CN103117975B - System of compensating MU-MAS communications and dynamically adapting communication characteristics of MU-MAS communication system - Google Patents

Abstract

The invention provides a system of compensating MU-MAS communications and dynamically adapting communication characteristics of an MU-MAS communication system. The system of compensating the MU-MAS communications comprises one or more coded modulation unit(s) used for coding and modulating information bits of each wireless client device of a plurality of the wireless client devices so as to generate coded and modulated information bits; one or more mapping unit(s) used for mapping the coded and modulated information bits into complex symbols; and an MU-MAS frequency / phase shift sensing pre-coding unit used for calculating MU-MAS frequency / phase shift sensing precoding weight by using channel condition information obtained from the wireless client devices through feedback. The MU-MAS frequency / phase shift sensing pre-coding unit uses the weight to pre-code the complex symbols obtained from the mapping unit(s) so as to pre-eliminate frequency/phase shift and/or interference between users.

Description

The communication characteristic of compensation MU-MAS communications and dynamically adapting MU-MAS communication systems System

The application be the applying date on 08 20th, 2008, it is Application No. 200880102933.4, entitled " distributed defeated The divisional application of the application for a patent for invention of the system and method for entering distributed output radio communication ".

Priority request

The application is the continuation application of the application NO.10/902,978 that July 30 in 2004 submits to.

Technical field

The present invention relates generally to field of wireless communications.Especially, the present invention relates to be used to use dividing for Space-time coding techniques The system and method for the radio communication of the distributed output of cloth input.

Background technology

The space-time code of signal of communication

Space multiplexing and space-time code are known newer development in wireless technology.Due to there is several antennas to be used in each Terminal, so a kind of space-time code of specific type is referred to as " multi-input multi output "（MIMO）.By using multiple days Line is sent and received, and multiple independent radio waves can be simultaneously transmitted in the identical frequency range.Following article is carried The general introduction of MIMO is supplied.

IEEE member David Gesbert, IEEE member Mansoor Shafi, IEEE member Da-shanShiu, IEEE The IEEEJOURNAL ON SELECTED AREAS IN of the senior member Ayman Naguib of member Peter J.Smith and IEEE COMMUNICATIONS,VOL.21,NO.3,APRIL 2003:“From theory to Practice:An Overview of MIMO Space-TimeCoded Wireless Systems”。

IEEE member David Gesbert, IEEE members Helmut Bolcskei, Dhananijay A.Gore and IEEE IEEE TRANSCTIONS ONCOMMUNICATIONS, VOL.50, the NO.12 of member Arogyaswami J.Paulraj, DECEMBER 2000:”Outdoor MIMOWireless Channels:Models and Performance Prediction”。

Substantially, MIMO technology is based on the spatially distributed day for producing coordinate spaces data flow in common band The application of line.Radio wave is propagated in this way, such that it is able to being separated in receiver and demodulating individual signals, even if they In identical transmitted in band, this can cause independent on multiple statistical significances（Namely efficiently separate）Communication channel. Therefore, compared with the standard wireless communication system for making great efforts to suppress multipath signal（That is, multiple delay time signals of same frequency, and shake There is modification in width and phase）, MIMO may rely on irrelevant or weak related multipath signal, realized in given frequency band compared with Throughput high and the signal to noise ratio for improving.Example shows, in power and signal to noise ratio（SNR）Under conditions of quite, MIMO technology reality Handling capacity higher is showed（throughput）, and traditional non-mimo system can only realize relatively low handling capacity.High pass is public Department（High pass is maximum wireless technology supplier）Websitehttp://www.cdmatech.com/products/what mimo delivers.jsp:This function is described on the page of subscript entitled " What MIMO Delivers "：“MIMO is the onlymultiple antenna technique that increases spectral capacity by delivering two ormore times the peak data rate of a system per channel or per MHz of spectrum.To be more specific,for wireless LAN or applications QUALCOMM'sfourth generation MIMO technology delivers speeds of 315Mbps in 36MHz ofspectrum or 8.8Mbps/MHz.Compare this to the peak capacity of 802.11a/ g(even with beam-forming or diversity techniques)which delivers only 54Mbps in17MHz of spectrum or 3.18Mbps/MHz”。

On generally, due to several reasons, mimo system is facing to each device less than 10 actual property limitations of antenna（Cause Improvement in this network is less than 10 × throughput）：

1. physical limit：There must be enough intervals between mimo antenna on setter, so that each is received The signal of statistical iteration.Even if although still can see the improvement of MIMO handling capacities in the antenna spacing of wavelength fraction, When antenna is more nearly, efficiency deteriorates rapidly, which results in relatively low MIMO throughput multiplication devices.

See, for example, below with reference to document：

[1] D.-S.Shiu, G.J.Foschini, M.J.Gans, and J.M.Kahn, " Fadingcorrelation And its effect on the capacity of multielement antenna systems, " IEEE Trans, Comm., vol.48, no.3, pp.502-513, Mar.2000.

[2] V.Pohl, V.Jungnickel, T.Haustein, and C.von Helmolt, " Antennaspacing In MIMO indoor channels, " Proc.IEEE Veh.Technol.Conf., vol.2, pp.749-753, May 2002.

[3] M.Stoytchev, H.Safar, A.L.Moustakas, and S.Simon, " Compactantenna Arrays for MIMO applications, " Proc.IEEE Antennas and Prop.Symp., vol.3, pp.708- 711, July 2001.

[4] A.Forenza and R.W.Heath Jr., " Impact of antenna geometry onMIMO Communication in indoor clustered channels, " Proc.IEEE Antennasand Prop.Symp., Vol.2, pp.1700-1703, June 2004.

Additionally, for small antenna spacing, coupling effect each other may reduce the performance of mimo system.

See, for example, below with reference to document：

[5] M.J.Fakhereddin and K.R.Dandekar, " Combined effect ofpolarization Diversity and mutual couplingon MIMO capacity, " Proc.IEEEAntennas and Prop.Symp., vol.2, pp.495-498, June 2003.

[7] P.N.Fletcher, M.Dean, and A.R.Nix, " Mutual coupling in multi-element Array antennas and its influence on MIMO channel capacity, " IEEEElectronics Letters, vol.39, pp.342-344, Feb.2003.

[8] V.Jungnickel, V.Pohl, and C.Von Hel molt, " Capacity of MIMOsystems With closely spaced antennas, " IEEE Comm.Lett., vol.7, pp.361-363, Aug.2003.

[10] J.W.Wallace and M.A.Jensen, " Termination-dependent diversityperformance of coupled antennas：Network theory analysis, " IEEE Trans.Antennas Propagat., vol.52, pp.98-105, Jan.2004.

[13] C.Waldsch midt, S.Schulteis, and W.Wiesbeck, " Complete RFsystem Model for analysis of compact MIMO arrays, " IEEE Trans.on Veh.Technol., vol.53, Pp.579-586, May 2004.

[14] M.L.Morris and M.A.Jensen, " Network model for MIMO systemswith Coupled antennas and noisy amplifiers, " IEEE Trans.AntennasPropagat., vol.53, Pp.545-552, Jan.2005.

And, when antenna it is crowded to together when, antenna generally has to be made smaller, and this can also influence antenna to imitate Rate.

See, for example, below with reference to document：

[15] H.A.Wheeler, " Small antennas, " IEEE Trans.AntennasPropagat., vol.AP- 23, n.4, pp.462-469, July 1975.

[16] J.S.McLean, " A re-examination of the fundamental limits on Theradiation Q of electrically small antennas, " IEEE Trans.Antennas Propagat., Vol.44, n.5, pp.672-676, May 1996.

Finally, with lower frequency and longer wavelength, the physical size of MIMO devices just becomes to be difficult to process.One pole The example at end is that, in HF wave bands, MIMO devices antenna must 10 meters separated from each other or bigger distance here.

2. noise limitation.The receiver of each MIMO/transmitter subsystem produces the noise of certain level.When more and more This subsystem when closing on placement mutually, ambient noise will rise.Meanwhile, when needs are recognized from multiple-antenna MIMO system When going out more unlike signals, it is desirable to lower ambient noise.

3. cost and power limit.Although cost and power consumption are not focuses in some MIMO applications, typical wireless In product, when developing a kind of successful product, cost and power consumption are all it is critical that restraining factors.For each MIMO days Line, it is necessary to separate RF subsystems, including separate analog-to-digital（A/D）Sum-mould（D/A）Converter.Unlike being come with Moore's Law Many aspects of the digital display circuit of measurement scale（The observation of the experience aspect done by cofounder Gordon mole of Intel As a result, the transistor size on the integrated circuit of microdevice will quadruple per every about 24 months；Source：http:// www.intel.com/technology/mooreslaw/）, so intensive analog subsystem is generally with certain physics knot Structure size and power requirement, its size are proportional to cost and power linear.Therefore, compared with single antenna devices, multiple antennas MIMO devices will become extremely expensive and with surprising energy consumption.

Used as result above, most of mimo systems expected from today are in 2 to 4 grades of antenna, to cause to gulp down The rising of 2 to 4 times of the amount of telling and some SNR caused due to the diversity benefit of multiaerial system（Signal to noise ratio）Rising. Anticipate 10 mimo systems of antenna（Especially because shorter wavelength and nearer antenna spacing in Microwave Frequency higher In rate）But, it is very unactual more than 10 antennas in addition to application special for some and insensitive to cost 's.

Virtual antenna array

A kind of special applications of the technology of MIMO type are virtual antenna arrays.Euroscience technical field research cooperative groups This system, EURO, Barcelona, Spain, the 15-17 of in January, 2003 days are suggested in the research file for knitting proposition：Center for Telecommunication Research,King’s CollegeLondon,UK:”A step towards MIMO: Virtual Antenna Arrays”,Mischa Dohler&Hamid Aghvami。

As described in file, virtual antenna array is cooperation wireless device system（Such as cell phone）, it is in separation Communicated in communication channel（If when they mutually close on enough）, rather than in their main communication channels with they Base station communication so that collaborative ground work（If for example, they are the GSM cell phones in UHF waveband, then this can be with It is the industrial scientific medical of 5GHz（ISM）Wireless wave band）.By in mutual relaying scope（Except in base station range）Interior Forwarding information between several devices, just looks like that they are that physically have the device work of multiple antennas the same so that Single antenna devices potentially realize the throughput hoisting as MIMO.

However, in fact, such system it is extremely difficult realization and use is limited.First of all, it is necessary to keep each device present A minimum of two different communication paths are lifted with realizing throughput, and the availability of its second repeated link is often uncertain 's.Being additionally, since them at least has the second communication subsystem and has bigger calculating demand, therefore the device is more expensive, Physical size is bigger, and consumes more power.Additionally, potentially through multiple communication links, the system is depended on very The complicated systematic live collaboration of institute.Finally, due to simultaneous channel usage increases（For example, using MIMO technology Simultaneous phone call transmission）, the computation burden for each device also increases（Generally with the linear increasing of channel usage Plus and exponentially increases）, this is very unpractiaca to the portable unit limited with strict power and size.

The content of the invention

The invention provides a kind of frequency and phase offset for compensating multi-user multi-aerial system MU-MAS communication System, the system includes：One or more coded modulation units, for wireless for each in multiple wireless client devices The information bit of customer set up is encoded and modulated with the information bit after generating coding and modulation；One or more mappings are single Unit, for the information bit after the coding and modulation to be mapped as into complex symbol；And the skew of MU-MAS frequency/phases is perceived Precoding unit, for calculating MU-MAS frequently using by feeding back the channel condition information obtained from the wireless client device Rate/phase offset perceives precoding weight, and the MU-MAS frequency/phases skew perceives precoding unit and uses the weight pair The complex symbol obtained from the map unit carries out precoding to eliminate frequency/phase skew and/or inter-user interference in advance.

Present invention also offers a kind of inphase quadrature for compensating multi-user multi-aerial system MU-MAS communications（I/Q）No The system of balance, the system includes：One or more coded modulation units, for every in for multiple wireless client devices The information bit of individual wireless client device is encoded and modulated with the information bit after generating coding and modulation；One or more Map unit, for the information bit after the coding and modulation to be mapped as into complex symbol；And MU-MAS IQ are perceived and prelisted Code unit, for using calculating by feeding back the channel condition information obtained from the wireless client device MU-MAS IQ senses Know precoding weight, the MU-MAS IQ perceive precoding unit using the weight to answering for being obtained from the map unit Numerical symbol is carried out precoding and eliminates the interference and/or inter-user interference that are brought due to I/Q gains and unbalance in phase with pre-.

Present invention also offers a kind of communication special for dynamically adapting multi-user multi-aerial system MU-MAS communication systems The system of property, the system includes：One or more coded modulation units, for for each in multiple wireless client devices The information bit of wireless client device is encoded and modulated with the information bit after generating coding and modulation；One or more reflect Unit is penetrated, for the information bit after the coding and modulation to be mapped as into complex symbol；And MU-MAS configurator units, use Sent out in based on the subset that user is determined by feeding back the channel characteristics data obtained from the wireless client device and MU-MAS Pattern is sent, and responsively controls the coded modulation unit and the map unit.

Describe a kind of being used for multi-user（MU）Send（“MU-MAS”）Multiaerial system（MAS）In frequency The system and method compensated with phase offset.For example, the method according to one embodiment of the present invention includes：Will be from base One or each wireless client device that the training signal of each antenna of standing is sent into multiple wireless client devices, client dress One in putting or each customer set up analyze each training signal to generate frequency offset compensation data, and are received in base station Frequency offset compensation data；MU-MAS precoders weight is calculated based on the frequency offset compensation data to launch with pre- elimination Frequency shift (FS) at machine；Precoding is carried out to training signal using the MU-MAS precoders weight, base station is directed to generate The precoding training signal of each antenna；Training signal after the precoding of each antenna from the base station is sent to institute Each wireless client device in multiple wireless client devices is stated, each customer set up analyzes each training signal to generate letter Road characteristic, and receive the channel characteristics data in the base station；Multiple MU- are calculated based on the channel characteristics data MAS precoding weights, the MU-MAS precoder weights be calculated to for it is pre- eliminate frequency and phase offset and/or user it Between interference；Precoding is carried out to data using MU-MAS precoders weight, to generate for the pre- of each antenna in base station Data-signal after coding；And pre-coded data signal after precoding is sent to it by each antenna of base station The client device of each.

Brief description of the drawings

With reference to accompanying drawing, below in detail description can obtain the present invention is better understood from, wherein：

Fig. 1 shows the mimo system of prior art.

Fig. 2 shows the N antenna base stations communicated with multiple single antenna customer set ups.

Fig. 3 shows three base stations of antenna communicated with three single antenna customer set ups.

Fig. 4 shows the training signal technology used in one embodiment of the present of invention.

Fig. 5 shows the channel characteristics data that base station is transferred to from customer set up according to an embodiment of the invention.

Fig. 6 shows the distributed output of multiple input according to an embodiment of the invention（“MIDO”）Downlink transfer.

Fig. 7 shows multi-input multi output according to an embodiment of the invention（“MIMO”）Uplink.

Fig. 8 show it is according to an embodiment of the invention by different clients group circulation to distribute the base of handling capacity Stand.

Fig. 9 shows according to an embodiment of the invention based on the custom partitioning for closing on.

Figure 10 shows the embodiments of the invention used in NVIS systems.

Figure 11 shows the implementation method of the DIDO emitters with I/Q compensation function units.

Figure 12 shows the DIDO receivers with I/Q compensation function units.

Figure 13 shows a kind of implementation method of the DIDO-OFDM systems with I/Q compensation.

Figure 14 shows the performances of DIDO 2 × 2 in the case where being compensated with and without I/Q（performance）One Plant implementation method.

Figure 15 shows a kind of implementation method of the performances of DIDO 2 × 2 in the case where being compensated with and without I/Q.

Figure 16 shows the SER for different Q AM planispheres in the case where being compensated with and without I/Q（Symbol is missed Code check）A kind of implementation method.

Figure 17 shows DIDO2 × 2 performance in the case where different user devices position has and is compensated without I/Q A kind of implementation method.

Figure 18 is shown in ideal（i.i.d.（Independent and same distribution））Have a case that in channel and without I/Q compensation A kind of implementation method of lower SER.

Figure 19 shows a kind of implementation method of the transmitter architecture of self adaptation DIDO systems.

Figure 20 shows a kind of implementation method of the receiver architecture of self adaptation DIDO systems.

Figure 21 shows a kind of implementation method of the method for self adaptation DIDO-OFDM.

Figure 22 shows a kind of implementation method of the antenna arrangement for DIDO measurements.

Figure 23 is shown for different stage（order）The implementation method of the array configuration of DIDO systems.

Figure 24 shows the performance of different stage DIDO systems.

Figure 25 shows a kind of implementation method of the aerial array for DIDO measurements.

Figure 26 shows the one of the performances of DIDO 2 × 2 of 4-QAM and 1/2FEC rates and the functional relation of location of user equipment Plant implementation method.

Figure 27 shows a kind of implementation method of the antenna arrangement for DIDO measurements.

Figure 28 shows how producing ratio has the DIDO 2 of low TX power demands for DIDO 8 × 8 in one embodiment × 2 bigger SE.

Figure 29 shows a kind of implementation method of the performances of DIDO 2 × 2 in the case of with day line options.

Figure 30 shows average BER of the different DIDO pre-coding schemes in i.i.d. channels（BER）Performance.

Figure 31 shows the function between the quantity of extra transmitting antenna in the snr gain and i.i.d. channels of ASel Relation.

Figure 32 shows that SNR threshold values are diagonal with for block in the case of having 1 and 2 exterior antenna in i.i.d. channels Change（BD）With the number of users of ASel（M）Between functional relation.

Figure 33 is shown for positioned at equal angular direction and with different angle spreads（AS）Two BER of user of value SNR average with every user.

Figure 34 shows the result similar with Figure 33, but there is angle higher to be spaced between user.

Figure 35 depicts the mean angle of arrival for user（AOA）Different value, function between AS and SNR threshold values closes System.

Figure 36 is shown for 5 SNR threshold values of the exemplary cases of user.

Situations of the Figure 37 for 2 users, there is provided in the case of with 1 and 2 additional antenna, SNR threshold values BD with The comparing of ASel.

Figure 38 shows the result similar with Figure 37, but is directed to 5 situations of user.

Figure 39 shows the SNR threshold values for the BD schemes with different AS values.

Figure 40 is shown for 1 and 2 BD and ASel of additional antenna, with AS=0.1 ° of space correlation letter SNR threshold values in road.

Figure 41 shows the calculating of the SNR threshold values of two other channel conditions for AS=5 °.

Figure 42 shows the calculating of the SNR threshold values of two other channel conditions for AS=10 °.

Figure 43-Figure 44 is respectively illustrated in the case of 1 and 2 additional antenna, SNR threshold values and number of users（M）With BD With the angle spread of ASel schemes（AS）Between functional relation.

Figure 45 shows the receiver equipped with frequency offset estimator/compensator；

Figure 46 shows the system models of DIDO 2 × 2 according to one embodiment of the present invention.

Figure 47 shows the method according to one embodiment of the present invention.

Figure 48 shown in the case of with and without frequency shift (FS), the SER results of the systems of DIDO 2 × 2.

Figure 49 compares the SNR threshold performances of different DIDO schemes.

Figure 50 compares the amount of overhead needed for distinct methods implementation method.

Figure 51 is shown in f_max=2Hz small frequency skew and without integer offset correction in the case of emulation.

Figure 52 shows the result when integer offset estimator is closed.

Specific embodiment

In the following description, for purposes of explanation, the present invention is thoroughly understood in order to provide, it is multiple special to illustrate Details.It may be evident, however, that for one of ordinary skilled in the art, even if without some specific details, still can be with Realize the present invention.Additionally, known construction and device is shown as block diagram format, to avoid obscuring ultimata of the present invention Change.

Fig. 1 shows mimo system in the prior art with transmitting antenna 104 and reception antenna 105.Such system Throughput can realize 3 times of the throughput generally realized in available channel.There are various different methods this to realize The details of mimo system, it had description in the published document on the theme, and explanation below will be described as one Method.

