step I i-3, the reception uplink reference signals Y (P that base station eNB is disturbed based on described removal _uL) and another estimation filter, estimate described up channel H _uL, wherein, this another estimation filter is relevant to uplink reference signals.

In an example, use a single-ended filter

be appreciated that due to P _uLbe uplink reference signals, base station eNB also should be able to be known this uplink reference signals in advance.

Up channel H so _uLestimation should calculate by following equation:

${\hat{H}}_{\mathrm{UL}}=Y\left(P_{\mathrm{UL}}\right)\frac{P_{\mathrm{UL}}^H}{P_{\mathrm{UL}}{P}_{\mathrm{UL}}^H}---\left(7\right)$

So far, can obtain estimated up channel

full duplex uplink receiving

Obtaining estimated up channel after, just can carry out the uplink receiving of full duplex.

If subscriber equipment sends uplink information s _uL, in step I ii-1, base station eNB is measured uplink receiving information signal Y _uL, described uplink receiving signal Y _uLcomprise uplink information signal s _uL, and comprise the downlink information x simultaneously sending _dLthe self-interference causing.After channel model, the uplink receiving signal Y that base station eNB receives _uLexpressed as follows

Y _UL＝H _RT·x _DL+H _UL·s _UL+n _UL (8)

In step I ii-2, base station eNB this downlink information based on known and estimated self-interference channel are removed the self-interference that downlink information causes from uplink receiving signal, obtain and remove the reception uplink information signal disturbing.More specifically, use has obtained as H _rTestimation, and x _dLknown.From Y _uLmiddle removal downlink information x _dLthe self-interference causing obtain and remove the reception uplink information signal Y (s disturbing _uL).

$Y\left(s_{\mathrm{UL}}\right)=Y_{\mathrm{UL}}-{\hat{H}}_{\mathrm{RT},\mathrm{eNB}}\·x_{\mathrm{DL}}---\left(9\right)$

Wherein, noise can be ignored.

In theory, Y (P _uL) the uplink information signal after up channel namely, i.e. H _uLs _uL.

Finally, in step I ii-3, use has obtained as H _uLestimation.Just can solve as s _uLestimation.

More concrete, as expressed in following equation

${\hat{s}}_{\mathrm{UL}}=\frac{{\hat{H}}_{\mathrm{UL}}^H}{{\hat{H}}_{\mathrm{UL}}^H{\hat{H}}_{\mathrm{UL}}}Y\left(s_{\mathrm{UL}}\right)---\left(10\right)$

Thus much, full duplex uplink receiving is achieved.

Before this, in the time describing equation (3), (4), once mentioning in the time estimating self-interference channel, is the impact of having ignored lower-powered upward signal.And at this moment, since at up channel and uplink information in equation (7) and equation (10), obtain estimation, so, the present invention more preferably proposes, use estimated go out up channel and uplink information, further estimated self-interference channel is adjusted, made it more accurate.

Concrete, in step x, base station eNB is adjusted estimated described self-interference channel according to equation (4), the self-interference channel after conversion can be adjusted for

${{\hat{H}}^{\′}}_{\mathrm{RT},\mathrm{eNB}}={\hat{H}}_{\mathrm{RT},\mathrm{eNB}}-{\hat{H}}_{\mathrm{UL}}\·{\hat{s}}_{\mathrm{UL}}\·\frac{P_{\mathrm{DL}}^H}{P_{\mathrm{DL}}{P}_{\mathrm{DL}}^H}---\left(11\right)$

At the self-interference channel obtaining after adjusting after, base station eNB can be used the self-interference channel after this adjustment again carry out above step I i and step I ii, obtain up channel more accurately and uplink information then, base station eNB can be used the up channel more accurately of up-to-date acquisition and uplink information again adjust self-interference channel, and repeat this process, until estimated self-interference channel, up channel and uplink information convergence.

the descending transmission of full duplex

Obtaining estimated up channel after, because full duplex is the duplex communication of same frequency simultaneously, so the reciprocity based on upper lower channel can obtain the estimation of down channel

Concrete, in one case, in step I i, base station eNB is with estimated up channel as estimated down channel

In another case, first, base station eNB is using estimated up channel as estimated down channel, and on this basis, base station eNB further improves the accuracy of estimated down channel with channel reciprocity calibration steps.Channel reciprocity calibration is the known a kind of technological means of one of ordinary skill in the art, and the present invention does not repeat them here.

Then, in step I ii, the down channel of base station eNB based on estimated and estimated self-interference channel, determine the delivery plan that downlink information is sent.The satisfied constraint of this delivery plan is: make to send the transmitted power maximum of downlink information, and make to send the self-interference minimum that downlink information produces receiving uplink information.

Then, base station eNB is used this delivery plan to send this downlink information.

In a concrete execution mode, this delivery plan is a pre-coding matrix W _fD.

Send for full duplex, transmitted signal (descending) is:

Y _DL＝H _DLW _FDs _DL+n _DL (12)

Wherein, s _dLdownlink user signal, n _dLdescending noise, and W _fDaccording to full duplex pre-coding matrix of the present invention.

Therefore the uplink receiving signal of, receiving at reception antenna is:

Y _UL＝H _ULs _UL+H _RT，eNBW _FDs _DL+n _UL，(13)

In present embodiment, obtain W with the standard that maximizes downlink transmission power and minimize self-interference power _fD, this can be expressed as optimization problem:

$W_{\mathrm{FD}}=\underset{w_{\mathrm{FD}}}{\arg }\max \left(\frac{{\left|\right|H_{\mathrm{DL}}{W}_{\mathrm{FD}}\left|\right|}^2}{{\left|\right|H_{\mathrm{RT},\mathrm{eNB}}{W}_{\mathrm{FD}}\left|\right|}^2}\right)---\left(14\right)$

Meet the W of above-mentioned standard _fDmaximized signal with self-interference than (SSIR).

This Solve problems is typical vague generalization Characteristic Problem.The solution of this problem is as follows:

W _FD＝GEV(H ^H _DLH _DL，H ^H _RT，eNBH _RT， _eNB) (15)

Wherein, function G EV (A, B) represents to get the maximum vague generalization Eigenvalues Decomposition of matrix A and B.So, obtaining the estimation of down channel with self-interference channel basis on, $W_{\mathrm{FD}}=\mathrm{GEV}({{\hat{H}}^H}_{\mathrm{DL}}{\hat{H}}_{\mathrm{DL}},{{\hat{H}}^H}_{\mathrm{RT}},\mathrm{eNB}{\hat{H}}_{\mathrm{RT},\mathrm{eNB}}).$

simulating, verifying

1) isolation between antennas

In HFSS, set up aerial array pattern according to the design in Fig. 3, and be evaluated at every antenna element between isolation between antennas.

As shown in Figure 6, aerial array is divided into the subarray of two orthogonal polarizations.Antenna port 1 to 8 is with the direction polarization of+45 degree, and antenna port 9 to 16 is with the direction polarization of-45 degree.The latitude of antenna is (x*y*z): 1100mm*240mm*20mm, and this is slightly smaller than conventional 8x2LTE TDD array (1400mm*320mm*100mm).This array is set to operate in 2.4-2.5GHz frequency range.

Assessment result is as follows:

orthogonal polarization port

Adjacent port:

S(8，9)＜-45dB S(4，9)＜-47dB

S(4，13)＜-53dB S(8，13)＜-58dB

1 port apart:

S(4，10)＜-54dB S(4，14)＜-55dB

S(8，10)＜-50dB S(8，14)＜-58dB

S(9，3)＜-54dB S(9，7)＜-56dB

S(13，3)＜-56dB S(13，7)＜-62dB

co-polarization port

Adjacent port:

S(4，3)＜-26dB S(4，8)＜-25dB

Note, S (a, b) is illustrated in the S parameter between port a and b.

Fig. 7 shows exemplary separate evaluation diagram.According to the result of Fig. 7, in this array, cross-polarized any antenna element between minimum isolation between antennas be 45dB, this is than the high 20dB of isolation between two co-polarization adjacent antennas.

2) full duplex sending and receiving scheme

In order to verify the performance of full duplex sending and receiving scheme, be provided with full duplex simulator.Configure as shown in the following table parameters:

Three scenes are carried out to emulation: 1) half-duplex scene; 2) do not have transmission full duplex scene (for example W _fD=1); And 3) utilize the full duplex scene of new transmission plan (for example to calculate W with equation 15 _fD).Suppose that, in scene 2 and 3, the new aerial array of the application of the invention provides main isolation 45dB, so that receiver can be unsaturated.