Before data are transmitted in the mimo system of Fig. 1, channel is by " characterization ".This is by starting " training letter Number " from each transmitting antenna 104 be transferred to each receiver 105 to realize.Training signal has coding and mod subsystem 102 Generation, and by D/A converter（It is not shown）Analog signal is converted into, is then changed from baseband signal by each transmitter 103 It is RF signals.The reception antenna 105 for being each coupled to its RF receiver 106 receives each training signal and is converted into base band Signal.Baseband signal is by D/A converter（It is not shown）Data signal is converted to, then signal processing subsystem 107 is characterized The training signal.The feature of each signal can include several factors, for example, it includes, relative to the reference inside receiver The phase and amplitude of signal, absolute reference signal, relative reference signal, characteristic noise or other factors.The feature of each signal It is normally defined the vector of phase and amplitude variations that the several aspects of signal are showed when signal is transmitted by channel.For example, Quadrature amplitude modulation（“QAM”）Modulated signal in, the feature is probably the phase and amplitude of several multipath images of signal The vector of skew.Another example is, in OFDM（“OFDM”）In the signal of modulation, it is probably OFDM frequency spectrums In several or all single component signals（sub-signal）Phase and amplitude excursion vector.

Signal processing subsystem 107 deposits the channel characteristics received by each reception antenna 105 and corresponding receiver 106 Storage is got up.All of three transmitting antennas 104 are completed after their training signal transmission, and signal processing subsystem 107 is by Stored three for the channel characteristics of each in three reception antennas 105, this results in 3 × 3 matrix 108, its table It is shown as channel characteristics matrix " H ".Each single matrix element Hi, j are the transmission antenna 104i that reception antenna 105j is received Training signal transmission channel characteristics.

At this point, signal processing subsystem 107 inverts matrix H 108 to produce H^-1, and wait from transmitting antenna The transmission of 104 real data.Note, various existing MIMO technologies described in available document can be used to ensure H-matrix 108 is reversible.

In force, the content of the data to be transmitted（payload）It is sent to data input subsystem 100.Then it is being sent to Before coding and mod subsystem 102, its allocated device（splitter）101 are divided into three parts.If for example, content is The ASCII bits of " abcdef ", it can just be allocated device and be divided into three sub- contents " ad ", " be " and " cf ".Then, each Sub- content is individually sent to coding and mod subsystem 102.

By using the statistical independence and the coded system of error correcting capability that are adapted to each signal, individually in every height Appearance is encoded.These include, and are not limited only to, Reed-Solomon codings, Viterbi coding（Viterbi coding） Encoded with enhancing（Turbo Codes）.Finally, using to the suitable modulator approach of channel to these three encode after sub- content in Each be modulated.Exemplary modulator approach is differential phase keying (DPSK) modulation（“DPSK”）, 64-QAM modulation and OFDM. It should be noted there that the diversity gain that MIMO is provided allows the modulation constellation of higher plate number, the modulation constellation to exist Use the SISO of same channel（Single-input single-output）It is also feasible in system.Then, each is encoded and the letter after modulating Number transferred out by the antenna 104 of its own, the transmission is followed in D/A converting units（It is not shown）D/A conversion and After the RF generations of each transmitter 103.

Assuming that have enough space diversitys to be present in sending and receiving between antenna, each reception antenna 105 will be from antenna 104 receive three various combinations of transmission signal.Each signal is received and converts them to base by each RF receiver 106 Band signal, then A/D converter（It is not shown）Signal is digitized again.If y_nIt is to be connect by n-th reception antenna 105 The signal for receiving, x_nIt is the signal sent by n-th transmitting antenna 104, N is noise, then this can just be described with following equalities.

y₁=x₁H₁₁+x₂H₁₂+x₃H₁₃+N

y₂=x₁H₁₂+x₂H₂₂+x₃H₂₃+N

y₃=x₁H₁₃+x₂H₃₂+x₃H₃₃+N

Assuming that this is one has three three systems of equation of unknown quantity, then here it is signal processing subsystem 107 derive x₁、x₂And x₃Linear algebra problem（Assuming that N is in sufficiently low level, it is allowed to which signal is decoded）：

x₁=y₁H^-1 ₁₁+y₂H^-1 ₁₂+y₃H^-1 ₁₃

x₂=y₁H^-1 ₂₁+y₂H^-1 ₂₂+y₃H^-1 ₂₃

x₃=y₁H^-1 ₃₁+y₂H^-1 ₃₂+y₃H^-1 ₃₃

Once derive three signal x of transmission_n, they are just demodulated by signal processing subsystem 107, decode and error correction, To recover original three bit streams separated by distributor 101.These bit streams merge in combiner unit 108, and from Single data stream is output as in data output 109.Assuming that system robust can overcome noise induced damage, then data output 109 is produced Raw bit stream is by as the bit stream being incorporated into data input 100.

Although described prior art systems are generally effectively until four antennas, perhaps until the antenna of as many as 10, Due to described in the background parts in the disclosure, with a large amount of antennas（Such as 25,100 or 1000）When it becomes very not It is actual.

Generally, such prior art systems are two-way, and return path is realized in an identical manner, but anti-mistake Come, all have in every side of communication channel and send and receive subsystem.

Fig. 2 shows one embodiment of the present of invention, wherein, base station（BS）200 are configured with wide area network（WAN）Interface （For example by T1 or other connections at a high speed）201 and it is provided with a number of（It is N number of）Antenna 202.We use art for the time being Language " base station " carries out any wireless site of radio communication to refer to one group of client of fixed position.The example of base station can be nothing Line LAN（WLAN）In access point, or WAN antenna or aerial array.There is some customer set ups 203-207, each has single Antenna, base station 200 is wirelessly serviced them.Although for the purpose of this example, being very easy to expect position In the base station of office environment, in this environment, it is the user's set 203- of the personal computer for being equipped with wireless network 207 provide services, but this structure will apply to substantial amounts of applicable cases, indoor and outdoors, and base station is served wirelessly herein Client.For example, the base station may be located on cellular tower, or in television broadcast towers.In one embodiment, base Stand and 200 be placed in ground, for HF frequencies（The frequency of such as 24MHz）Up transmission, by signal from ionospheric reflection Return, as on April 20th, 2004 proposes, Serial No. No.10/817,731, entitled SYSTEM AND METHOD FOR ENHANCING NEAR VERTICALINCIDENTCE SKYWAVE(“NVIS”)COMMUNICATION USINGSPACE- While TIME CODING pending application description as, its by the agent of branch dispensing the application, herein as ginseng Examine.

Some details being associated with base station 200 and illustrated customer set up are used for the purpose of the purpose of illustration, without Necessary to be cardinal principle of the invention.For example, the base station can be connected to multiple different types via wan interface 201 Wide area network, it includes private wide area network, and for example those are used for the wide area network that digital video sends.Similarly, customer set up can To be any kind of wireless data processing and/or communicator, it includes, and is not limited solely to, cell phone, individual number Word assistant（“PDA”）, receiver and wireless camera.

In one embodiment, the n antenna 202 of base station is spatially separated, so that each sends and receives The related signal of non-space, just look like the base station be prior art MIMO transceiver it is the same.As retouched in the introduction State, antenna is with λ 6（That is 1/6 wavelength）The experiment being spaced apart has been made, and it successfully realizes and is carried from the handling capacity of MIMO But in general rise, these antenna for base station are got over and are spaced further apart placing, and the performance of system is better, λ 2 be it is gratifying most Small distance.Certainly, cardinal principle of the invention any specific separation between being not limited to antenna.

Note, its antenna can be positioned over far distance by single base station 200 well.For example, in HF frequency spectrums, antenna There can be 10 meters or farther（For example, during the NVIS being generally noted above is realized）.If using 100 such antennas, the base station Aerial array can just occupy several square kilometres of area.

In addition to space diversity technology, in order to improve the effective throughput of system, one embodiment of the present of invention will be believed Number polarization.It is a kind of known technology that channel capacity is improved by polarizing, and it has been used very by satellite television providers For many years.Using Polarization technique, multiple can be made（Such as three）Base station or user antenna with each other closely, and are still Non-space correlation.Although traditional RF systems mostly just benefit from the two dimension of polarization（Such as x and y）Diversity, but retouch here The structure stated can further benefit from the three-dimensional of polarization（X, y and z）Diversity.

In addition to space and polarity diversity, one embodiment of the present invention is using almost orthogonal antenna pattern （pattern）, to improve link performance via directional diagram diversity.Directional diagram diversity can improve the capacity and error code of mimo system Rate performance, and it can be found in following article compared to the advantage of other diversity antenna technologies：

[17] L.Dong, H.Ling, and R.W.Heath Jr., " Multiple-input multiple-output

Wireless communication systems using antenna pattern diversity, " Proc.IEEE Glob.Telecom.Conf., vol.1, pp.997-1001, Nov.2002.

[18] R.Vaughan, " Switched parasitic elements for antenna diversity, " IEEE

Trans.Antennas Propagat., vol.47, pp.399-405, Feb.1999.

[19] P.Mattheijssen, M.H.A.J.Herben, G.Dolmans, and L.Leyten, " Antenna- Pattern diversity versus space diversity fof use at handhelds, " IEEETrans.on Veh.Technol., vol.53, pp.1035-1042, July 2004.

[20] C.B.Dietrich Jr, K.Dietze, J.R.Nealy, and W.L.Stutzman, " Spatial, Polarization, and pattern diversity for wireless handheld terminals, " Proc.IEEE Antennas and Prop.Symp., vol.49, pp.1271-1281, Sep.2001.

[21] A.Forenza and R.W.Heath, Jr, " Benefit of Pattern Diversity Via2- Element Array of Circular Patch Antennas in Indoor Clustered MIMOChannels ", IEEE Trans.on Communications, vol.54, no.5, pp.943-954, May2006.

By using directional diagram diversity, multiple base stations or user antenna are may be such that each other closely, although and such as This also will not spatially be associated.

Fig. 3 provides the additional detail of one embodiment of the base station 200 and customer set up 203-207 shown in Fig. 2.For The purpose for simplifying, the base station 300 is only displayed as three antennas 305 and three customer set up 306-308.However, it is necessary to note Meaning, embodiments of the invention described herein can use the antenna 305 of virtually limitless quantity（That is, only by available sky Between and noise limit）Realized with customer set up 306-308.

Fig. 3 is similar with MIMO structures in the prior art shown in Fig. 1, wherein, both have three in every one end of communication channel Antenna.Significant difference is, in the mimo system of prior art, three antennas 105 on the right side of Fig. 1 be from each other it is fixed away from From（For example, being integrated in single device）, the signal received from each antenna 105 is together in signal processing subsystem 107 Processed.By contrast, in figure 3, three antennas 309 on the right side of figure are each coupled to different customer set up 306- On 308, each described customer set up can be distributed in the range of base station 305 Anywhere.So, each customer set up The signal for receiving can be in its coding, modulation, signal processing subsystem 311 independently of other two signals for receiving Processed.Therefore, with multiple input（That is antenna 105）Multiple output（That is antenna 104）" MIMO " system compare, Fig. 3 Show multiple input（That is antenna 305）Distribution output（That is antenna 305）System, below refers to " MIDO " system.

Note, the application uses the term usage different from application before, it is used to conform better to academia and industry Example.In entitled " the SYSTEMANDMETHOD FOR ENHANCING NEAR that 20 days cited before April in 2004 submits to VERTICAL INCIDENCE SKYWAVE（“NVIS”）COMMUNICATION USING SPACE-TIME CODING's " is total to With the application NO.10/902,978 that copending application NO.10/817,731 and on July 30th, 2004 submit to（The application is this The continuation application of application）In, " input " and " output "（In the environment of SIMO, MISO, DIMO and MIDO）The meaning with should Term expressing the meaning in this application is opposite.In application before, " input " refers to input to reception antenna（For example, in Fig. 3 Antenna 309）Wireless signal, and " output " refers to transmitting antenna（For example, antenna 305）The wireless signal of output.In academia In wireless industry, usually using the antisense of " input " and " output ", wherein " input " refers to the wireless signal of input to channel （That is, the wireless signal for being sent from antenna 305）, and " output " refers to the wireless signal from channel output（That is, antenna 309 is received Wireless signal）.The application uses this term usage, and the usage is opposite with the usage in the application that this section is cited before.Cause This, below depicts the term usage equivalent form of value between several applications：

10/817,731 and 10/902,978 the application

SIMO = MISO

MISO = SIMO

DIMO = MIDO

MIDO = DIMO

MIDO structures shown in Fig. 3 realize real in SISO systems similar to MIMO for the transmitting antenna of given quantity Existing capacity boost.However, a difference of the specific MIDO embodiments shown in MIMO and Fig. 3 is, to realize multiple base stations day The capacity boost that line is provided, each MIDO customer set up 306-308 is only required to single reception antenna, and for MIMO, each Customer set up at least requires the reception antenna with the capacity multiple for wishing to realize as many.Assuming that generally have the limitation of an implementation, Its limitation how many antenna can be placed in customer set up（As explained in the introduction）, this just limits mimo system on typical case System is between 4 to 10 antennas（4 times to 10 times of capacity）.Because base station 300 generally from fixation and fills dynamic position Many customer set ups are served, is expanded to and is exceeded well over 10 antennas, and with appropriately distance separate antenna realizing space Diversity is very actual.As described, each antenna arrangement has transceiver 304 and part coding, modulation and Signal Processing Element 303 disposal ability.It is worth noting that, in this embodiment, no matter base station 300 extend how much, each customer set up 306- 308 will be only required to an antenna 309, therefore cost for single user customer set up 306-308 will be very low, and base station 300 cost can be shared in the user of large cardinal.

In Fig. 4 to Fig. 6, it is shown that how to complete the example from base station 300 to the MIDO of customer set up 306-308 transmission Son.

In one embodiment of the invention, before MIDO transmission starts, channel is characterized.For mimo system, often Individual antenna 405 is transmitted one by one to training signal.Fig. 4 only show first transmission of training signal, but right For three antennas 405, three separate transmission are had.Each training signal is by coding, modulation and signal processing subsystem 403 generations, are converted into analog signal, and sent by each RF transceiver 404 as RF signals by D/A converter. Available a variety of codings, modulation and signal processing technology include, and are not limited to those technology described above（Example Such as, Reed Solomon, Viterbi coding（Viterbi Coding）；QAM, DPSK, QPSK modulation etc.）.

Each customer set up 406-408 is received training signal and is believed the training by transceiver 410 by its antenna 409 Number it is converted into baseband signal.A/D converter（It is not shown）Be encoded in the signal, modulate and signal processing subsystem 411 at The place of reason converts thereof into data signal.Then signal characteristic logic unit 320 recognizes the feature of gained signal（For example, knowing Not above-mentioned phase and amplitude distortion）And this feature is stored in memory.This characteristic processing process is similar to existing skill The processing procedure of the mimo system of art a, significant difference is that each customer set up only calculates one antenna, rather than The n characteristic vector of antenna.For example, as it is known that the training signal of pattern is by the coding of customer set up 406, modulation and signal Reason subsystem 420 is initialized（It is received by the information for sending when producing, or by other initialization process）.The same day When line 405 sends the training signal in a known pattern, coding, modulation and signal processing subsystem 420 use correlation method Find most strong training signal reception pattern, it saves phase and amplitude excursion, then its by the pattern from receiving Signal in the middle of cut.Next, it finds the last the second reception pattern related to the training signal, phase and amplitude is inclined Shifting is saved, and then it cuts the second strong mode from the signal for receiving.The treatment is carried out always, until preserving The phase of certain fixed qty and amplitude excursion（For example, 8）Or detectable training signal pattern drops to given background Under noise.The vector of phase/amplitude skew turns into the element H of vector 413₁₁.At the same time, customer set up 407 and 408 Coding, modulation and signal processing subsystem perform same treatment, produce their vector element H₂₁And H₃₁。

The memory of channel characteristics storage can be nonvolatile memory, such as flash memory, or hard disk, and/or volatibility Memory, such as random access memory（For example, SDRAM, RDAM）.Additionally, different user's sets can simultaneously using not The memory of same type stores characteristic information（Such as PDA can be and use flash memory, and notebook computer can be and use hard disk）. On various customer set ups or base station, ultimata of the present invention is not limited to any certain types of storing mechanism.

As described above, according to the scheme for being used, because each customer set up 406-408 only has an antenna, each is only Only store 1 × 3 row 413-415 of H-matrix.Fig. 4 shows the stage after the transmission of the first training signal, here, 1 × 3 row 413- 415 first row stores three channel characteristic informations of the first of antenna for base station 405 antenna.Remaining two row is stored from it The channel characteristics that two ensuing two training signals of antenna for base station of remaininging are transmitted.Note, for the purpose of illustration, described three Individual training signal pattern is in separate time tranfer.If selected for three training signal patterns so as to orthogonal, then it With simultaneous transmission, therefore the training time can be reduced.

As shown in figure 5, after all three pilot transmission is completed, each customer set up 506-508 will have been stored 1 × 3 row 513-515 of matrix H send back base station 500.For purposes of simplicity, client's dress is only shown in Figure 5 Put 506 transmission its characteristic informations.With reference to appropriate Error Correction of Coding（Such as Reed Solomon, Viterbi coding（Viterbi Coding）And/or enhancing coding（TurboCodes））, it is possible to use suitable modulator approach（Such as DPSK, 64QAM, OFDM） Come ensure base station 500 be accurately received row 513-515 in data.

In Fig. 5, although all three antenna 505 shows reception signal, for receiving every 1 × 3 row 513-515's Transmission, the single antenna of base station 500 and single transceiver are enough.However, under certain condition, using many or all antennas 505 and transceiver 504 come receive each transmission（That is, prior art is used in coding, modulation and signal processing subsystem 503 The multiple output of single input（“SIMO”）Treatment technology）Can realize than single antenna 505 and the more preferable signal to noise ratio of transceiver 504 （SNR）.

Coding, modulation and signal processing subsystem 503 when base station 500 receive described 1 from each customer set up 507-508 When × 3 row 513-515, be stored in 1 × 3 row 513-515 in 3 × 3 H-matrix 516 by it.For customer set up, base Standing can carry out storage matrix 516 using many different memory technologies, and it includes, but are not limited to, non-volatile mass storage Device（Such as hard disk）And/or volatile memory（Such as SDRAM）.Fig. 5 shows that base station is received and stores from visitor The stage of 1 × 3 row 513 of family device 509.When 1 × 3 row 514 and 515 is transmitted from remaining customer set up, they can be with It is transmitted and is stored in H-matrix 516, until whole H-matrix 516 is stored.

With reference to Fig. 6, the embodiment from base station 600 to the MIDO of customer set up 606-608 transmission will now be described.Because every Individual customer set up 606-608 is independent device, so each device receives different data transfers.So, the reality of base station 600 Applying example includes being located at wan interface 601 and encodes, carries out liaison to them between modulation and signal processing subsystem 603 Router 602, it receives multiple data flows from wan interface 601（Form is bit stream）, correspond respectively to each customer set up 606-608 is sent the data flow by separate data flow u1-u3.For this purpose, the router 602 can using it is various The route technology known.

As shown in fig. 6, by three bit streams, u₁-u₃, it route into the coding, modulation and signal processing subsystem In 603, the error correction stream of statistical iteration is encoded that as（For example, using ReedSolomon, Viterbi or enhancing coding）, and With to the suitable modulator approach of channel（Such as DPSK, 64QAM or OFDM）They are modulated.Additionally, the embodiment bag that Fig. 6 shows Signal precoding logic unit 630 is included, based on signal characteristic matrix 616, the signal precoding logic unit 630 is used for from every The signal that individual antenna 605 sends carries out unique encodings.Especially, in this embodiment, precoding logic unit 630 is by Fig. 6 In three bit stream u₁-u₃It is multiplied to generate three new bit stream u' with the inverse matrix of H-matrix 616₁-u'₃, rather than general Each coding and modulated bit stream are routed to separate antenna（As done in Fig. 1）.Then, D/A converter（Do not show Go out）Described three precoding bits circulation is changed to analog signal, transceiver 604 and antenna 605 are sent out as RF signals See off.