First, the inhibition that checking is provided by full duplex transmission scheme of the present invention, as shown in Figure 8.

As seen from Figure 8, can realize by full duplex transmission scheme of the present invention (TX inhibition) the self-interference inhibition of about 18dB.It should be noted that this inhibition ability is almost independent of upstream or downstream channel quality.This inhibition ability mainly determined by the accuracy that self-channel is estimated, this is independent of upstream or downstream channel quality.Can affect this inhibition ability and realize extra self-interference inhibition, and increase the full duplex quality of reception.

Subsequently, will full duplex transmission be assessed the impact of handling up.Be different from half-duplex system (therein unidirectional handle up be used as evaluation criteria), will use together uplink and downlink to handle up at this, be i.e. two-way handling up as evaluation criteria.

As shown in Figure 9, with respect to half-duplex operation, full duplex can be brought about 30% throughput gain.Although it can provide significant self-interference to suppress, but as cost, use the W calculating according to equation 15 _fDdigital self-interference suppress to cause the decline of handling up, this is especially obvious in the time that channel SNR is lower.Reason is, for multiaerial system, the degree of freedom of whole Spatial Dimension is determined by antenna configuration and channel, therefore fixed for given scene.The operation that self-interference suppresses needs the spatial degrees of freedom for downlink data transmission originally, has therefore caused the descending decline of handling up.

conclusion

Embodiments of the present invention have proposed to have the new full duplex scheme of following design main points:

1. there is a kind of new antenna array design of the sending and receiving antenna element of cross polarization and vertical separation.

2. a new sending and receiving scheme, its aerial array based on new and uplink/downlink/feeding back channel state information is measured and is had the self-interference of transmission and suppress and be received to disturb to suppress to eliminate function.

According to emulation provided by the present invention, new design can realize the isolation of 45dB, this full duplex transmission for conventional radio node enough, and with respect to half-duplex operation, full duplex can be brought about 30% throughput gain.

It should be noted that, above-described embodiment is only exemplary, but not limitation of the present invention.According to full duplex sending and receiving scheme of the present invention and do not rely on aerial array according to the present invention and exist, and go for estimating self-interference channel, up channel and the down channel in any full duplex system, and and then realize this full duplex and receive and dispatch scheme.

Within any technical scheme that does not deviate from spirit of the present invention all should fall into protection scope of the present invention, this comprises the different technologies feature that use occurs in different embodiment, and installation method can combine, to obtain beneficial effect.In addition, any Reference numeral in claim should be considered as limiting related claim; " comprise " that a word do not get rid of device unlisted in other claims or specification or step.

Claims (15)

1. an aerial array, is characterized in that, comprising:

The-the first subarray, has multiple first day line elements;

The-the second subarray, has multiple the second antenna elements;

At one time, one in described the first subarray and described the second subarray for transmitted signal, and another is for receiving signal, to carry out full-duplex communication,

Wherein, described the first subarray separates with the second subarray, and the polarised direction of the polarised direction of this first day line element of described the first subarray and the second antenna element of described the second subarray is orthogonal.

2. aerial array according to claim 1, is characterized in that,

Different transmitting-receiving units, or receiver, or transmitter is connected respectively to each first day line element, and,

Different transmitting-receiving units, or receiver, or transmitter is connected respectively to each the second antenna element.

3. aerial array according to claim 2, it is characterized in that described transmitting-receiving unit, or receiver, or the signal that transmitter receives/sends each first day line element is respectively weighted control, be shaped with the vertical beam of the sending/receiving that realizes the first subarray;

Described transmitting-receiving unit, or receiver, or the signal that transmitter receives/sends each the second antenna element is respectively weighted control, be shaped with the vertical beam of the sending/receiving that realizes the second subarray.

4. aerial array according to claim 3, is characterized in that, described vertical beam is shaped for being reduced to the Power leakage of adjacent antenna unit, and suppresses the self-interference between the first subarray and the second subarray.