Before explaining how customer set up 606-608 receives the not special stream, will describe what precoding module 630 was performed Operation.Similar to the example of MIMO in above figure 1, the coding and modulated letter of each bit stream in three original bit streams Number will be indicated as u_n.In the embodiment shown in fig. 6, each u_iComprising the number that three bit streams that router 602 is route come According to, each such bit stream will as three user's set 606-608 one of those.

However, not as the MIMO examples in Fig. 1, there, each x_iThere is each antenna 104 to send, in the present invention shown in Fig. 6 Embodiment in, receive each u in each customer set up antenna 609_i（Plus noise N any in channel）.It is such to realize As a result, the output of each in three antennas 605（We are denoted as v_i）It is u_iWith the H for characterizing each customer set up The function of matrix.In embodiment, the precoding logic unit 630 in coding, modulation and signal processing subsystem is by performing Following equalities calculate each v_i：

v₁=u₁H^-1 ₁₁+u₂H^-1 ₁₂+u₃H^-1 ₁₃

v₂=u₁H^-1 ₂₁+u₂H^-1 ₂₂+u₃H^-1 ₂₃

v₃=u₁H^-1 ₃₁+u₂H^-1 ₃₂+u₃H^-1 ₃₃

Therefore, unlike MIMO, there, channel calculates each x after translating the signals into receiver_i, and described herein Inventive embodiment solved each v before channel is translated the signals into transmitter_i.Each antenna 609 is received from other For the u of other antennas 609_n-1The u separated in bit stream_i.The signal that each transceiver 610 will be received respectively is converted into Baseband signal, here A/D converter（It is not shown）It is digitized, each coding, modulation and signal processing subsystem 611 To its x_iBit stream is demodulated and decodes, and its bit is streamed into the data-interface 612 that customer set up is used（For example, objective Application program on the device of family）.

Embodiments of the invention described herein can be realized using various different codings and modulator approach.For example, In OFDM realizations, its intermediate frequency spectrum is divided into multiple split-band, and technology described herein can be used to characterize each single point Frequency band.However, as described above, cardinal principle of the invention is not limited to any specific modulator approach.

If customer set up 606-608 is portable data processing device, such as PDA, notebook computer and/or radio If words, then because customer set up may move to another from a position, then channel characteristics can frequently change Become.So, in one embodiment of the invention, the channel characteristics matrix 616 of base station is constantly updated.In an implementation In example, base station 600 is periodically（Every 250 milliseconds）Send new training signal to each customer set up, each customer set up by its Channel eigenvectors constantly send back base station 600 to ensure that channel characteristics keep accurate (if for example, environment changes or client Device movement is so as to have influence on channel).In one embodiment, it is right in the actual data signal for be sent to each customer set up Training signal is interleaved.Typically, the handling capacity of the training signal is far below the handling capacity of the data-signal, therefore this The throughput total to system will have little to no effect.Correspondingly, in this embodiment, channel characteristics matrix 616 in base station actively Renewal can be continuously available when being communicated with each customer set up, so that when customer set up moves to next position from a position Put, or environment to change and keep accurate channel characteristics so as to having influence on when channel.

One embodiment of the present of invention shown in Fig. 7 improves uplink communication channel using MIMO technology（That is, from client Channels of the device 706-708 to base station 700）.In this embodiment, the uplink channel characteristics logic unit 741 in base station is continuous Channel to coming from each customer set up is analyzed and characterizes.Especially, each customer set up 706-708 sends training letter Number to base station 700, there channel characteristics logic unit 741 is analyzed to produce the channel characteristics matrix 741 of N × M, and N is visitor here The quantity of family device, M is the quantity of the antenna that base station is used.Embodiment shown in Fig. 7 uses three Hes of antenna 705 in base station Three customer set up 706-708, which results in 3 × 3 channel characteristics matrixes 741 for depositing in base station 700.Customer set up can be by MIMO uplink shown in Fig. 7 is used to send data back to base station 700 and channel eigenvectors are sent back into base station 700, such as Shown in Fig. 5.But with unlike the embodiment shown in Fig. 5, in Figure 5, the channel eigenvectors of each customer set up with point The time opened is transmitted, and the method shown in Fig. 7 allows that channel eigenvectors are transmitted back into base simultaneously from multiple customer set ups 700 are stood, so as to substantially reduce influence of the channel eigenvectors to Return Channel throughput.

As described above, the feature of each signal can include several factors, for example, it is included relative to inside receiver Reference signal, absolute reference signal, the phase of relative reference signal, characteristic noise or other factors and amplitude.For example, orthogonal In the signal that amplitude modulation(PAM) is modulated, the feature can be the phase and amplitude excursion vector of several multipath images of signal. Another example is, in the signal that OFDM is modulated, the feature can be several or all in OFDM frequency spectrums The phase of single component signal and amplitude excursion vector.The training signal can be by the coding of each customer set up and modulation subsystem System 711 is generated, D/A converter（It is not shown）The training signal is converted into analog signal, then the transmitter of each customer set up It is converted into RF signals by 709 from baseband signal.In one embodiment, in order to ensure the synchronization of training signal, customer set up Only training signal is transmitted when base station requests（For example, in circulation（round robin）In the case of）.Furthermore, it is possible to Training signal is interleaved in the actual data signal sent from each customer set up, or training signal can with it is described Actual data signal is transmitted together.Therefore, even if customer set up 706-708 is mobile, uplink channel characteristics logic unit 741 The training signal can also be continuously transmitted and analyze, so that it is guaranteed that channel characteristics matrix 741 keeps updating.

Total channel capacity that previously described embodiments of the present invention is supported can be defined as min（N, M）, here, M is visitor The quantity of family device, and N is the quantity of antenna for base station.That is, capacity is limited by the antenna amount of base station side or customer side It is fixed.In this way, one embodiment of the present of invention ensures to be no more than min within preset time using simultaneous techniques（N, M）Individual antenna Sending/receiving.

In the typical case, the quantity of the antenna 705 of base station 700 will be less than the quantity of customer set up 706-708.Fig. 8 An exemplary situation is shown, it allows 5 customer set up 804-808 to be led to the base station with three antennas 802 Letter.In this embodiment, it is determined that the quantity of total customer set up 804-808 and detecting necessary channel characteristic information （For example, above description）Afterwards, base station 800 selects first crowd of three client 810 being in communication with（Because min（N, M） =3, so being in this instance three clients）.After the specified time that communicated with first crowd of client 810, base station just selects another Three clients 811 that group communicates with.In order to evenly distribute communication channel, the selection of base station 800 is not comprised in first group Two customer set ups 807,808.Further, since extra antenna is available, base station 800 just selects to be included in first group Extra customer set up 806.In one embodiment, base station 800 is circulated in the client masses by this way such that it is able to effectively Distribute to the handling capacity of each client equal number in time.For example, in order to evenly distribute handling capacity, base station can be then Select three any combinations of customer set up in addition to customer set up 806（That is, because customer set up 806 is used in beginning Communicated with base station in two circulations）.

In one embodiment, in addition to the data communication of standard, base station can transmit training using aforementioned techniques Signal receives training signal and signal characteristic data to each customer set up and from each customer set up.

In one embodiment, some customer set ups or customer set up group can be assigned to the handling capacity of varying level, example Such as, customer set up can be distinguished order of priority, the customer set up such that it is able to ensure of a relatively high priority must be relatively low preferential There are more communication cycles at client's dress family of level（That is, more handling capacities）.Based on a number of variable, can be to client's " priority " is selected, and the variable includes, for example, the subscription fee to wireless bandwidth of user（For example, for being ready to be volume Outer handling capacity pays more）, and/or communicate to/from the data type of customer set up（For example, real-time Communication for Power, for example call voice And video, obtain the priority higher than non-realtime traffic, such as Email）.

In the present load required based on each customer set up, the embodiment of base station dynamically distributes handling capacity.If for example, The live video stream of customer set up 804, and other devices 805-808 is performing the non real-time function of such as Email, then base Stand 800 can distribute relatively large number of handling capacity to the client 804.It is to be noted, however, that cardinal principle of the invention is not It is limited to any specific throughput distribution technology.

As shown in figure 9, two customer set ups 907,908 can be closely so that the channel characteristics of the client are in reality It is the same on border.As a result, base station will receive and storage two actually equal channel characteristics of customer set up 907,908 to Amount, therefore this will can not produce, signal of spatial distribution unique for each client.Correspondingly, in one embodiment, base station Will ensure that mutual distance any two closely or more customer set up is assigned to different groups.For example, in fig .9, Base station 900 first with first group 910 of customer set up 904,905 and 908 communication, then with customer set up 905,906,907 Second group 911 communication, which ensure that customer set up 907 and 908 is in different groups.

Selectively, in one embodiment, base station 900 is communicated with customer set up 907 and 908 simultaneously, but is made Communication channel is multiplexed with known multi-channel Technology.For example, base station can use time division multiplex（“TDM”）, frequency Divide multiplexing（“FDM”）Or CDMA（“CDMA”）Technology separates single, space correlation between customer set up 907 and 908 Signal.

Although above-mentioned each customer set up is equipped with single antenna, can be come by using the customer set up with multiple antennas Cardinal principle of the invention is realized to improve handling capacity.For example, when with above-mentioned wireless system, with 2 antennas Client will realize 2 times of throughput hoisting, and 3 times throughput hoisting, etc. will be realized with 3 clients of antenna（I.e., it is assumed that Space and angular separation between antenna are enough）.When being circulated by the customer set up with multiple antennas, base station Same general rule can be applied.For example, it can regard each antenna as separate client, and that is given by throughput distribution Individual " client ", is that any other client is the same just as it（For example, it is ensured that each client is provided with enough or suitable communication weeks Phase）.

As described above, one embodiment of the present of invention uses above-mentioned MIDO and/or MIMO signal transmission technology intimate Vertical incidence sky wave（“NVIS”）It is middle to improve signal to noise ratio and handling capacity.With reference to Figure 10, in one embodiment of the invention, equipment The NVIS base stations 1001 for having the matrix of N number of antenna 1002 are used for and M customer set up 1004 is communicated.Described NVIS days The antenna 1004 of line 1002 and various user's sets with vertical direction about into the angle within 15 degree by signal uplink transmit with Obtain desired NVIS and be preferably minimized surface wave disturbing effect.In one embodiment, antenna 1002 and customer set up 1004 assigned frequency using above-mentioned various MIDO and MIMO technology in NVIS frequency spectrums（For example in carrier frequency or it is less than The frequency of 23MHz, but usually less than in the frequency of 10MHz）Multiple independent data flows 1006 are supported, so as to significantly improve The handling capacity of assigned frequency（That is, it is directly proportional with the quantity of the data flow with statistical iteration）.

The NVIS antennas for serving given base station can have far physical distance each other.Assuming that being less than The long range that the long wavelength of 10MHz and signal are propagated（300 miles of round distance）, hundreds of yards, even several miles of antenna Physical separation can provide benefit in diversity.In such a situa-tion, single aerial signal can be withdrawn into centre bit Put, it is processed with traditional wired or wireless communication system.Selectively, each antenna can have local device Its signal is processed, the data transfer is then gone back into center using traditional wired or wireless communication system.In the present invention One embodiment in, NVIS base stations 1001 have to internet 1010（Or other wide area networks）Wideband link 1015 so that It is supplied to that customer set up 1003 is long-range, high speed, wireless network access.

In one embodiment, base station and/or user can utilize polarization/direction figure diversity（patterndiversity） Technology, while diversity is provided with lifting handling capacity, to reduce array size and/or user distance.For example, being passed with HF In defeated DIMO systems, due to polarization/direction figure diversity, user can be located at same position and their signal will not be associated. Especially, by using directional diagram diversity, a user can be communicated via earthwave with base station, and other users can be via NVIS and communicated with base station.

Additional embodiments of the invention

Ⅰ、DIDO-OFDM precodings are carried out using I/Q imbalances

One embodiment of the present invention is using for OFDM（OFDM）Distributed input distribution Formula is exported（DIDO）Inphase quadrature in system（I/Q）The system and method that imbalance is compensated.In short, according to this implementation Mode, user equipment is estimated channel, and the information is fed back into base station；Base station calculates pre-coding matrix, to eliminate Interference between carrier wave caused by I/Q imbalances and user between；And parallel data stream is sent out via DIDO precodings Deliver to multiple user equipmenies；The user equipment is forced via zero（ZF）, least mean-square error（MMSE）Or maximum likelihood（ML）Connect Receipts machine is demodulated to data, suppressing residual interference.

As detailed below, some notable features of the implementation method of the invention include, but are not limited to：

Precoding for eliminate ofdm system in from mirror image adjust（mirror tone）Inter-carrier interference（ICI）（Cause Caused by I/Q is mismatched）；

Precoding is for the inter-user interference and ICI in elimination DIDO-OFDM systems（Caused by I/Q is mismatched）；

For via using block diagonalization（BD）DIDO-OFDM systems in ZF receivers eliminate ICI（Because of I/Q not Caused by matching）Technology；

For via the precoding in DIDO-OFDM systems（At emitter）With ZF or MMSE wave filters（In receiver Place）To eliminate inter-user interference and ICI（Caused by I/Q is mismatched）Technology；

For via the precoding in DIDO-OFDM systems（At emitter）With similar to maximum likelihood（ML）Detector Nonlinear detector（At receiver）To eliminate inter-user interference and ICI（Caused by I/Q is mismatched）Technology；

Done for the intercarrier from mirror image tune in eliminating ofdm system using the precoding based on channel condition information Disturb（ICI）（Caused by I/Q is mismatched）；

Using the precoding based on channel condition information for the carrier wave from mirror image tune in elimination DIDO-OFDM systems Between disturb（ICI）（Caused by I/Q is mismatched）；

In base station known DIDO precoders are mismatched using I/Q（I/Q mismatch aware DIDOprecoder）, and DIDO receivers known to I/Q are used at user terminal；

In base station known DIDO precoders are mismatched using I/Q（I/Q mismatch aware DIDOprecoder）, DIDO receivers known to I/Q are used at user terminal, and use I/Q known channel estimators；

Known DIDO precoders are mismatched using I/Q in base station, uses DIDO known to I/Q to receive at user terminal Machine, and use I/Q known channels estimator and DIDO feedback generators known to I/Q（The maker by channel condition information from User terminal is sent to website）；

Known DIDO precoders are mismatched using I/Q in base station, and uses DIDO configurators known to I/Q（This is matched somebody with somebody Put device and various functions performed using I/Q channel informations, including user's selection, adaptive coding and modulation, the mapping of space-time frequency or Precoder is selected）；

Use DIDO receivers known to I/Q（The receiver is via using block diagonalization（BD）The DIDO-OFDM of precoder ZF receivers in system eliminate ICI（Caused by I/Q is mismatched））；

Use DIDO receivers known to I/Q（The receiver is via the precoding in DIDO-OFDM systems（At emitter） With similar to maximum likelihood（ML）The nonlinear detector of detector（At receiver）To eliminate inter-user interference and ICI（Cause Caused by I/Q is mismatched））；And

Use DIDO receivers known to I/Q（The receiver disappears via ZF the or MMSE wave filters in DIDO-OFDM systems Except ICI（Caused by I/Q is mismatched））.

a、Background

Exemplary radio communication system sends and receives signal comprising inphase quadrature（I/Q）Component.In actual system, The inphase quadrature component may the distortion due to the defect in mixing and baseband operations.These distortions（distortion）Performance For I/Q phases, gain and delay are mismatched.Unbalance in phase is by the sine in modulator/demodulator（sine）And cosine （cosine）It is correct it is orthogonal caused by.Gain imbalance is caused by the different amplification between inphase quadrature component 's.Due to I the and Q tracks in analog circuit（rail）Between delay it is different, it is also possible to there is additional distortion, the distortion is referred to as It is uneven to postpone.

In OFDM（OFDM）In system, I/Q imbalances can cause the intercarrier for carrying out spontaneous emission tune uneven （ICI）.The influence is studied in some data, and in following information, it has been proposed that for defeated to single input list The I/Q gone out in SISO-OFDM systems mismatches the method for compensating：M.D.Benedetto and P.Mandarini, “Analysis of the effect of the I/Qbaseband filter mismatch in an OFDM modem,” Wireless personalcommunications,pp.175-186,2000；S.Schuchert and R.Hasholzner, " A novelI/Q imbalance compensation scheme for the reception of OFDM signals,” IEEETransaction on Consumer Electronics,Aug.2001；M.Valkama, M.Renfors and V. Koivunen,“Advanced methods for I/Q imbalance compensation incommunication receivers,”IEEE Trans.Sig.Proc,Oct.2001；R.Rao and B.Daneshrad, " Analysis of I/Q mismatch and a cancellation scheme for OFDMsystems,"IST Mobile Communication Summit,June 2004；A.Tarighat, R.Bagheri and A.H.Sayed, " Compensation schemes and performance analysis of IQimbalances in OFDM receivers,”Signal Processing, IEEE Transactions on are [referring also to Acoustics, Speech, and Signal Processing, IEEE Transactions on],vol.53,pp.3257-3268,Aug.2005。

Extension of the work to multiple-input and multiple-output MIMO-OFDM systems is shown in following information：R.Rao and B.Daneshrad,“I/Q mismatch cancellation for MIMO OFDM systems,”in Personal, Indoor and Mobile Radio Communications,2004；PIMRC 2004.15th IEEE International Symposium on,vol.4,2004,pp.2710-2714.For spatial reuse（SM）, refer to R.M.Rao,W.Zhu,S.Lang,C.Oberli,D.Browne,J.Bhatia,J.F.Frigon,J.Wang,P;Gupta, H.Lee, D.N.Liu, S.G.Wong, M.Fitz, B.Daneshrad, and O.Takeshita, " Multiantenna testbeds for research and education inwireless communications,”IEEE Communications Magazine,vol.42,no.12,pp.72-81,Dec.2004；S.Lang, M.R.Rao and B.Daneshrad,“Design anddevelopment of a 5.25GHz software defned wireless OFDM communicationplatform,”IEEE Communications Magazine,vol.42,no.6,pp.6-12,June 2004；For orthogonal space time packet（OSTBC）, refer to A.Tarighat and A.H.Sayed, " MIMOOFDM receivers for systems with IQ imbalances,"IEEE Trans.Sig.Proc,vol.53,pp.3583- 3596,Sep.2005。

Unfortunately, how introduction is not currently existed to distributed input distributed output（DIDO）In communication system The data that I/Q gains and unbalance in phase error are corrected.Invention described below implementation method provides a kind of solution The scheme of these problems.

DIDO systems include a base station with spaced antenna, and the base station is using the nothing for being same as traditional SIO systems Line resource（That is, identical time-slot duration and frequency band）, send parallel data stream（Through precoding）To multiple users, to increase Strong downlink throughput.Entitled " the System and that S.G.Perlman and T.Cotter was submitted in July 30 in 2004 The application of Method for DistributedInput-Distributed Output Wireless Communications " No.10/902,978（" earlier application "）The detailed description of DIDO systems is given, this application is transferred and gives receiving for the application People, and this application is allowed to be incorporated herein as reference.

There are various ways and realize DIDO precoders.A kind of scheme is the block diagonalization described in following information （BD）：Q.H.Spencer, A.L.Swindlehurst and M.Haardt, " Zero forcingmethods for downlink spatial multiplexing in multiuser MIMO channels,”IEEETrans.Sig.Proc,vol.52, pp.461-471,Feb.2004；K.K.Wong, R.D.Murch, and K.B.Letaief, " A joint channel diagonalization for multiuser MIMO antennasystems,”IEEE Trans.Wireless Comm., vol.2,pp.773-786,JuI 2003；L.U.Choi and R.D.Murch, " A transmit preprocessing technique for multiuser MIMOsystems using a decomposition approach,”IEEE Trans.Wireless Comm.,vol.3,pp.20-24,Jan 2004；Z.Shen, J.G.Andrews, R.W.Heath and B.L.Evans,“Lowcomplexity user selection algorithms for multiuser MIMO systems With blockdiagonalization, " IEEE Trans.Sig.Proc, Sep.2005 are published in by receiving；Z.Shen, R.Chen, J.G.Andrews, R.W.Heath and B.L.Evans, " Sum capacity of multiuserMIMO Broadcast channels with block diagonalization, " it is submitted to IEEE Trans.Wireless Comm.,Oct.2005；R.Chen, R.W.Heath, and J.G.Andrews, " Transmitselection diversity for unitary precoded multiuser spatial multiplexing systemswith linear Receivers, " received to IEEE Trans, on Signal Processing, 2005.The use gone out given in these materials BD precoders are contemplated in the method for I/Q compensation, and the precoder can be scaled up to any types of DIDO precodings.

In DIDO-OFDM systems, I/Q is mismatched can cause two kinds of influences：ICI and inter-user interference.With SISO-OFDM Similar in system, the former is because the interference adjusted from mirror image is caused.The latter is due to the fact that what is caused, i.e. I/Q Mismatch can destroy the orthogonal of DIDO precoders, so as to produce interference between users.Can by method described herein, The interference of this two class is eliminated at transmitter and receiver.There is described herein three kinds of sides for the I/Q compensation in DIDO-OFDM systems Method, and mismatched for and without I/Q, compare their performance.Based on using performed by DIDO-OFDM prototypes Emulation and actual measurement, illustrate result.

Present embodiment is the extension of earlier application.Especially, these implementation methods have with the following characteristics of earlier application Close：

System described in earlier application, wherein I/Q tracks can be influenceed by gain and unbalance in phase；

At emitter, the DIDO with I/Q compensation is calculated using the training signal used for channel estimation pre- Encoder；And

Signal characteristic data take into account the distortion caused by I/Q imbalances, and at emitter, according to this material The method for being proposed, DIDO precoders are calculated using the signal characteristic data.

b、Embodiments of the present invention

First, Mathematical Modeling of the invention and framework will be described.

Before this programme is shown, explain that core mathematics concept is highly useful.We are by assuming that I/Q gains and phase Position is uneven（Phase delay is not included in this description, but the phase delay will be automatic in the algorithm of DIDO-OFDM forms Treatment）To be explained to it.To explain basic thought, it is assumed that we are want two plural number s=sI+jsQ and h=hI+jho phases Multiply, and cause x=h*s.We represent inphase quadrature component using subscript.Call below equation：

x_I=s_Ih_I-s_Qh_Q

And

x_Q=s_Ih_Q+s_Qh_I

Its matrix form is rewritable to be：

By channel matrix（H）To mark normalization to convert.It is now assumed that s is transmitted symbol, and h is channel.Can lead to Cross the following non-normalized change of establishment and bring and I/Q gains and the presence of unbalance in phase are modeled：

The effect of the skill is to confirm to be written as：

It is now right（A）It is written over：

We carry out defined below：

And

The two matrixes have normalization structure, therefore may be expressed as plural form：

h_e=h₁₁+h₂₂+j(h₂₁-h₁₂)

And

h_c=h₁₁-h₂₂+j(h₂₁+h₁₂)

By using all these knowledge, we can be derived back with two channels effective equation（Equivalent channels he and Conjugate channel hc）Scalar form.Therefore,（5）In efficient transformation be changed into：

x=h_es+h_cs^*

First channel is referred to as equivalent channels by us, and second channel is referred to as conjugate channel.If there is no I/Q gains and phase Position is uneven, then the equivalent channels are our channels to be observed.

By using similar argument, the discrete time MIMON × M systems with I/Q gains and unbalance in phase Input output Relationship can be shown as（Their matrix corresponding form is set up by using the scalar equivalent form of value）：

Wherein, t is discrete time index, h_e,h_c∈C^M×N, s=[s₁,...,s_N], x=[x₁..., x_M] and L take out for channel Head（channel tap）Number.

In DIDO-OFDM systems, received signal in frequency domain is illustrated.If meeting below equation, from signal Re-called with system：

FFT_K{ s [t] }=S [k] then FFT_K{s^*[t] }=S^*[(- k)]=S^*[K-k] for k=0,1 ..., K-1

Using OFDM, for subcarrier k, the Input output Relationship of equal value of MIMO-OFDM systems is：

Wherein, k=0,1 ..., K-1 are OFDM subcarrier indexes, H_eAnd H_cOf equal value and conjugate channel matrix is represented respectively, it is fixed Justice is as follows：

And

（1）In the second base value be from mirror image adjust interference.Can be by building following stacked（stacked）Matrix system （Please carefully note conjugate）To process it：

WhereinWithRespectively send and receive symbol vector in a frequency domain.

By using the method, active matrix can be set up, be operated for DIDO.For example, using DIDO 2 × 2 input- Output relation（Assuming that each user has single reception antenna）, first user equipment is contemplated that below equation（In the absence of making an uproar During sound）：

And second user notes below equation：

Wherein,Matrix H is represented respectively_eAnd H_cM rows, and W ∈ C^4x4It is DIDO precoding squares Battle array.According to（2）With（3）, it may be noted that the symbol that user m is receivedBy two interference sources caused by I/Q imbalances （That is, the inter-carrier interference adjusted from mirror image（That is,）And inter-user interference（That is,Andp ≠m））Influence.（3）In DIDO pre-coding matrixes W be designed to eliminate the two distracters.

There are multiple different implementation methods in the DIDO precoders that can be used for herein, this depends on being applied at receiver Joint-detection.In one embodiment, can use according to composite channel（Rather than）The block for being calculated Diagonalization（BD）（Refer to for example, Q.H.Spencer, A.L.Swindlehurst, and M.Haardt, " Zeroforcing methods for downlink spatialmultiplexing in multiuser MIMO channels,”IEEE Trans.Sig.Proc,vol.52,pp.461-471,Feb.2004.K.K；Wong, R.D.Murch, and K.B.Letaief, " A jointchannel diagonalization for multiuser MIMO antenna systems,”IEEE Trans.Wireless Comm.,vol.2,pp.773-786,JuI 2003；L.U.Choi and R.D.Murch, " Atransmit preprocessing technique for multiuser MIMO systems using adecomposition approach,”IEEE Trans.Wireless Comm.,vol.3,pp.20-24,Jan2004；Z.Shen, J.G.Andrews, R.W.Heath, and B.L Evans, " Low complexityuser selection algorithms For multiuser MIMO systems with blockdiagonalization, " IEEE is published in by receiving Trans.Sig.Proc,Sep.2005；Z.Shen, R.Chen, J.G.Andrews, R.W.Heath, and B.L Evans, " Sum Capacity of multiuserMIMO broadcast channels with block diagonalization, " carried Hand over to IEEE Trans.Wireless Comm., Oct.2005）.Therefore, current DIDO Systematic selections precoder, to cause：

Wherein, α_{I, j}It is constant, andThe method is highly profitable, because by using the precoding Device, because the influence that I/Q gains and unbalance in phase are completely eliminated at emitter can make its other party of DIDO precoders Keep intact in face.

DIDO precoders can be also designed as eliminating inter-user interference in advance, without eliminating in advance because IQ imbalances are led The ICI of cause.Using the method, receiver（Rather than emitter）Can by using one of receiving filter as described below come to IQ Imbalance is compensated.Therefore,（4）In Precoding Design standard can be modified to：

And

Wherein for m-th sending signal,AndThe symbolic vector received by user m.

In receiving side, in order to sending symbolic vectorEstimated, user m uses ZF wave filters, and estimated Symbolic vector is given as：

Although ZF wave filters are easiest to understand that receiver can also be applied any number of known to one of skill in the art Other wave filters.A kind of popular selection is MMSE wave filters, wherein：

And ρ is signal to noise ratio.Alternatively, the executable maximum likelihood symbol detection of user（Or the change of Sphere decoder device, iteration）. For example, first user can be used ML receivers, and solve following optimization：

Wherein, S is the set of all possible vectorial s, and depending on constellation sizes.The ML receivers are given can be preferable Performance, but complexity higher is required at receiver.One group of similar equation can be applied to second user.

Note,（6）With（7）InWithIt is assumed to be with zero.The hypothesis is only in transmitting precoder Can be completely eliminated and be directed to（4）In standard inter-user interference in the case of effectively.Similar,WithOnly in hair Penetrating precoder can be completely eliminated inter-carrier interference（That is, adjusted from mirror image）In the case of be diagonal matrix.

Figure 13 shows a kind of implementation method of the framework of the DIDO-OFDM systems with I/Q compensation, the DIDO- Ofdm system includes being located at base station（BS）Interior IQ-DIDO precoders 1302, send channel 1304, in user equipment Channel estimation logic 1306 and ZF, MMSE or ML receiver 1308.The channel estimation logic 1306 is via training signal To channelWithEstimated, and these estimations are fed back to the precoder in AP.BS calculates DIDO precoders Weight（Matrix W）, disturbed with eliminating interference caused by I/Q gains and unbalance in phase and user in advance, and by data Sent to user by wireless channel 1304.User equipment m uses ZF, MMSE or ML receiver 1308, by using unit 1304 channel estimations for being provided eliminate residual interference, and data are demodulated.

This I/Q backoff algorithm can be realized using three below implementation method.

Method 1-TX is compensated：In this embodiment, emitter according to（4）In standard calculate pre-coding matrix. At receiver, user equipment uses " simplification " ZF receivers, whereinWithIt is assumed to be diagonal matrix.Therefore, Formula（8）It is reduced to：

Method 2-RX is compensated：In this embodiment, emitter is based on R.Chen, R.W.Heath, andJ.G.Andrews,"Transmit selection diversity for unitary precoded multiuserspatial multiplexing systems with linear receivers,"accepted to IEEE Trans, onSignal Processing, the traditional BD methods described in 2005 calculate pre-coding matrix, and be not directed to（4）In Standard eliminate intercarrier and inter-user interference.Using the method,（2）With（3）In pre-coding matrix be reduced to：

At receiver, user equipment is such as（8）In like that use ZF wave filters.Note, the method is not such as above-mentioned method 1 Like that, interference is eliminated in advance at emitter.Therefore, it eliminates inter-carrier interference at receiver, but can not eliminate user Between disturb.Additionally, being fed back compared to the requirement of method 1WithIn method 2, user only needs feedback for emitter VectorTo calculate DIDO precoders.Therefore, method 2 is particularly suitable for the DIDO systems with low-rate feedback channel. On the other hand, method 2 need at user equipment have computation complexity somewhat higher, with（8）（Rather than（11））Middle calculating ZF receivers.

Method 3-TX-RX is compensated：In one embodiment, above-mentioned two method is merged.Emitter is such as（4）Count like that Calculate pre-coding matrix, and receiver according to（8）To estimate sending symbol.

I/Q is uneven（Either unbalance in phase, gain are uneven, or are to postpone imbalance）Can be to radio communication system Signal quality in system causes harmful degradation.For that reason, conventional circuit is both designed to relatively low imbalance pin. However, as described above, can correct this and ask by using the Digital Signal Processing of transmitting precoded form and/or specific receiver Topic.One embodiment of the present invention includes the system with multiple New function units, and each unit is for realizing ofdm communication I/Q corrections in system or DIDO-OFDM communication systems are critically important.

One embodiment of the present invention come from ofdm system using the precoding based on channel condition information with being eliminated The inter-carrier interference that mirror image is adjusted（ICI）（Cause because I/Q is mismatched）.As shown in figure 11, the DIDO transmittings according to present embodiment Machine includes subscriber selector unit 1102, multiple coded modulation units 1104, corresponding multiple map units 1106, DIDO IQ Known precoding unit 1108, multiple RF transmitter units 1114, user feedback unit 1112 and DIDO configurator units 1110。

The subscriber selector unit 1102 is based on the feedback information acquired in feedback unit 1112, selection and multiple users U₁-U_MAssociated data, and provide this information to each coded modulation unit in multiple coded modulation units 1104 1104.Each coded modulation unit 1104 is encoded and demodulated to the information bit of each user, and sends them to reflect Penetrate unit 1106.Input bit is mapped to complex symbol by the map unit 1106, and result is sent to pre- known to DIDO IQ Coding unit 1108.The channel conditions that the precoding unit 1108 known to DIDO IQ is obtained using feedback unit 1112 from user Information, calculates precoding weight known to DIDO IQ, and incoming symbol to being obtained from map unit 1106 carries out precoding.Often One pre-code data stream is sent to OFDM units 1115 by precoding unit 1108 known to DIDO IQ, the OFDM units 1115 calculate IFFT, and add Cyclic Prefix.The information is sent to D/A units 1116, and the D/A units 1116 carry out digital-to-analogue and turn Change, and send it to RF units 1114.The RF units 1114 send it to baseband signal raising frequency to intermediate frequency/radio frequency Transmitting antenna.

The precoder is operated together to the conventional mirror image tune that reconciles, uneven to compensate I/Q.Arbitrary Digit can be used The precoder design standard of amount, including ZF, MMSE or weighting MMSE designs.In a preferred embodiment, precoder can be complete The full ICI removed caused by I/Q is mismatched, so that receiver need not perform any ancillary relief.

In one embodiment, the precoder uses block diagonalization standard, not exclusively to eliminate each user I/Q influence（This needs accessory receiver to process）In the case of, inter-user interference is completely eliminated.In another embodiment, Inter-user interference and ICI interference caused by I/Q imbalances are completely eliminated using zero pressure standard for the precoder. The implementation method can use traditional DIDO-OFDM processors at receiver.

One embodiment of the present invention uses the precoding based on channel condition information, to eliminate DIDO-OFDM systems In from mirror image adjust inter-carrier interference（ICI）（Caused by I/Q is mismatched）, and each user uses DIDO known to IQ to receive Machine.As shown in figure 12, in one embodiment of the invention, system（Including receiver 1202）Including multiple RF units 1208th, correspondingly multiple A/D units 1210, IQ known channels estimator 1204 and DIDO feedback generator units 1206。

The RF units 1208 receive the signal sent from DIDO transmitter units 1114, by the signal down to base band, And the signal after the frequency reducing is supplied to A/D units 1210.Afterwards, 1210 pairs of signals of the A/D units carry out analog-to-digital conversion, and Send it to OFDM units 1213.The OFDM units 1213 remove Cyclic Prefix, and carry out FFT, by the signal reports extremely Frequency domain.During cycle of training, OFDM units 1213 transmit output to IQ known channels estimation unit 1204, the known letters of the IQ Road estimation unit 1204 calculates channel estimation in a frequency domain.Alternatively, the channel estimation can in the time domain be calculated.In data week Phase（data period）Period, OFDM units 1213 transmit output to receiver unit 1202 known to IQ.The IQ is known to be received Machine unit calculates IQ receivers, and the signal is demodulated/decoded, to obtain data 1214.The IQ known channels are estimated Meter unit 1204 sends the channel estimation to DIDO feedback generators unit 1206, and the feedback generator unit 1204 can be to institute State channel estimation to be quantified, and emitter is sent it back to via feedback control channel 1112.

Receiver 1202 shown in Figure 12 can be in any number of standard known in those skilled in the art（Including ZF, MMSE, maximum likelihood or MAP receivers）Lower work.In a preferred embodiment, receiver is eliminated using MMSE wave filters ICI caused by the IQ imbalances on mirror image tune.In another preferred embodiment, receiver is used similar to maximum likelihood The nonlinear detector of search carrys out the symbol that joint-detection mirror image is adjusted.The method has good performance, but with higher Complexity.

In one embodiment, receiver coefficient is determined using IQ known channels estimator 1204, to remove ICI. Therefore, we require that DIDO-OFDM systems（The load adjusted from mirror image is eliminated using the precoding based on channel condition information Disturbed between ripple（ICI）（Caused by I/Q is mismatched））, DIDO receivers known to IQ and IQ known channel estimators rights and interests. The channel estimator can be used traditional training signal, or the instruction for being usable in the special structure of transmission on inphase quadrature signal Practice signal.Any number of algorithm for estimating, including least square method, MMSE or maximum likelihood can be implemented.The IQ known channels Estimator receiver known to IQ provides input.

Channel condition information can be provided to website by channel reciprocity or by feedback channel.Of the invention one is real Applying mode includes DIDO-OFDM systems, and the system has precoder known to I/Q, and for the letter of user terminal in future Feedback channel known to the I/Q that road condition information transmits to website.The feedback channel can be physically or logically control channel.It can be By special or shared in RACH.Can be by using user terminal（We have also required that the rights and interests of the user terminal）Place DIDO feedback generators generate feedback information.The DIDO feedback generators are defeated by the I/Q known channels estimator Go out as input.It can quantized channel coefficient, or usable any amount Limited Feedback algorithm known in the field.

The distribution of user, modulation and encoding rate, the mapping to space-time frequency coding time slot can feed back according to the DIDO and generate The result of device and change.Therefore, an implementation method includes DIDO configurators known to IQ, and the configurator is used from one or more The IQ known channels of user are estimated to configure precoder known to DIDO IQ, select modulation rate, encoding rate, allow the use of transmission The subset at family and their mapping to space-time frequency coding time slot.

In order to evaluate the performance of proposed compensation method, three systems of DIDO 2 × 2 will be compared：

1st, mismatched with I/Q：By all of tune（Except DC mediations edge is adjusted）It is transmitted, and not to I/Q not With compensating；

2nd, compensated with I/Q：It is transmitted by all of tune, and comes to mismatch I/Q by using above-mentioned " method 1 " Compensate；

3rd, preferably：Only by odd number adjust be transmitted, with avoid inter-user interference and because I/Q mismatch caused by Intercarrier（That is, adjusted from mirror image）Interference.

After this, illustrate in true propagation situation and measure acquired result using DIDO-OFDM prototypes.Figure 14 depict the 64-QAM planispheres acquired in above three system.These planispheres are in same customer location and fixation Average signal-to-noise ratio（~45dB）In the case of obtain.First planisphere 1401 is very noisy（By I/Q imbalances are led Cause the interference adjusted from mirror image）.Second planisphere 1402 shows that some are improved（Due to I/Q compensation）.Note, the second planisphere 1402 ideal situation not shown in planisphere 1403 is pure like that（Due to existing there may be inter-carrier interference（ICI）'s Phase noise）.

Figure 15 shown in the case of unmatched with and without I/Q, the DIDO 2 of 64-QAM and 3/4 encoding rate × The average SER of 2 systems（SER）1501 and per user goodput（goodput）1502.OFDM is with a width of 250KHZ, adjusts and circulating prefix-length L with 64_cp=4.Because in the ideal case, we are only sent out by the subset adjusted Data are sent, therefore according to the transmission power averagely often adjusted（Rather than total transmission power）To evaluate SER and goodput performance, To ensure the fair comparing between different situations.Additionally, in following result, we use the normalized value of transmission power（With Decibel sign）, because the target of ours is compare different schemes relative herein（Rather than definitely）Performance.Figure 15 shows and is depositing In the case of I/Q is unbalanced, SER saturations and miss the mark SER（~10^-2）, this and A.Tarighat and A.H.Sayed,"MIMO OFDM receiversfor systems with IQ imbalances,"IEEE The result reported in Trans.Sig.Proc, vol.53, pp.3583-3596, Sep.2005 is consistent.The saturation effect be by On the fact that caused, i.e. signal power and jamming power（Adjusted from mirror image）Increase with the increase of TX power.So And, pass through proposed I/Q compensation methodes, interference can be eliminated, and obtain preferably SER performances.Note, due to 64-QAM modulation Larger transmission power is needed, therefore, SER can be caused to have at SNR high because of the amplitude saturation effect in DAC trickle Increase.

Additionally, can be observed, existing in the case that I/Q compensates, SER performances closely ideal situation.This two kinds of feelings Between condition, the 2dB gaps of TX power are due to phase noise（The phase noise may produce additional between adjacent OFDM tune Interference）Cause.Finally, goodput curve 1502 shows that it can send out compared to ideal situation when application I/Q methods Send the data of twice because we used all of data adjust rather than only odd number adjust（For ideal situation）.