5. aerial array according to claim 3, is characterized in that, described the first subarray and the second subarray are also configured to carry out horizontal beam shaping, realize the beam forming in three dimensions, to reduce the power leakage of adjacent antenna unit, suppress self-interference, improve user simultaneously and accept performance.

6. for a communication means for full-duplex communication, it is characterized in that, comprise the steps:

The self-interference channel of the self-interference that i. estimation transmission downlink information produces reception uplink information;

Ii. according to this estimated self-interference channel, estimation of communication channels;

Iii. according to this estimated communication channel of this estimated self-interference channel, carry out corresponding upward signal reception or downstream signal and send.

7. method according to claim 6, is characterized in that, described step I comprises:

A. send downlink reference signal, measure uplink receiving signal simultaneously, wherein, in uplink receiving signal, contain the self-interference that downlink reference signal causes;

B. based on described uplink receiving signal and an estimation filter, estimate described self-interference channel, wherein, described estimation filter is relevant to downlink reference signal.

8. method according to claim 7, is characterized in that, in described step b, described estimation filter comprises single-ended filter, and the estimated matrix of this filter is

wherein, P _dLthe vector of downlink reference signal,

Estimated described self-interference channel is

wherein, Y _uLbe described uplink receiving signal, it comprises downlink reference signal.

9. according to the method described in claim 7 or 8, it is characterized in that, is to estimate up channel in described step I i, and this step I i comprises:

Ii-1. measure uplink receiving reference signal, described uplink receiving reference signal comprises uplink reference signals, and comprises simultaneously the self-interference that the downlink information that sends causes;

Ii-2. this downlink information based on known and estimated described self-interference channel are removed the self-interference that described downlink information causes from uplink receiving signal, obtain and remove the reception uplink reference signals disturbing;

The reception uplink reference signals that ii-3 disturbs based on described removal and another estimation filter, estimate described up channel, and wherein, described another estimation filter is relevant to uplink reference signals.

10. method according to claim 9, is characterized in that, in described step I i-3, described another estimation filter comprises single-ended filter, and the estimated matrix of this filter is

wherein, P _uLthe vector of uplink reference signals,

Estimated described up channel is

wherein, Y (P _uL) be to remove the described reception uplink reference signals disturbing.

11. methods according to claim 9, is characterized in that, are to carry out uplink communication based on estimated up channel in described step I ii, and described step I ii comprises:

Iii-1. measure uplink receiving information signal, described uplink receiving information signal comprises uplink information signal, and comprises simultaneously the self-interference that the downlink information that sends causes;

Iii-2. this downlink information based on known and estimated described self-interference channel are removed the self-interference that described downlink information causes from uplink receiving signal, obtain and remove the reception uplink information signal disturbing;

Iii-3. the described up channel based on estimated recovers uplink information from remove the reception uplink information signal disturbing.

12. methods according to claim 11, is characterized in that, also comprise the steps:

X. according in described step I i estimated go out described up channel and described step I ii in recover uplink information, adjust estimated described self-interference channel;

Y. according to the estimated described self-interference channel after adjusting, again carry out described step I i and estimate that up channel and described step I ii recover uplink information;

Repeat above step x and described step y, until estimated self-interference channel, up channel and uplink information convergence.

13. methods according to claim 9, in described step I i also using estimated described up channel as estimated down channel;

Or, in described step I i:

-using estimated described up channel as estimated down channel;

-on this basis, improve the accuracy of estimated down channel with channel reciprocity calibration steps.

14. methods according to claim 13, is characterized in that, described step I ii comprises:

-the down channel based on estimated and estimated self-interference channel, determine the delivery plan that downlink information is sent;

-use this delivery plan to send this downlink information;

Wherein, this delivery plan meets following constraint:

-make to send the transmitted power maximum of downlink information, and make to send the self-interference minimum that downlink information produces receiving uplink information.

15. methods according to claim 14, is characterized in that, described delivery plan comprises pre-coding matrix, and described pre-coding matrix is

W _fD=GEV (H ^h _dLh _dL, H ^h _{rT, eNB}h _rT, _eNB), wherein GEV () is the maximum vague generalization Eigenvalues Decomposition of getting two matrixes, H _dLestimated down channel, H _rT, _eNBdescribed estimated self-interference channel.