Figure 16 is illustrated in the case where being compensated with I/Q or being compensated without I/Q, the SER of different Q AM planispheres Energy.We can be observed, and in this embodiment, the method for being proposed is particularly advantageous for 64-QAM planispheres. For 4-QAM and 16-QAM, I/Q compensation methodes meeting producing ratio has the I/Q worse performances of unmatched situation, and this may Because the method bigger power of requirement for being proposed is eliminated the interference that carries out data is activation and adjust from mirror image.Additionally, The unmatched shadows of I/Q are subject to due to the larger minimum range between constellation point, 4-QAM and 16-QAM and not as 64-QAM Ring.Referring to A.Tarighat, R.Bagheri, and A.H.Sayed, " Compensation schemes and performance analysis of IQ imbalances in OFDMreceivers,"Signal Processing,IEEE Transactions on are [referring also to Acoustics, Speech, and Signal Processing, IEEE Transactions on],vol.53,pp.3257-3268,Aug.2005.Also observable Figure 16 and by by I/Q mismatch with It is compared for the ideal situation of 4-QAM and 16-QAM and draws the conclusion.When therefore, for 4-QAM and 16-QAM For, eliminated with interference（Adjusted from mirror image）DIDO precoders required for secondary power can not be compensated for I/Q Slight interests go bail for.Note, the problem can be solved by using above-mentioned I/Q compensation methodes 2 and 3.

Finally, under different propagation conditions, the relative SER performances of above three method are measured.Also describe and depositing It is for reference in the SER performances of the unmatched situations of I/Q.It is 450.5MHZ and with a width of that Figure 17 is depicted for carrier frequency The systems of 64-QAM DIDO 2 × 2 of 250KHz, in the SER measured by two different customer locations.In position 1, user and place In not chummery and in NLOS（Without sighting distance）The BS of state is at a distance of ~ 6 λ.In position 2, user and with LOS（Sighting distance）BS phases Away from ~ λ.

Figure 17 shows a case that all three compensation method is had outstanding performance than not compensating.It should be noted, however, that , under any channel conditions, method 3 surpasses other two kinds of compensation methodes.The relative performance of method 1 and 2 depends on passing Broadcast situation.By actual measurement activity, can show that method 1 generally surpasses method 2, because it is eliminated in advance（In emitter Place）Inter-user interference caused by I/O imbalances.When the inter-user interference very little, as shown in the curve map 1702 of Figure 17, side Method 2 can surpass method 1, because it is not exposed to the power attenuation caused by I/Q compensation precoders.

Up to the present, by only considering limited group of propagation situation（As shown in figure 17）And distinct methods are compared Compared with.After this, in preferable i.i.d.（Independent and tool is with distribution）The relative performance of these methods is measured in channel.Using hair I/Q phases and gain imbalance with receiving side is penetrated to emulate DIDO-OFDM systems.Figure 18 is shown in only transmitter side to be had In the case of gain balance（That is, there is gain 0.8 on the I rails of the first transmitting chain, there is gain 1 on other rails）, institute The performance of the method for proposition.Can be seen that, method 3 has surpassed every other method.Additionally, with position in the curve map 1702 of Figure 17 Obtain result at 2 to compare, in i.i.d. channels, the comparable method 2 of method 1 is better carried out.

Therefore, it is uneven to compensate the I/Q in above-mentioned DIDO-OFDM systems to give three kinds of novel methods, and method 3 surpasses Other compensation methodes for being proposed.In the system with low-rate feedback channel, method 2 can be used to reduce DIDO precodings Required feedback quantity, but poor SER performances can be caused.

Ⅱ、Self adaptation DIDO delivery plans

By description for strengthening distributed input distributed output（DIDO）The system and method for the performance of system it is another Implementation method.Radio resource allocation is dynamically given different user equipmenies by the method by tracking the channel status for changing, with Increase handling capacity while some target error rates are met.The user equipment estimated channel quality, and its is anti- It is fed to base station（BS）；The base station is processed the channel quality for being obtained from user equipment, to select to be used for what is sent next time Optimal user cluster tool, DIDO schemes, modulation/coding scheme（MCS）And array configuration；The base station is via precoding Parallel data is sent to multiple user equipmenies, and signal and is demodulated at receiver.

One system for effectively distributing resource for DIDO Radio Links is also described.The system is included with DIDO configurators DIDO base stations, the base station is processed the feedback for receiving personal family, with select for send next time optimal user set, DIDO schemes, modulation/coding scheme（MCS）And array configuration；Receiver in DIDO systems, the receiver to channel and its He measures relevant parameter, to generate DIDO feedback signals；And DIDO feedback control channels, for by from the anti-of user Feedforward information is transferred to base station.

As detailed in the following, some notable features of the implementation method of the invention may include, but be not limited to：

For based on channel quality information, being adaptive selected number of users, DIDO delivery plans（That is, day line options or Multiplexing）, modulation/coding scheme（MCS）And array configuration, to minimize SER, or maximize spectrum efficiency per user or under The technology of line link spectrum efficiency；

Technology for defining combination of multigroup DIDO sending modes as DIDO schemes and MCS；

For giving different time slots, the OFDM skill for reconciling DIDO subflows by different DIDO mode assignments according to channel status Art；

Different DIDO patterns are dynamically assigned to the technology of different user for the channel quality based on different user；

For being swashed to self adaptation DIDO switchings based on the link quality metric calculated in time domain, frequency domain and spatial domain Standard living；

Standard for being switched into line activating to self adaptation DIDO based on look-up table.

The DIDO systems in base station with DIDO configurators as shown in figure 19, the system can be believed based on channel quality Breath, is adaptive selected number of users, DIDO delivery plans（That is, day line options or multiplexing）, modulation/coding scheme（MCS）With And array configuration, to minimize SER, or maximize the spectrum efficiency or downlink tone spectrum efficiency per user；

As shown in figure 20 having DIDO configurators and at each user equipment there is DIDO to feed back in base station is generated The DIDO systems of device, the system uses the other specification at estimated channel conditions and/or receiver（Similar to estimated SNR）, to generate the feedback message of input to DIDO configurators.

DIDO systems, the system has DIDO configurators（In base station）, DIDO feedback generators and DIDO feedback control Channel processed（The DIDO feedback channels are used to from user transmit DIDO specific configuration informations to base station）.

a、Background

In multiple-input and multiple-output（MIMO）In system, it is contemplated that diversity scheme（For example, orthogonal space time packet（OSTBC） （Referring to V.Tarokh, H.Jafarkhani, and A.R.Calderbank, " Spacetime block codes from orthogonal designs,”IEEE Trans.Info.Th.,vol.45,pp.1456-467,JuI.1999）Or day line selection Select（Referring to R.W.Heath Jr., S.Sandhu, andA.J.Paulraj, " Antenna selection for spatial multiplexing systems with linearreceivers,”IEEE Trans.Comm.,vol.5,pp.142-144, Apr.2001）, to prevent fading channel, improve link reliability（The reliability can be exchanged into more preferably coverage rate）.The opposing party Face, spatial reuse（SM）Can be sent using multiple parallel datas and carry out strengthening system handling capacity as means.Referring to G.J.Foschini, G.D.Golden, R.A.Valenzuela, and P.W.Wolniansky, " Simplifedprocessing for high spectral effciency wireless communication employingmultielement arrays,”IEEE Jour.Select.Areas in Comm.,vol.17,no.11,pp.1841-1852,Nov.1999.Root According to from L.Zheng and D.N.C.Tse, " Diversity andmultiplexing:a fundamental tradeoff in multiple antenna channels,”IEEE Trans.Info.Th.,vol.49,no.5,pp.1073-1096, The theoretical diversity of May 2003/multiplexing compromise, these benefits can be realized simultaneously in mimo systems.One actual form of implementation is logical The channel status of tracking change is crossed, self adaptation switching is carried out between diversity and multiplexing delivery plan.

It has been proposed that a large amount of adaptive MIMO transmission technologies.R.W.Heath and A.J.Paulraj, " Switching between diversity and multiplexing in MIMO systems,”IEEETrans.Comm.,vol.53, Diversity/multiplexing changing method in no.6, pp.962-968, Jun.2005 is designed to, based on momentary channel quality information, change Enter the BER sent for fixed rate（Bit error rate）.Alternatively, can such as S.Catreux, V.Erceg, D.Gesbert, and R.W.Heath.Jr.,“Adaptive modulation and MIMO coding for broadband wireless Datanetworks, " in IEEE Comm.Mag., vol.2, pp.108-115, June 2002 (" Catreux ") like that, use Statistic channel information enters line activating to self adaptation, so as to reduce the quantity of feedback overhead and control message.In Catreux When self adaptation transmission algorithm is designed to based on channel/frequency selection designator, for OFDM（OFDM）In system The predeterminated target bit error rate, strengthens spectrum efficiency.Also directed to narrowband systems, it is proposed that similar low feedback adaptive method, the party Method is selectively come to be switched between diversity scheme and spatial reuse using channel space.See, for example, A.Forenza, M.R.McKay, A.Pandharipande, R.W.Heath.Jr., and I.B.Collings, " Adaptive MIMOtransmission for exploiting the capacity of spatially correlated channels,”accepted to the IEEE Trans,on Veh.Tech.,M ar.2007；M.R.McKay, I.B.Collings,A.Forenza,and R.W.Heath.Jr.,“Multiplexing/beamformingswitching For coded MIMO in spatially correlated Rayleigh channels, " received to IEEE Trans, on Veh.Tech.,Dec.2007；A.Forenza, M.R.McKay, R.W.Heath.Jr., and I.B.Collings, “Switching between OSTBC and spatial multiplexing withlinear receivers in spatially correlated MIMO channels,”Proc.IEEE Veh.Technol.Conf.,vol.3, pp.1387-1391,May 2006；M.R.McKay, I.B.Collings, A.Forenza, and R.W.Heath Jr., " A throughput-based adaptive MIMO BICMapproach for spatially correlated Channels, " appear in Proc.IEEE ICC, June 2006.

In the data, the working range represented in various previous disclosures is extended to DIDO-OFDM systems by us.Ginseng See such as R.W.Heath and A.J.Paulraj, " Switching betWeendiversity and multiplexing in MIMO systems,”IEEE Trans.Comm.,vol.53,no.6,pp.962-968,Jun.2005；S.Catreux, V.Erceg, D.Gesbert, and R.W.Heath Jr., " Adaptive modulation and MIMO coding for broadband wireless datanetworks,”IEEE Comm.Mag.,vol.2,pp.108-115,June 2002； A.Forenza, M.R.McKay, A.Pandharipande, R.W. Heath Jr., and I.B.Collings, “AdaptiveMIMO transmission for exploiting the capacity of spatially correlated channels,”IEEE Trans,on Veh.Tech.,vol.56,n.2,pp.619-630,Mar.2007； M.R.McKay, I.B.Collings, A.Forenza, and R.W.Heath Jr., " Multiplexing/ beamformingswitching for coded MIMO in spatially correlated Rayleigh Channels, " received to IEEE Trans, on Veh.Tech., Dec.2007；A.Forenza,M.R.McKay, And I.B.Collings, " Switching between OSTBC and spatial multiplexing R.W.HeathJr., withlinear receivers in spatially correlated MIMO channels,”Proc.IEEE Veh.Technol.Conf.,vol.3,pp.1387-1391,May 2006；M.R.McKay,I.B.Collings, A.Forenza, and R.W.Heath Jr., " A throughput-based adaptive MIMO BICMapproach for Spatially correlated channels, " appear in Proc.IEEE ICC, June 2006.

There is described herein NEW ADAPTIVE DIDO sending strategys, the strategy with based on channel quality information in varying number Switch over as a kind of means to improve systematic function between user, the transmitting antenna of varying number and delivery plan.Note Meaning, M.Sharif and B.Hassibi, " On the capacity ofMIMO broadcast channel with partial Side information, " IEEE Trans.Info.Th., vol.51, p.506522, Feb.2005 and W.Choi, A.Forenza, J.G.Andrews, and R.W.Heath Jr., " Opportunistic space division multiple Access with beam selection, " IEEE Trans are appeared in, had been presented for many on Communications The scheme of adaptively selected user in user's mimo system.However, these opportunistics in disclosing（opportunistic）It is empty Multiplexing is divided to access（OSDMA）Scheme is designed to maximize total capacity by using multi-user diversity, and they only can Realize dirty paper（dirty paper）The part of theory capacity of code, because eliminating interference in advance completely not at emitter. In DIDO transmission algorithms described in this, inter-user interference is eliminated using block diagonalization in advance.However, proposed self adaptation sends Strategy can apply to any DIDO systems, the type without considering precoding technique.

This patent application describes the extension of the invention described above and the implementation method of earlier application, including but not limited to Lower supplementary features：

1st, can by wireless client device using in earlier application be used for channel estimation training symbol come to self adaptation DIDO Link quality metric in scheme is evaluated.

2nd, as described in earlier application, base station receives the signal characteristic data from client device.Current real Apply in mode, signal characteristic data are defined for entering self adaptation the link quality metric of line activating.

3rd, one is prior application described for selecting the mechanism of antenna and number of users, and defines throughput distribution.This Outward, the handling capacity of different stage can be dynamically assigned to different clients as earlier application.Current implementation of the invention Mode defines the novel standard related to the selection and throughput distribution.

b、Embodiments of the present invention

The target of the self adaptation DIDO technologies for being proposed is by by the Radio Resource dynamic in time, frequency and space The different user in system is distributed to strengthen the spectrum efficiency or downlink tone spectrum efficiency of every user.The overall self adaptation mark It is mutatis mutandis in while target error rate is met, improve handling capacity.According to spread state, it is also possible to use the adaptive algorithm via Diversity scheme improves the link-quality of user（Or coverage rate）.The flow chart that Figure 21 shows describes self adaptation DIDO schemes Step.

In 2102, base station（BS）Collect the channel condition information from all users.2104, according to the CSI for being received, Base station calculates link quality metric in time domain/frequency domain/spatial domain.2106, being selected using these link quality metrics will be under The user of being serviced in one transmission, and for the sending mode of each user.Note, sending mode include modulation/coding with And the various combination of DIDO schemes.Finally, user is sent data to via DIDO precodings in 2108, BS.

In 2102, the base station selected channel condition information from all user equipmenies（CSI）.2104, base station uses should CSI determines the instantaneous or statistical channel quality of all user equipmenies.In DIDO-OFDM systems, can be in time domain, frequency domain and sky Domain is to channel quality（Or link quality metric）Estimated.Afterwards, 2106, base station is determined most using link quality metric Good user's subset and the sending mode for current propagation state.DIDO sending mode set is combined into DIDO schemes（That is, Its line options or multiplexing）, modulation/coding scheme（MCS）And the combination of array configuration.2108, by using selected user Quantity and sending mode, send data to user equipment.

Can be by look-up table（LUT）（The look-up table is based on the bit error rate performance in DIDO systems difference communication environments It is pre-calculated）To carry out model selection.Channel quality information is mapped to bit error rate performance by these LUT.In order to build LUT, can evaluate DIDO systems bit error rate performance in different propagation situations according to SNR.Be can be seen that from ber curve, can calculated Minimum SNR needed for realizing a certain predeterminated target bit error rate.The SNR requirement definitions are SNR threshold values by we.Afterwards, in difference Propagation situation and evaluate SNR threshold values for different DIDO sending modes, and store it in LUT.For example, can make LUT is built with SER results in Figure 24 and Figure 26.Afterwards, according to the LUT, base station may be selected the transmission mould for active user Formula, the pattern can improve handling capacity while the predeterminated target bit error rate is met.Finally, base station via DIDO precodings by number According to transmission to selected user.Note, different DIDO mode assignments can be adjusted and DIDO subflows to different time slots, OFDM, with So that self adaptation can be carried out in time domain, frequency domain and spatial domain.

Figure 19-Figure 20 shows a kind of implementation method using the system of DIDO self adaptations.Introduce some new functions Unit implements proposed DIDO adaptive algorithms.Specifically, in one embodiment, DIDO configurators 1910 can base In the channel quality information 1912 that user equipment is provided, various functions, including selection number of users, DIDO delivery plans are performed （That is, day line options and multiplexing）, modulation/coding scheme（MCS）And array configuration.

Subscriber selector unit 1902 is based on the feedback information as acquired in DIDO configurators 1910, selection and multiple users U₁-U_MAssociated data, and the information is provided per each coded modulation unit in multiple coded modulation unit 1904.Often Individual coded modulation unit 1904 is encoded and modulated to the information bit of each user, and they are sent to map unit 1906.Input bit is mapped to complex symbol by the map unit 1906, and sends it to precoding unit 1908.Coding is adjusted , using the information for being obtained from DIDO configurators unit 1910, it is each user institute to select for unit processed 1904 and map unit 1906 The modulation/coding scheme type of use.Described information can be provided by configurator unit 1910 by using feedback unit 1912 The channel quality information of each user calculate.DIDO precoding units 1908 are utilized and obtained by DIDO configurators unit 1910 The information for taking calculates DIDO precoding weights, and incoming symbol to being obtained from map unit 1906 carries out precoding.By DIDO precoding units 1906 send to OFDM units 1915 data flow after each precoding, and the OFDM units 1915 are calculated IFFT simultaneously adds Cyclic Prefix.D/A units 1916 are sent this information to, the D/A units 1916 carry out digital-to-analogue conversion, and will most Whole analog signal is sent to RF units 1914.Be sent to for baseband signal raising frequency to intermediate frequency/radio frequency by the RF units 1914 To transmitting antenna.

The RF units 2008 of each client device receive the signal sent from DIDO transmitter units 1914, by the signal Base band is downconverted to, and the signal after frequency reducing is supplied to A/D units 2010.Afterwards, the A/D units 2010 are by the signal from mould Plan is converted to numeral, and sends it to OFDM units 2013.The OFDM units 2013 remove Cyclic Prefix, and perform FFT, with By signal reports to frequency domain.In cycle of training, OFDM units 2013 transmit output to channel estimating unit 2004, and the channel is estimated Meter unit 2004 calculates channel estimation in a frequency domain.Alternatively, can be in time-domain calculation channel estimation.During the data cycle, OFDM units 2013 transmit output to receiver unit 2002, and the receiver unit 2002 is demodulated to signal/decodes, with Obtain data 2014.The channel estimating unit 2004 sends to DIDO feedback generators unit 2006 channel estimation, should DIDO feedback generators unit 2006 can quantify to channel estimation, and hair is sent it back to via feedback control channel 1912 Penetrate machine.

The DIDO configurators 1910 are usable in the information that base station is obtained, or in a preferred embodiment, additionally make With the DIDO feedback generators 2006 worked at each user equipment（Referring to Figure 20）Output.The DIDO feedback generators 2006 are generated using the other specification similar to estimated SNR at estimated channel conditions 2004 and/or receiver The feedback message of DIDO configurators 1910 will be input to.The DIDO feedback generators 2006 can enter at receiver to information Row compresses, quantifies and/or uses some Limited Feedback strategies known in the field.

The DIDO configurators 1910 can be used the information recovered from DIDO feedback control channels 1912.DIDO feedback controls Channel is logic or physical control channel, and the channel can be used to from user send the output of DIDO feedback generators 2006 to base Stand.Control channel 1912 can be implemented in any number of mode known in the field, and can be that logic or physical control are believed Road.Used as physical channel, it may include to be assigned to the dedicated time slot/frequency band of user.It can also be by all users to share with Machine access channel.The control channel can be pre-assigned, or can occupy bit by predetermined way in existing control channel （stealing bits）To create.

In the following discussion, will be measured by using DIDO-OFDM prototypes described in true propagation environment and be obtained The result for taking.These results show that in self adaptation DIDO systems potential gain realizability.Represent different stage first The performance of DIDO systems, shows that antenna/user quantity can be increased, to realize bigger downlink throughput.Afterwards, description with The relevant DIDO performances in the position of user equipment, demonstrate the need for the channel status of tracking change.Finally, to using diversity technique The performance of DIDO systems is described.

The performance of i, different stage DIDO systems

Using increasing transmitting antenna（N=M, wherein M are number of users）To evaluate the performance of different DIDO systems. The performance of following system is compared：SISO, DIDO 2 × 2, DIDO 4 × 4, DIDO 6 × 6 and DIDO 8 × 8.DIDO N × M refers to has N number of transmitting antenna and the M DIDO of user at BS.

Figure 22 shows that transmit/receive antenna is laid out.Transmitting antenna 2201, and user position are arranged with square array configuration Around emission array.In Figure 22, T refers to " transmitting " antenna, and U refers to " user equipment " 2202.

Different antennae subset in 8 yuan of emission arrays is active, and this depends on the N selected by different measurements Value.For every DIDO ranks（N）, the true area of maximum that selection can be targeted to the constraint of the fixed size of 8 element array is carried out The antenna subset of covering.The standard is expected to that the space diversity of given N values can be strengthened.

Figure 23 shows available true regional for being adapted to（That is, dotted line）Different DIDO ranks array configuration.It is square Dashed box have 24 " × 24 " size, corresponding to 450MHz carrier frequencies ~ λ × λ.

Based on the commentary related to Figure 23 and refer to Figure 22, now define and in relatively following system each system property Energy：

SISO with T1 and U1（2301）

DIDO 2 × 2 with T1,2 and U1,2（2302）

DIDO 4 × 4 with T1,2,3,4 and U1,2,3,4（2303）

DIDO 6 × 6 with T1,2,3,4,5,6 and U1,2,3,4,5,6（2304）

DIDO 8 × 8 with T1,2,3,4,5,6,7,8 and U1,2,3,4,5,6,7,8（2305）

Figure 24 is shown in 4-QAM and 1/2FEC（Forward error correction）In the case of rate, SER, BER, SE in above-mentioned DIDO systems （Spectrum efficiency）With goodput performance and transmitting（TX）The functional relation of power.Observation show that SER and BER performances can be because of N Value increases and declines.The influence is caused by following two phenomenons：For fixed TX power, the input power of DIDO arrays Can be in increasing user（Or data flow）Between be divided；Space diversity can be with the number of users in actual DIDO channels Increase and reduce.

As shown in figure 24, in order to compare the relative performance of different stage DIDO systems, target BER is fixed as 10^-4（The value Can be changed according to system）, the value corresponds roughly to SER=10^-2.The TX performance numbers for corresponding to the target are referred to as TX by us Power threshold（TPT）.For any N, if TX power is less than TPT, it will be assumed that can not possibly be sent out under DIDO level ns Send, and we need to switch to the other DIDO of even lower level.Additionally, in Figure 24, observable draws, when TX power exceedes for any During the TPT of N values, SE and goodput performance can reach saturation.According to these results, self adaptation sending strategy can be designed to Switched between different stage DIDO, to strengthen SE or goodput for the fixed predeterminated target bit error rate.

Performance under II, user variable situation

Aiming at for the experiment, is emulated via in space correlation channel, evaluates the DIDO of different user position Performance.The systems of DIDO 2 × 2 are considered as with 4QAM and 1/2FEC rates.As shown in figure 25, user 1 is located at the side of emission array Penetrate（broadside）Direction, and the position of user 2 penetrates direction and is changed into end-fire from side（endfire）Direction.Transmitting antenna interval- λ/2, and it is separated by -2.5 λ with user.

Figure 26 shows the diverse location for user equipment 2, SER and the SE results per user.From the side of emission array Penetrate orientation measurement, the angle of arrival of user equipment（AOA）It is 0 ° to 90 °.Observation show that the angular distance with user equipment increases, DIDO performances will be lifted, because DIDO channel memories are in bigger diversity.Additionally, in target SER=10^-2Place, AOA2=0 ° with There is the gap of 10dB between AOA2=90 ° of both of these case.Emulated for 10 ° of angle spread in the result and Figure 35 Result is consistent.Furthermore, it is noted that in the case of AOA1=AOA2=0 °, coupling effect is there may be between two users（Cause Their antenna is adjoining caused）, this may be such that their performance is different from the simulation result in Figure 35.

III, the preferred situation for DIDO 8 × 8

Figure 24 shows the bigger SE of the producing ratio even lower level DIDO of DIDO 8 × 8, but with TX power demands higher.This point Aiming at for analysis shows to have the situation, i.e. DIDO 8 × 8 not only in peaks spectrum efficiency（SE）Aspect, but also in TX work( Rate demand（Or TPT）Aspect, surpasses DIDO2 × 2, to realize the peak value SE.

Note, in i.i.d.（It is preferable）In channel, TX power between the SE of DIDO 8 × 8 and DIDO 2 × 2 exist ~ The gap of 6dB.The gap be because of the fact caused by, i.e. DIDO 8 × 8 has been carried out between 8 data flows TX power point Cut, and DIDO 2 × 2 is split between only being flowed at two.The result is illustrated via the emulation in Figure 32.

However, in space correlation channel, TPT is communication environments characteristic（For example, array direction, customer location, angle expand Exhibition）Function.For example, Figure 35 show between two different user devices positions low angle extension ~ 15dB gaps. Similar result is illustrated in the application Figure 26.

Similar to mimo system, when user is located at the end-on direction of TX arrays, the performance of DIDO systems can decline（Because lacking Caused by few diversity）.The influence can measure by using current DIDO prototypes and observe and draw.Therefore, one kind shows The mode that DIDO 8 × 8 surpasses DIDO 2 × 2 is that user is placed in the end-on direction relative to the arrays of DIDO 2 × 2.In this feelings Shape, DIDO 8 × 8 has surpassed DIDO 2 × 2, because 8- aerial arrays provide diversity higher.

In this analysis, it is contemplated that following system：

System 1：The DIDO 8 × 8 of 4-QAM（8 parallel data streams are sent per time slot）；

System 2：The DIDO 2 × 2 of 64-QAM（Every 4 time slots, are once sent to sending user X and Y）.For this System, it is contemplated that four kinds of combinations of TX and RX aerial positions：a）T1,T2U1,2（End-on direction）；b）T3,T4U3,4（End-fire Direction）；c）T5,T6U5,6（It is separated by ~ 30 ° with end-on direction）；d）T7,T8U7,8（NLOS（Without sighting distance））；

System 3：The DIDO 8 × 8 of 64-QAM；And

System 4：The MISO 8 × 1 of 64-QAM（Every 8 time slots, are once sent to sending user X）.

For all these situations, 3/4 FEC rates are used.

Figure 27 depicts the position of user.

In Figure 28, SER results are shown because different array direction and customer location are between system 2a and 2c A ~ 15dB gap（It is similar to the simulation result in Figure 35）.The first subgraph in a second row shows SE curve saturations TX power value（That is, corresponding to BER 1e-4）.It is observed that system 1 than system 2 for relatively low TX power demands（It is small In ~ 5dB）Generate the SE of each bigger user.Spatial multiplexing gain of the DIDO 8 × 8 on DIDO 2 × 2 is additionally, since, Benefit of DIDO8 × 8 compared to DIDO 2 × 2 is for DL（Downlink）It is more obvious for SE and DL goodputs.By In the array gain of beam forming（That is, with the MRC of MISO 8 × 1）, system 4 is than system 1 with lower TX power demands （Less than 8dB）.But system 4 only generates the 1/3 of the SE of each user compared to system 1.Poor performance of the system 2 than system 1 （That is, relatively low SE is generated for larger TX power demands）.Finally, system 3 than system 1 for larger TX power demands （~15dB）Generate much bigger SE（Because larger exponent number (larger order) is modulated）.

According to these results, can be inferred that to draw a conclusion：

A kind of channel conditions are confirmed to be DIDO 8 × 8 surpasses DIDO 2 × 2（Produced for relatively low TX power demands Larger SE）；

In the channel conditions, DIDO 8 × 8 generates each bigger user's than DIDO 2 × 2 and MISO 8 × 1 SE and DLSE；And

Can be by with larger TX power demands（More than ~ 15dB）For cost uses high order modulation（That is 64-QAM, without It is 4-QAM）Further to increase the performance of DIDO 8 × 8.

Iv. there is the DIDO of day line options

Below, we assess received by IEEE journals on Signal Processing 2005 by R.Chen, " the Transmit selection diversityfor unitary precoded that R.W.Heath and J.G.Andrews are delivered Day line options described in multiuser spatial multiplexing systems with linear receivers " The benefit of algorithm.We are presented for a result for specific DIDO systems with the FEC rates of two users, 4-QAM and 1/2. Following system is compared in figure 27：

DIDO 2 × 2 with T1,2 and U1,2；And

The DIDO 3 × 2 of the use day line options with T1,2,3 and U1,2.

Position of transmitting antenna and user device location are identical with Figure 27.

Figure 29 shows DIDO 3 × 2 and the systems of DIDO 2 × 2 with day line options（Without selection）Compared to can be with The gain of offer ~ 5dB.Notice that channel is almost static state（I.e. no Doppler effect）, so selection algorithm is applied to path Loss is related to channel space, rather than rapid decay.We should see not in the situation with Doppler effect high Same gain.And, in the particular experiment, it was observed that Antenna Selection Algorithem selection antenna 2 and 3 is used to send.

Iv. it is used for the SNR threshold values of LUT

In selection [0171], we are stated model selection and are realized by LUT.LUT can be by assessing SNR threshold values come quilt Precomputation with realize in different communication environments be used for DIDO sending modes specific predefined target error rate performance.Under Face, we provides the performance of the DIDO systems having and without day line options and transformable number of users, the performance Can serve as constructing the guidance of LUT.Although Figure 24, Figure 26, Figure 28, Figure 29 are obtained by with the actual measurement of DIDO prototypes, under The figure in face is obtained by emulation.Following BER results assume no FEC.

Figure 30 shows the average BER performances of different DIDO pre-coding schemes in independent same distribution channel.Indicate and " do not have Have selection " curve refer to situation using BD.In same figure, day line options（ASel）Performance for varying number Additional antenna（For the user of varying number）And be illustrated.As can be seen that as the quantity of additional antenna increases, ASel is provided More preferable diversity gain（The slope of the BER curve in Yi Gao SNR areas is characterized）, generate preferably covering.If for example, we Target BER is fixed to 10-2（For the actual value of uncoded system）, then by the SNR gain of ASel offers with antenna Quantity increases.

Figure 31 shows the number as the extra transmitting antenna in independent same distribution channel for different target BER The SNR gain of the ASel of the function of amount.As can be seen that only by adding 1 or 2 antenna, ASel generates huge compared with BD SNR gain.In following part, we will be fixed to 10 only for 1 or 2 situation of additional antenna by by target BER^-2 （For uncoded system）To assess the performance of ASel.

Figure 32 is shown for having 1 and 2 conduct of the BD and ASel of additional antenna to use in independent same distribution channel Amount amount（M）Function SNR threshold values.It is observed that due to the larger reception SNR demands for large number of user, SNR threshold values increase with M.Note, it will be assumed that for the total transmission power that any number of user is fixation（Use varying number Transmitting antenna）.Additionally, Figure 32 shows the gain due to day line options for any amount in independent same distribution channel User for be constant.

Below, the performance of the DIDO systems in spatial correlation channel is we illustrated.We by X.Zhuang, F.W.Vook, K.L.Baum, T.A.Thomas and M.Cudak are in September, 2004 in IEEE 802.16Broadband " the Channel models for link and system delivered on Wireless Access Working Group COST-259 spatial Channel Models described in level simulations " emulate the channel of each user.We generate is used for The single group of each user.As a kind of case study, it will be assumed that NLOS channels, there is uniform linear array in emitter （ULA）, element spacing is 0.5 λ.When 2 custom systems, we are for the first and second users respectively with arrival The average angle of AOA1 and AOA2 carrys out emulation group.AOA is measured relative to the side surface direction of ULA.When having in systems more than two Individual user, we generate has in scope [- φ_m, φ_m] in evenly spaced average AOA user group, wherein we Definition

K is the quantity of user, and △ φ are the angular distances between the average AOA of user.Note angular range [- φ_m, φ_m] center It it is 0 °, direction is penetrated in the side corresponding to ULA.Below, we study work with BD and ASel delivery plans and different numbers of users For channel angle is distributed（AS）The BER performances of the DIDO systems of the function of the angular distance and between user.

Figure 33 is shown for positioned at same angle direction（Direction, AOA1=AOA2=0 ° are penetrated in side i.e. relative to ULA）'s The BER of the average SNR relative to each user of two users with different AS values.As can be seen that as AS increases, BER Performance improvement and close to independent same distribution situation.In fact, AS higher is statistically generated in two character modules of user The better performance of less covering and BD precoders between formula.

Figure 34 shows the result similar to Figure 33, but has angular distance higher between users.It is contemplated that AOA1= 0 °, AOA2=90 °（I.e. 90 ° angular distances）.Best performance is realized in the case of low AS.In fact, for angular distance high Situation, when angular distance is low, there is less crossover between the feature mode of user.It is interesting that it is observed that for just Mention it is identical the reasons why, the BER performances in low AS are better than independent same distribution channel.

Next, for 10 in different relevant situations^-2Target BER, we calculate SNR threshold values.Figure 35 is depicted The SNR threshold values of the function as AS of the different value for the average AOA of user.For low user's angular distance, with rational SNR demands（That is 18dB）Reliable transmission be possible only for the channel being characterized with AS high.On the other hand, when user exists , it is necessary to less SNR meets identical target BER when being spatially separated.

Figure 36 is shown a case that for 5 SNR threshold values of user.According to（13）In definition generation have different angles The average AOA of user of the value away from △ φ.It is observed that for △ φ=0 ° and AS<15 °, due to the small angular distance between user, BD poor performances, do not meet target BER.For the AS for increasing, the SNR demands for meeting fixed target BER reduce.The opposing party Face, for △ φ=30 °, minimum SNR demands is obtained in low AS, consistent with the result in Figure 35.As AS increases, SNR thresholds Value is saturated in independent same distribution channel.Note, correspond to [- 60 °, 60 °] with 5 △ φ=30 ° of user AOA scopes, this is for being typical with the base station in 120 ° of cellular systems of sector element.

Next, we have studied the performance of the ASel delivery plans in spatial correlation channel.Figure 37 compare for Two user situations with the 1 and 2 SNR threshold value of the BD and ASel of additional antenna.We consider the angular distance between user Two kinds of different situations：{ AOA1=0 °, AOA2=0 ° } and { AOA1=0 °, AOA2=90 ° }.For BD schemes（I.e. without day Line options）Curve it is identical with Figure 35.It is observed that ASel is generated with 1 and 2 volumes respectively for AS high The SNR gain of the 8dB and 10dB of outside antenna.With AS reduce, on BD due to ASel gain due to MIMO broadcast believe The quantity of the free degree in road is reduced and becomes smaller.It is interesting that for AS=0 °（I.e. close to LOS channels）And situation { AOA1=0 °, AOA2=90 ° }, ASel does not provide any gain due to difference in the spatial domain.Figure 38 shows and figure 37 similar results, but during for 5 users.

We are calculated as number of users in systems for BD and ASel delivery plans（M）Function SNR thresholds Value（Assuming that 10^-2General objectives BER）.SNR threshold values correspond to average SNR, to cause that total transmission power is permanent for any M Fixed.We assume that in azimuth coverage [- φ_m,φ_m- 60 ° of]=[, 60 °] in each customer group average AOA between maximum Interval.Then, the angular distance between user is △ φ=120 °/(M-1).

Figure 39 shows the SNR threshold values for the BD schemes with different AS values.It is observed that due between user Big angular distance, for the user with relatively small number（That is K≤20）AS=0.1 °（I.e. low angular spread）, obtain most Low SNR demands.However, for M>50, because △ φ are very small and BD can not be carried out, SNR demands are far longer than 40dB.This Outward, for AS>10 °, SNR threshold values are held nearly constant for any M, and the DIDO systems in spatial correlation channel are close to independently With the performance of distribution channel.

In order to reduce the value of SNR threshold values and improve the performance of DIDO systems, we apply ASel delivery plans.Figure 40 shows For with the 1 and 2 SNR threshold value in the spatial correlation channel with AS=0.1 ° of the BD and ASel of additional antenna. In order to refer to, we also reported for figure 32 illustrates independent same distribution situation curve.It can be seen that, for less User（That is M≤10）, due to lacking diversity in DIDO broadcast channels, day line options do not help reduce SNR demands.With Number of users increases, and ASel is benefited from multi-user diversity, generates SNR gain（It is 4dB i.e. for M=20）.Additionally, for M ≤ 20, the performance with 1 or 2 ASel of additional antenna in spatial correlation channel high is identical.

Then we calculate the SNR threshold values of the channel conditions other for two kinds：The AS in AS=5 ° and Figure 42 in Figure 41 =10°.Figure 41 shows due to larger angular spread that ASel is generated and is also used for relatively few number of use compared with Figure 40 Family（That is M≤10）SNR gain.As reported in Figure 42, for AS=10 °, SNR threshold values are further reduced, due to ASel's Gain becomes higher.

Finally, we summarize the result for being proposed for correlated channels at present.Figure 43 and Figure 44 show has 1 respectively With 2 additional antennas as the number of users for BD and ASel schemes（M）And angular spread（AS）Function SNR thresholds Value.Note, AS=30 ° of situation actually corresponds to independent same distribution channel, we are used to scheme in figure using this value of AS Shape is represented.It is observed that, although BD is influenceed by channel space correlation, and ASel generates almost identical for any AS Performance.Additionally, for AS=0.1 °, due to multi-user diversity, ASel is similar to BD performances for low M, and for big M（I.e. M≥20）More than BD.

Figure 49 compares the performance of different DIDO schemes in terms of SNR threshold values.The DIDO schemes for being considered are：BD、 ASel, with feature mode selection（BD-ESel）BD and maximum-ratio combing（MRC）.Notice that MRC is not eliminated in hair in advance The interference penetrated at machine（Unlike other methods）, but larger gain is provided in the case where user is spatially separated.In Figure 49 In, we depict when two users respectively be located at penetrate direction into -30 ° and 30 ° with the side of emission array when, for DIDO N × 2 systems for target BER=10^-2SNR threshold values.It is observed that for low AS, MRC schemes are carried compared with other schemes The gain of 3dB has been supplied, because the space channel of user is separated well, the influence very little of the interference between user.Note, The gain of the MRC on DIDO N × 2 is due to array gain.For the AS more than 20 °, QR-ASel schemes exceed other schemes And the gain of about 10dB is generated compared with the BD 2 × 2 without selection.QR-ASel and BD-ESel are any for AS Value provides about the same performance.

Described above is the new self adaptation transmission technology for DIDO systems.The method is between DIDO sending modes Dynamic translation strengthens the handling capacity for fixed target error rate to different users.The property of the DIDO systems of different stage Can be measured under different propagation conditions, it was observed that the huge gain in handling capacity can be used as propagation feelings by dynamic select The DIDO patterns and number of users of the function of condition is realized.

III. the precompensation of frequency and phase difference

A. background

As described above, wireless communication system transmits information using carrier wave.These carrier waves are typically sine wave, its amplitude And/or phase response is modulated in the information for being sent.The nominal frequency of sine wave is known as carrier frequency.In order to create this Waveform, emitter synthesizes one or two sine wave, and is created using up-conversion and overlap with designated carrier frequency The signal after modulation on sine wave.This can be realized by directly conversion, wherein, signal is risen on the carrier wave or by multiple The frequency conversion stage is from directly modulated.In order to process the waveform, receiver must demodulate received RF signals, and effectively move Except modulation carrier wave.This needs receiver to synthesize one or more sinusoidal signals to be reversed in the modulated process at emitter, it is known that For frequency reducing is changed.Regrettably, the sine wave signal in transmitter and receiver generation is obtained from different reference oscillators.Not yet There is reference oscillator to create perfection（perfect）Frequency reference；In fact, generally there is some deviations with actual frequency.

In a wireless communication system, the difference of the output of the reference oscillator at transmitter and receiver is at receiver Create the phenomenon for being known as carrier frequency shift or simple frequency shift (FS).Substantially, after down conversion, received Signal in have some residue modulation（Corresponding to the difference sent and received in carrier wave）.This create in received signal Distortion, result in bit error rate and relatively low handling capacity higher.

In the presence of the different technologies for processing carrier frequency shift.Most methods estimate the carrier frequency at receiver Skew, then using offset correction of carrier frequency algorithm.Carrier frequency offset estimation algorithm is blindness using following methods （blind）：Offset-QAM（T.Fusco and M.Tanda, " BlindFrequency-offset Estimation for OFDM/ OQAM Systems, " Signal Processing, IEEE Transactions on [referring also to Acoustics, Speech, and Signal Processing,IEEETransactions on],vol.55,pp.1828-1838,2007）；Cyclophysis （E.Serpedin, A.Chevreuil, G.B.Giannakis and P.Loubaton, " Blind channel and carrier frequencyoffset estimation using periodic modulation precoders,"Signal Processing, IEEETransactions on [referring also to Acoustics, Speech, and Signal Processing, IEEETransactions on],vol.48,no.8,pp.2389-2405,Aug.2000）；Or OFDM （OFDM）Cyclic Prefix in structural approach（J.J.van de Beek, M.Sandell and P.O.Borjesson, " ML estimation of time and frequency offset in OFDM systems,"Signal Processing, IEEE Transactions on are [referring also to Acoustics, Speech, and SignalProcessing, IEEE Transactions on],vol.45,no.7,pp.1800-1805,July 1997；U.Tureli, H.Liu and M.D.Zoltowski,"OFDM blind carrier offset estimation:ESPRIT,"IEEE Trans.Commun.,vol.48,no.9,pp.1459-1461,Sept.2000；M.Luise, M.Marselli and R.Reggiannini,"Low-complexity blind carrier frequencyrecovery for OFDM signals over frequency-selective radio channels,"IEEE Trans.Commun.,vol.50, no.7,pp.1182-1188,July 2002）.

Alternatively, special training signal can be utilized, including the data symbol for repeating（P.H.Moose,"A technique for orthogonal frequency division multiplexing frequency offsetcorrection,"IEEE Trans.Commun.,vol.42,no.10,pp.2908-2914,Oct.1994）；Two Different symbols（T.M.Schmidl and D.C.Cox,"Robust frequency and timingsynchronization for OFDM,"IEEE Trans.Commun.,vol.45,no.12,pp.1613-1621, Dec.1997）；Or the known symbol sebolic addressing for periodically inserting（M.Luise and R.Reggiannini, " Carrier frequency acquisition and tracking for OFDM systems,"IEEE Trans.Commun., vol.44,no.11,pp.1590-1598,Nov.1996）.Correction can occur in analog or digital mode.Receiver can be with Carry out the signal transmitted by precorrection using Carrier frequency offset estimation and eliminate skew.Because multicarrier and ofdm system are to frequency The sensitivity of skew, offset correction of carrier frequency is extensively studied for multicarrier and ofdm system（J.J.van de Beek, M.Sandell and P.O.Borjesson, " ML estimation of time and frequency offset InOFDM systems, " Signal Processing, IEEE Transactions on [referring also to Acoustics, Speech,and Signal Processing,IEEE Transactions on],vol.45,no.7,pp.1800-1805, July 1997；U.Tureli, H.Liu and M.D.Zoltowski, " OFDM blindcarrier offset estimation: ESPRIT,"IEEE Trans.Commun.,vol.48,no.9,pp.1459-1461,Sept.2000；T.M.Schmidl and D.C.Cox,"Robust frequency andtiming synchronization for OFDM,"IEEE Trans.Commun.,vol.45,no.12,pp.1613-1621,Dec.1997;M.Luise, M.Marselli and R.Reggiannini,"Low-complexity blind carrier frequency recovery for OFDM signals overfrequency-selective radio channels,"IEEE Trans.Commun.,vol.50, no.7,pp.1182-1188,July 2002）.

Frequency offset estimation and correction are for multiple antenna communication or more generally MIMO（Multiple-input and multiple-output）System It is important problem.In mimo systems, transmitting antenna is locked into a frequency reference, and receiver is locked into another frequently Rate benchmark, there is single skew between the transmitter and receiver.Propose several algorithms and ask processing this using training signal Topic（K.Lee and J.Chun, " Frequency-offset estimation for MIMO and OFDM systems using orthogonaltraining sequences,"IEEE Trans.Veh.Technol.,vol.56,no.1,pp.146-156, Jan.2007；M.Ghogho and A.Swami, " Training design for multipath channel andfrequency offset estimation in MIMO systems,"Signal Processing, IEEETransactions on [referring also to Acoustics, Speech, and Signal Processing, IEEETransactions on],vol.54,no.10,pp.3957-3965,Oct.2006；And adaptivetracking C.Oberli and B.Daneshrad, " Maximum likelihood tracking algorithms forMIMOOFDM, " in Communications,2004IEEE International Conference on,vol.4,June 20-24,2004, pp.2468-2472）.Prior problem is encountered in mimo systems, wherein, transmitting antenna is not locked into same frequency Rate benchmark, but reception antenna is locked into together.This actually occurs at space division multiple access access（SDMA）The up-link of system In, SDMA systems are considered as mimo system, and wherein different user corresponds to different transmitting antennas.In this case, frequency The compensation of skew is more complicated.Specifically, frequency shift (FS) creates the interference in the different MIMO streams for being sent.Can make It is corrected with complicated Combined estimator and equalization algorithm（A.Kannan, T.P.Krauss and M.D.Zoltowski, " Separation of cochannel signals under imperfecttiming and carrier synchronization,"IEEE Trans.Veh.Technol.,vol.50,no.1,pp.79-96,Jan.2001）, and Equilibrium after Frequency offset estimation（T.Tang and R.W.Heath, " Joint fequency offset estimation and interference cancellation forMIMO-OFDM systems[mobile radio]," 2004.VTC2004-Fall.2004IEEE 60thVehicular Technology Conference,vol.3,pp.1553- 1557,Sept.26-29,2004；X.Dai,"Carrier frequency offset estimation for OFDM/SDMA systems usingconsecutive pilots,"IEEE Proceedings-Communications,vol.152, pp.624-632,Oct.7,2005）.A few thing has processed the relevant issues of excess phase shift and tracking error, wherein surplus Remaining phase offset is estimated and compensates after Frequency offset estimation, but this work only accounts for the upper of SDMAOFDMA systems Line link（L Haring, S.Bieder and A.Czylwik, " Residual carrierand sampling fequency synchronization in multiuser OFDM systems,"2006.VTC 2006-Spring.IEEE 63rd Vehicular Technology Conference,vol.4,pp.1937-1941,2006）.When all transmittings and reception antenna During with different frequency references, there is the situation of most serious in mimo systems.On the only available work of this topic The asymptotic analysis of the evaluated error in flat fading channel is only processed（O.Besson and P.Stoica, " On parameterestimation of MIMO flat-fading channels with frequency offsets," SignalProcessing, IEEE Transactions on are [referring also to Acoustics, Speech, and SignalProcessing,IEEE Transactions on],vol.51,no.3,pp.602-613,Mar.2003）.

When the different transmitting antennas of mimo system do not have identical frequency reference, and reception antenna is independently processed from signal When, situation about having been furtherd investigate occurs.This generation is being known as distributed input and output（DIDO）Communication system（In document In also referred to as MIMO broadcast channels）Middle generation.DIDO systems include an access point with spaced antenna, the antenna hair Send data streams in parallel（Via precoding）Strengthen the handling capacity of downlink to multiple users, now wirelessly provided using identical Source（That is identical time-slot duration and frequency band）As conventional SISO systems.DIDO systems are described in detail in S.G.Perlman and T.Cotter, entitled " the System and method fordistributed that in July, 2004 submits to In the U.S. Patent application 20060023803 of input-distributed output wireless communications " Propose.There is the mode of many implementation DIDO precoders.One solution is block diagonalization（BD）, in such as documents below Description：Q.H.Spencer, A.L.Swindlehurst and M.Haardt, " Zero-forcing methods for downlink spatial multiplexing inmultiuser MIMO channels,"IEEE Trans.Sig.Proc, vol.52,pp.461-471,Feb.2004;K.K.Wong, R.D.Murch and K.B.Letaief, " A joint- channeldiagonalization for multiuser MIMO antenna systems,"IEEE Trans.WirelessComm.,vol.2,pp.773-786,JuI 2003;L.U.Choi and R.D.Murch, " A transmitpreprocessing technique for multiuser MIMO systems using a decompositionapproach,"IEEE Trans.Wireless Comm.,vol.3,pp.20-24,Jan 2004; Z.Shen, J.G.Andrews, R.W.Heath and B.L Evans, " Low complexity user Selectionalgorithms for multiuser MIMO systems with block diagonalization, " quilt Receiving is published in IEEE Trans.Sig.Proc, Sep.2005;Z.Shen, R.Chen, J.G.Andrews, R.W.Heath and B.L Evans, " Sum capacity of multiuser MIMO broadcast channels withblock Diagonalization, " is submitted to IEEE Trans.Wireless Comm., Oct.2005;R.Chen,R.W.Heath And J.G.Andrews, " Transmit selection diversity for unitaryprecoded multiuser Spatial multiplexing systems with linear receivers, " are received to IEEE Trans, on Signal Processing,2005。

In DIDO systems, send precoding and be used to separate the data flow for different user.When transmitting antenna radio frequency When chain does not share same frequency benchmark, carrier frequency shift result in the Railway Project related to system implementation.When this generation When, each antenna is effectively sent with slightly different carrier frequency.This destroys the integrality of DIDO precoders, cause each User is subjected to extra interference.Forth below are several solutions to this problem.In an implementation of solution In mode, DIDO transmitting antennas share a frequency reference by wired, optical or wireless network.In solution Another implementation method in, one or more users estimate frequency offset difference（The relative mistake in skew between antenna pair It is different）And the information is sent back into emitter.Then emitter precorrection frequency shift (FS) and proceed for DIDO training and Precoder estimates phase.The implementation method is problematic when feedback channel has delay.Reason is that possible have by trimming process The residual phase error of establishment, the trimming process does not consider subsequent channel estimation.In order to solve this problem, one other Implementation method uses new frequency shift (FS) and phase estimating device, by estimating that delay solves this problem.Based on emulation and logical The actual measurement of DIDO-OFDM prototypes execution is crossed to provide result.

The frequency and phase offset compensation method for proposing in this document may be to the estimations due to the noise at receiver Application condition is sensitive.Therefore, another implementation method proposes the method estimated for the time and frequency shift, in low SNR bars It is also very strong under part.

There are the different methods for performing time and frequency shift estimation.Due to its sensitivity to synchronous error, this Many methods in a little methods are proposed exclusively for OFDM waveforms.

These algorithms are therefore general for single carrier wave and overloading waveform without the structure for typically using OFDM waveforms It is enough.Algorithm described below is using known fiducial mark（For example, training data）To assist synchronous technology Among class.Many methods are the extensions of the frequency offset estimator of Moose（See P.H.Moose, " A technique for orthogonal frequency divisionmultiplexing frequency offset correction,"IEEE Trans.Commun.,vol.42,no.10,pp.2908-2914,Oct.1994）.Moose is proposed and is used two instructions of repetition Practice signal and obtain frequency shift (FS) using the phase difference between received signal.The method of Moose can only be corrected Fraction（fractional）Frequency shift (FS).The extension of the method for Moose is proposed by Schmidl and Cox（T.M.Schmidl and D.C.Cox,"Robust frequency and timing synchronization forOFDM,"IEEE Trans.Commun.,vol.45,no.12,pp.1613-1621,Dec.1997）.Their main innovation is to use one The training symbol of periodic OFDM symbol and other differential coding.The integer that differential coding in second symbol is realized Offset correction.Coulson is considered in T.M.Schmidl and D.C.Cox, " Robust frequency and timing synchronization forOFDM,"IEEE Trans.Commun.,vol.45,no.12,pp.1613-1621, Similar setting described in Dec.1997, and in A.J.Coulson, " Maximum likelihood synchronization forOFDM using a pilot symbol:analysis,"IEEE J.Select.Areas Commun., vol.19, no.12, pp.2495-2503, Dec.2001 and A.J.Coulson, " Maximum likelihoodsynchronization for OFDM using a pilot symbol:algorithms,"IEEE J.Select.Areas Commun., vol.19, no.12, pp.2486-2494, provide algorithm with analysis in Dec.2001 It is discussed in detail.One main difference is that the correlation properties that are provided using the maximal-length sequence for repeating of Coulson.He Also suggestion uses linear frequency modulation（chirp）Signal, because its constant envelope properties in time domain and frequency domain.Coulson is examined The details of reality is considered but not including integer estimation.Multiple repeat training signals by Minn et.al.in H.Minn, V.K.Bhargava and K.B.Letaief, " A robust timing and frequency synchronization for OFDM systems, " IEEETrans.Wireless Commun., vol.2, no.4, pp.822-839, July 2003 is examined Consider, but the structure of training is not optimised.Shi and Serpedin propose training structure has the one of the idea for forming frame synchronization A little optimalitys（K.Shi and E.Serpedin, " Coarse frame and carriersynchronization of OFDM systems:a new metric and comparison,"IEEE Trans.Wireless Commun.,vol.3,no.4, pp.1271-1284,July 2004）.An embodiment of the invention performs frame using the method for Shi and Serpedin Synchronous and fractional frequency offset is estimated.

Many methods in the literature are concentrated in frame synchronization and fractional frequency offset correction.Integer offset correction is used T.M.Schmidl and D.C.Cox, " Robust frequency and timingsynchronization for OFDM, " IEEE Trans.Commun., vol.45, no.12, pp.1613-1621, other training symbol in Dec.1997 are solved Certainly.For example, Morrelli etc. is in M.Morelli, A.N.D'Andrea and U.Mengali, " Frequency ambiguity Resolution inOFDM systems, " IEEE Commun.Lett., vol.4, no.4, pp.134-136, in Apr.2000 T.M.Schmidl and D.C.Cox is obtained, " Robust frequency and timingsynchronization for The OFDM, " modified version of IEEE Trans.Commun., vol.45, no.12, pp.1613-1621, Dec.1997.Using not The interchangeable method of same preamble structure is proposed by Morelli and Mengali（M.Morelli and U.Mengali, " An improvedfrequency offset estimator for OFDM applications,"IEEE Commun.Lett., vol.3,no.3,pp.75-77,Mar.1999）.This method has used the correlation between the M identical training symbol for repeating Property come by the M factors increase fractional frequency offset estimator scope.This is best linear unbiased estimator and receives most Big skew（With suitable design）, but the timing synchronization not provided.

System describe

An embodiment of the invention eliminates the frequency in DIDO systems using the precoding based on channel condition information Rate and phase offset.See Figure 11 and for the associated description above the description of the implementation method.

In an embodiment of the invention, each user is using being equipped with connecing for frequency offset estimator/compensator Receipts machine.Go out as shown in Figure 45, in an embodiment of the invention, including the system of receiver includes multiple RF units 4508th, corresponding multiple A/D units 4510, the receiver and DIDO that are equipped with frequency offset estimator/compensator 4512 are anti- Feedback maker unit 4506.

RF units 4508 receive the signal sent from DIDO transmitter units, and signal down is transformed into base band, and will drop Signal after frequency is changed is provided to A/D units 4510.Then signal from analog is converted to numeral by A/D units 4510, and by it It is sent to frequency offset estimator/compensator units 4512.Frequency offset estimator/compensator units 4512 estimate frequency shift (FS) And compensating frequency deviation, as described here, the signal after compensation is then sent to OFDM units 4513.OFDM units 4513 removal Cyclic Prefix simultaneously run FFT（FFT）By signal reports to frequency domain.During the training period, OFDM is mono- Unit 4513 sends output to channel estimating unit 4504 to calculate channel estimation in a frequency domain.Alternatively, channel estimation Can be calculated in the time domain.During the data cycle, OFDM units 4513 send output to DIDO receiver units 4502, The DIDO receiver units 4502 are demodulated/decode to obtain data to signal.Channel estimating unit 4504 is by channel estimation DIDO feedback generators unit 4506 is sent to, the DIDO feedback generators unit 4506 can be estimated and via anti-with quantized channel They are sent back emitter by feedback control channel, as shown.

A description for implementation method to the algorithm for the situations of DIDO 2 × 2

The implementation method of the algorithm for the frequency/phase migration in DIDO systems is described below.DIDO systems System model starts to be described in the case where being offset with and without frequency/phase.For simplicity, there is provided DIDO 2 × 2 The particular implementation of system.However, general principle of the invention can also be carried out in high-order DIDO systems.

With/DIDO the system models without frequency and phase offset

The received signal of DIDO 2 × 2 can be write as first user：

r₁[t]=h₁₁(w₁₁x₁[t]+w₂₁x₂[t])+h₁₂(w₁₂x₁[t]+w₂₂x₂[t]) （1）

And write as second user：

r₂[t]h₂₁(w₁₁x₁[t]+w₂₁x₂[t])+h₂₂(w₁₂x₁[t]+w₂₂x₂[t]) （2）

Wherein t is discrete time index, h_mnAnd w_mnIt is respectively the channel between m-th user and n-th transmitting antenna With DIDO precoding weights, x_mIt is for the sending signal of user m.Note, h_mnAnd w_mnIt is not the function of t, as we assume that Channel is constant on the cycle between training and data is activation.

In the presence of frequency and phase offset, the signal for receiving is represented as

And

Wherein, T_sIt is symbol period；For n-th transmitting antenna, ω T_n=2 ∏ f_Tn；For m-th user, W_Um=2 ∏ F_Um；And f_TnAnd f_UmIt is respectively for the practical carrier frequency of n-th transmitting antenna and m-th user（By bias effect）.Value t_mnRepresent in channel h_mnOn cause the random delay of phase offset.Figure 46 depicts the system models of DIDO 2 × 2.

For the time, we use defined below：

Δω_mn=ω_Um-ω_Tn（5）

For representing the frequency shift (FS) between m-th user and n-th transmitting antenna.

The description of an embodiment of the invention

The method of an implementation method of the invention is illustrated in Figure 47.The method includes following general step （Including sub-step, as shown）：For the cycle of training 4701 of Frequency offset estimation；For the cycle of training of channel estimation 4702；Via the data is activation 4703 for having balanced DIDO precodings.These steps are described in detail in the following.

（a）For the cycle of training of Frequency offset estimation（4701）

During the first cycle of training, one or more training signals from each transmitting antenna are sent to use by base station One in family（4701a）.As described here, " user " is wireless client device.When DIDO 2 × 2, The signal received by m-th user is given by：

Wherein, p₁And p₂It is respectively the training sequence sent from the first and second antennas.

M-th user can use any type of frequency offset estimator（I.e. by the convolution of training sequence）And estimate Shifted by delta ω_mlWith Δ ω_m2.Then, according to these values, user calculates the frequency shift (FS) between two emitting antennas：

Δω_T=Δ ω_m2-Δω_m1=ω_T1-ω_T2（7）

Finally, exist（7）In value be fed back to base station（4701b）.

Note,（6）In p₁And p₂Be designed to it is orthogonal so that user can estimate Δ ω_mlWith Δ ω_m2.Can Alternatively, in one embodiment, identical training sequence is used in two continuous time slots, and user therefrom estimates skew. Additionally, in order to improve（7）In skew estimation, all users of the calculating for DIDO systems of same as described above（No Only for m-th user）For can be done, last estimation can be the value obtained from all users（After weighting） Average value.However, this solution needs more calculating time and feedback quantities.Finally, the renewal of Frequency offset estimation There are when frequency shift (FS) is changed over time just needs.Therefore, according to the stability of the clock at emitter, 4701 the step of algorithm Can be performed in long-term basis（I.e. for each data is activation）So that above-mentioned feedback is reduced.

（b）For the cycle of training of channel estimation（4702）

During the second cycle of training, base station first from m-th user or from multiple users or must have（7）In value Frequency shift (FS) feedback.（7）In value be used for precompensation transmitting terminal frequency shift (FS).Then, base station is by training data All users are sent to come for channel estimation（4702a）.

For the systems of DIDO 2 × 2, the signal received at first user is given by：

And at second user：

Wherein,And Δ t is random or known between the first transmission of base station and second send Postpone.Additionally, p₁And p₂It is respectively the training sequence sent from the first and second antennas of user's frequency shift (FS) and channel estimation.

Note, precompensation is only applied to the second antenna in this embodiment.

Launch（8）, we obtain

It is similarly for second user：

Wherein,

In receiving terminal, user is by using training sequence p₁And p₂Carry out compensating frequency offset residue.Then, user is by instruction Practice vector channel to be estimated（4702b）：

（12）In these channels or channel condition information（CSI）It is fed back to base station（4702b）, base station such as lower face DIDO precoders are calculated described in point.

（c）DIDO precodings with precompensation（4703）

Base station receives from user（12）In channel condition information（CSI）And by block diagonalization（BD）To calculate precoding Weight（4703a）, to cause

Wherein, vector h₁（12）In be defined, and w_m=[w_m1,w_m2].Note, the present invention for proposing in the disclosure Can be used in any other DIDO method for precoding in addition to BD.Base station also by using（7）In estimation come mend in advance Frequency shift (FS) is repaid, and the delay between sending and being currently transmitted is trained by estimating second（Δt₀）To pre-compensate for phase offset （4703a）.Finally, base station transmits data to user via DIDO precoders（4703b）.

After transmission process, the signal received at user 1 is given by：

Wherein,Use attribute（13）, we obtain

Similarly, for user 2, we obtain：

Launch（16）：

Wherein,

Finally, user calculates frequency offset residue and channel estimation carrys out demodulated data stream x₁[t] and x₂[t]（4703c）.

It is generalized to DIDO N × M

In the portion, the technology for describing before is generalized to the DIDO systems with N number of transmitting antenna and M user System.

I. the cycle of training of user's Frequency offset estimation

During the first cycle of training, due to the letter that the training sequence that is sent from N number of antenna is received by m-th user Number it is given by：

Wherein, p_nIt is from n-th training sequence of antenna transmission.

Estimating shifted by delta ω_mnAfterwards,M-th user calculate first and n-th transmitting antenna it Between frequency shift (FS)：

Δω_{T, 1n}=Δ ω_mn-Δω_m1=ω_T1-ω_Tn（19）

Finally,（19）In value be fed back to base station.

Ii. it is used for the cycle of training of channel estimation

During the second cycle of training, base station first from m-th user or from multiple users obtain with（19）In The frequency shift (FS) feedback of value.（19）In value be used for precompensation transmitting terminal frequency shift (FS).Then, base station will train number Come for channel estimation according to all users are sent to.

For DIDO N × M systems, the signal received at m-th user is given by：

Wherein,And Δ t is the first and second transmissions of base station Between random or known delay.Additionally, P_nIt is the training sequence sent from n-th antenna for frequency shift (FS) and channel estimation Row.

In receiving side, user is by using training sequence P_nCarry out compensating frequency offset residue.Then, each user m passes through Trained vector channel is estimated：

And base station is fed back to, base station calculates DIDO precoders as described in following part.

Iii. there is the DIDO precodings of precompensation

Base station receives from user（12）In channel condition information（CSI）And by block diagonalization（BD）To calculate precoding Weight, with cause

Wherein, vector h_m（21）In be defined, and w_m=[w_m1,w_m2,...,w_mN].Base station also by using（19）In Estimation pre-compensate for frequency shift (FS), and by estimating that the second training sends and be currently transmitted between delay（Δt₀）Mend in advance Repay phase offset.Finally, base station transmits data to user via DIDO precoders.

After transmission process, the signal received at user i is given by：

Wherein,Use attribute（22）, we obtain：

Finally, user calculates frequency offset residue and channel estimation carrys out demodulated data stream x_i[t]。

As a result

Figure 48 shows the SER results of the systems of DIDO 2 × 2 having and without frequency shift (FS).It can be seen that, carried The method for going out completely eliminates frequency/phase skew, generates and the system identical SER without skew.

Next, we assess fluctuation of the proposed compensation method for frequency offset error and/or real time offset Sensitivity.Therefore, we will（14）It is rewritten as：

Wherein, ε represents evaluated error and/or the change of the frequency shift (FS) between training and data is activation.Note, the effect of ε Fruit is destruction（13）In orthogonal property, with cause（14）With（16）In distracter at emitter not disappeared in advance completely Remove.Because so, SER performances are reduced with the ε values of increase.

Figure 48 shows the SER performances of the frequency offset compensation method for different ∈ values.These results assume T_s= 0.3ms（There is the signal of 3KHz bandwidth）.It is observed that for ε=0.001Hz（Or it is less）, SER performances with without inclined The situation of shifting is similar.

F. it is used for a description for implementation method of the algorithm that the time and frequency shift is estimated

Below, we describe to perform the other implementation method that the time and frequency shift is estimated（4701b in Figure 47）.Examine The transmission signal structure of worry in H.Minn, V.K.Bhargava and K.B.Letaief, " A robust timing and frequency synchronization for OFDM systems,"IEEETrans.Wireless Commun.,vol.2, Proposed in no.4, pp.822-839, July 2003, in K.Shi and E.Serpedin, " Coarse frame and carrier synchronization of OFDM systems:a new metric and comparison,"IEEE Studied in detail in Trans.Wireless Commun., vol.3, no.4, pp.1271-1284, July 2004.Generally have The sequence of good association attributes is used for training.For example, for our system, Chu sequences are used, and Chu sequences such as exist D.Chu,"Polyphasecodes with good periodic correlation properties(corresp.)," It is described in IEEE Trans.Inform.Theory, vol.18, no.4, pp.531-532, July 1972.These sequences have Interesting attribute, i.e., they are with perfect circulation correlation.Allow L_cpRepresent the length of Cyclic Prefix, N_tRepresent component training sequence The length of row.So that N_t=M_t, wherein M_tIt is the length of training sequence.Under these assumptions, the transmitted symbol for the beginning Sequence can be written to：

s[n]=t[n-N_t] for n=-1 ... ,-L_cp

S [n]=t [n] for n=0 ..., N_t-1

s[n]=t[n-N_t] for n=N_t,…,2N_t-1

s[n]=-t[n-2N_t] for n=2N_t,…,3N_t-1

s[n]=t[n-3N_t] for n=3N_t,…,4N_t-1

Noting the structure of the training signal can be scaled up to other length, but block structure is repeated.For example, in order that with 16 training signals, it is contemplated that a kind of structure, for example：

[CP,B,B,-B,B,B,B,-B,B,-B,-B,B,-B,B,B,-B,B,]。

By using the structure, and make N_t=4M_t, all algorithms that will be described can without modification in the case of be made With.Effectively, our repetition training sequences.This is particularly useful in the case of suitable training signal possibility is disabled.

After in the filtering matched to symbol rate and to down-sampling, it is considered to received signal below：

Wherein ε is unknown discrete-time frequency skew, and Δ is unknown vertical shift, and h [l] is unknown discrete time letter Road coefficient, and v [n] is additional noise.In order to explain the key idea in following part, ignore the presence of additional noise.

I. rough frame synchronization

The purpose of rough frame synchronization is to solve unknown vertical shift Δ.Let us is made defined below：

The rough frame synchronization algorithm for being proposed from K.Shi and E.Serpedin, " Coarse frame andcarrier synchronization of OFDM systems:a new metric and comparison," Algorithm in IEEETrans.Wireless Commun., vol.3, no.4, pp.1271-1284, July 2004 takes a hint, Obtained according to maximum-likelihood criterion.

The improved rough frame synchronization of method 1-：Rough frame synchronization estimator solves following optimization：

Wherein,

So that the signal being corrected is defined as：

Small inceptive impulse that other correction term is used to compensate in channel can be simultaneously conditioned based on application.This volume Outer delay will be included in the channel afterwards.

Ii. fractional frequency offset correction

Fractional frequency offset is corrected after rough frame synchronization block.

The improved fractional frequency offset corrections of method 2-：Fractional frequency offset is following solution：

This is known as fractional frequency offset, because algorithm only can be with correcting offset

This problem will be solved in next part.Fine frequency offset correction signal is allowed to be defined as：

Note, method 1 and 2 is for preferable K.Shi, the E.Serpedin, " Coarse of being worked in frequency selective channel frame and carrier synchronization of OFDM systems:a newmetric and Comparison, " IEEE Trans.Wireless Commun.'s, vol.3, no.4, pp.1271-1284, July 2004 changes Enter.Here one especially innovation be the use of r recited above andUse improve former estimator because it Have ignored the sampling being affected because internal symbol is disturbed.

Iii. integer frequency deviation correction

In order to correct integer frequency deviation, it is necessary to write one for received after fine frequency offset correction The equivalent system model of signal.The timing error of reservation is absorbed in channel, does not have noisy received signal to have Following structure：

Wherein n=0,1 ..., 4N_t–1.Integer frequency deviation is k, and unknown equivalent channels are g [l].

The improved integer frequency deviation corrections of method 3-：Integer frequency deviation is following solution：

Wherein：

R=D [k] Sg

These give the estimation of total frequency shift (FS)：

In fact, method 3 has complexity very high.In order to reduce complexity, following observation can be made.First, multiply Product S (S*S)^-1S can be expected to calculate.Regrettably, this still leaves sizable matrix multiplication.Alternatively use Observation with the training sequence for being proposed, S*S ≈ I.This generates the method for the support type of following reduction.

The improved integer frequency deviation correction of method 4- low-complexities：

The integer frequency deviation estimator of low-complexity is solved

Iv. result

In the portion, we compare the performance of the different estimators for being proposed.

First, in Figure 50, we compare every kind of method required for expense amount.Notice that two kinds of new methods will be opened Pin reduces 10 times to 20 times.For the performance of relatively more different estimators, MonteCarlo experiments are performed.What is considered sets It is to send waveform from our common NVIS of the linear modulation construction with 3K symbols symbol rate per second to put, and is corresponded to The pass band width of 3KHz, and the cosine impulse shaping for rising.For each Monte Carlo realize, frequency shift (FS) from [- f_max,f_max] on be uniformly distributed and generate.

With f_maxThe small frequency shift (FS) and the emulation without integer offset correction of=2Hz are illustrated in Figure 51.Can be with Find out from the Performance comparision, with N_t/M_t=1 performance slightly degrades from original estimator, although substantially reduce expense. With N_t/M_t=4 performance is more preferable, almost 10dB.Due to the error in integer bias estimation, all curves are in low SNR Point experienced complications.Small error in integer skew can create big frequency error and big splicing square error.It is whole Number offset correction can be turned off to improve performance in small skew.

In the presence of multi-path channel, the performance of frequency offset estimator is typically reduced.However, in Figure 52, Close integer offset estimator and present extraordinary performance.Therefore, in multi-path channel, the rough correction of robust is being performed Improved fine correction algorithm afterwards is prior.Note, with N_t/M_t=4 offset behavior is good in the case of multipath Much.

Embodiments of the present invention can include various steps suggested above.The step can be executable with machine The mode of instruction realizes that the instruction causes universal or special computing device particular step.For example, base described above Stand/AP and customer set up in various assemblies may be implemented as on universal or special processor perform software.In order to keep away Exempt from the various known personal computer of the fuzzy parties concerned of the invention, computer storage, hard disk, input unit etc. Component is saved from figure.

Alternatively, in one embodiment, various functions module herein shown and correlation step can pass through Specialized hardware components comprising the hardwired logic for performing step（Such as application specific integrated circuit（ASIC））Or by program calculation Any combination of thermomechanical components and custom hardware components and be performed.

In one embodiment, the particular module of coding as described above, modulation and signal processing logic 903 can With in programmable digital signal processor（DSP）（Or DSP groups）On be carried out, the DSP for example uses Texas Instrument（Texas Instruments）TMS320x frameworks DSP（For example, TMS320C6000, TMS320C5000 etc.）.In this embodiment DSP can be embedded in the package card of personal computer, for example, pci card.Certainly, general principle of the invention is being met In the case of, a variety of DSP architectures can be used.

Various parts of the invention can also be provided for storing the machine readable media of machine-executable instruction.Machine Device computer-readable recording medium can include but is not limited to flash memory, CD, CD-ROM, DVD ROM, RAM, EPROM, EEPROM, magnetic card or light Learn card, communication media or be suitable to store the other types of machine readable media of e-command.For example, the present invention can be downloaded Be computer program, the computer program can by way of being included in the data-signal of carrier wave or other communication medias via Communication link（For example, modem or network connection）From remote computer（Such as server）It is sent to requesting computer （Such as client）.

Throughout described above, for purposes of explanation, many specific details are proposed to provide the complete of system and method Foliation solution.It will be apparent, however, to one skilled in the art that system and method can be without these certain details It is implemented in the case of some in section.Therefore, the scope of the present invention and essence should be judged according to appended claims.

Additionally, in the foregoing written description, many documents are cited to provide of the invention being more completely understood.It is all these to draw Bibliography is by reference to being integrated into the application.

Claims (33)

1. one kind is for compensating the frequency of multi-user multi-aerial system (MU-MAS) communication and the system of phase offset, the system bag Include：

One or more coded modulation units, for the letter for each wireless client device in multiple wireless client devices Breath bit is encoded and modulated with the information bit after generating coding and modulation；

One or more map units, for the information bit after the coding and modulation to be mapped as into complex symbol；And

The skew of MU-MAS frequency/phases perceives precoding unit, for calculating MU-MAS frequencies/phase using channel condition information Position skew perceives precoding weight, and MU-MAS frequency/phases skew perceives precoding unit using the weight to from institute Stating the complex symbol of map unit acquisition carries out precoding to eliminate frequency/phase skew and/or inter-user interference in advance.

2. system according to claim 1, wherein the channel condition information is by from the wireless client device to institute The skew of MU-MAS frequency/phases is stated to perceive the feedback of precoding unit and be obtained.

3. system according to claim 1, wherein the channel condition information in the downlink is in the MU-MAS Frequency/phase skew is perceived precoding unit and is obtained from up-link by using channel reciprocity.

4. system according to claim 1, the system also includes：One or more OFDM (OFDM) units, Perceive the signal after the precoding of precoding unit from MU-MAS frequency/phases skew and marked according to OFDM for receiving Standard modulates the signal after the precoding.

5. system according to claim 4, wherein the OFDM standards include calculating fast fourier inverse transformation (IFFT) With addition Cyclic Prefix.

6. system according to claim 4, the system also includes：One or more D/A units, for mono- in the OFDM Digital-to-analogue (D/A) conversion is performed in the output of unit to generate analog baseband signal；And one or more radio frequency (RF) units, it is used for The analog baseband signal up-conversion is carried out into sending signal for radio frequency and using corresponding one or more transmitting antennas.

7. system according to claim 1, wherein MU-MAS frequency/phases skew perceives precoding unit being carried out For the MMSE precoders after least mean-square error (MMSE) precoder, weighting, zero force (ZF) precoder or block diagonal Change (BD) precoder.

8. system according to claim 1, wherein being sent out by between multiple transmitting antennas or the wireless client device Training sequence is sent to estimate the channel condition information, and the channel condition information is used to estimating between the transmitting antenna Frequency or phase offset.

9. a kind of for compensating inphase quadrature (I/Q) unbalanced system that multi-user multi-aerial system MU-MAS communicates, this is System includes：

One or more coded modulation units, for the letter for each wireless client device in multiple wireless client devices Breath bit is encoded and modulated with the information bit after generating coding and modulation；

One or more map units, for the information bit after the coding and modulation to be mapped as into complex symbol；And

MU-MAS IQ perceive precoding unit, and precoding weight is perceived for calculating MU-MASIQ using channel condition information, The MU-MAS IQ are perceived precoding unit the complex symbol obtained from the map unit is prelisted using the weights Code eliminates the interference and/or inter-user interference that are brought due to I/Q gains and unbalance in phase with pre-.

10. system according to claim 9, wherein the channel condition information is by from the wireless client device to institute MU-MAS IQ are stated to perceive the feedback of precoding unit and be obtained.

11. systems according to claim 9, wherein the channel condition information in the downlink is in the MU-MAS IQ is perceived precoding unit and is obtained from up-link by using channel reciprocity.

12. systems according to claim 9, the system also includes：

One or more OFDM (OFDM) units, precoding unit is perceived for receiving from the MU-MASIQ Signal after precoding and the signal after the precoding according to OFDM classical modulations.

13. systems according to claim 12, wherein the OFDM standards include calculating fast fourier inverse transformation (IFFT) and addition Cyclic Prefix.

14. systems according to claim 12, the system also includes：

One or more D/A units, for performing digital-to-analogue (D/A) conversion in the output of the OFDM units to generate simulation base Band signal；And

One or more radio frequency (RF) units, for being radio frequency and use corresponding by the analog baseband signal up-conversion Individual or multiple transmitting antennas carry out sending signal.

15. systems according to claim 9, wherein the MU-MAS IQ perceive precoding unit be implemented as wave beam into Shape or maximum-ratio combing (MRC).

16. systems according to claim 9, wherein the MU-MAS IQ perceive precoding unit is implemented as minimum MMSE precoders, zero pressure (ZF) precoder or block diagonalization (BD) after square error (MMSE) precoder, weighting is pre- Encoder.

17. systems according to claim 9, wherein being sent out by between multiple transmitting antennas or the wireless client device Training sequence is sent to estimate the channel condition information, and the channel condition information is used to estimate the I/ of the transmitting antenna Q-gain and unbalance in phase.

18. systems according to claim 9, including one or more RF units are with the conversion signal between base band and RF.

A kind of system of 19. communication characteristics for dynamically adapting multi-user multi-aerial system (MU-MAS) communication system, including：

One or more coded modulation units, for the letter for each wireless client device in multiple wireless client devices Breath bit is encoded and modulated with the information bit after generating coding and modulation；

One or more map units, for the information bit after the coding and modulation to be mapped as into complex symbol；And

MU-MAS configurator units, subset and MU-MAS sending modes for determining user based on channel characteristics data, and Responsively control the coded modulation unit and the map unit.

20. systems according to claim 19, wherein the channel characteristics data by from the wireless client device to The feedback of the MU-MAS configurators unit and be obtained.

21. systems according to claim 19, wherein the channel characteristics data in the downlink are in the MU- MAS configurators unit is obtained by using channel reciprocity from up-link.

22. systems according to claim 19, the system also includes：

MU-MAS precoding units, operate to calculate in data-signal quilt under the control of the MU-MAS configurators unit The precoding weight of precoding is carried out before being sent to the wireless client device to the data-signal.

23. systems according to claim 22, the system also includes：

One or more OFDM (OFDM) units, for receiving the precoding from the MU-MAS precoding units Signal after rear signal and the precoding according to OFDM classical modulations.

24. systems according to claim 23, wherein the OFDM standards include calculating fast fourier inverse transformation (IFFT) and addition Cyclic Prefix.

25. systems according to claim 23, the system also includes：

One or more D/A units, for performing digital-to-analogue (D/A) conversion in the output of the OFDM units to generate simulation base Band signal；And

One or more radio frequency (RF) units, for being radio frequency and use corresponding by the analog baseband signal up-conversion Individual or multiple transmitting antennas carry out sending signal.

26. systems according to claim 22, wherein the MU-MAS precoding units are implemented as least mean-square error (MMSE) MMSE precoders after precoder, weighting, zero pressure (ZF) precoder or block diagonalization (BD) precoder.

27. systems according to claim 19, wherein by between multiple transmitting antennas and the wireless client device Training sequence is sent to estimate the channel characteristics data, and the channel characteristics data are used for evaluating wireless customer set up A subset, transmitting antenna or the transmission mode for the MU-MAS communication systems.

28. systems according to claim 19, wherein the MU-MAS configurators unit uses polarization and/or directional diagram point Method of the collection technology as array sizes are reduced, while obtaining diversity on wireless links.

29. systems according to claim 19, wherein communicate occurring via nearly vertical incidence sky wave (NVIS) and/or earthwave As increase diversity and the method for downlink throughput.

30. systems according to claim 19, wherein directional diagram diversity are used to be carried out with specific user via earthwave Communicate and communicated with other users via NVIS.

31. systems according to claim 29, wherein each client using earthwave and NVIS links be spatially separating come As the method for the space diversity for increasing link.

32. systems according to claim 19, the system also includes base station, and the base station is used to be based on the letter from client Road Quality Feedback adaptively to switch between different array geometry structures and different antennae diversity technique, as increase chain The diversity on road and the method for downlink throughput.

33. systems according to claim 19, the system also includes base station, and the base station defines user's group and is based on theirs Priority and/or channel conditions dispatch different groups of user to be transmitted.