CN101969417A - Low-return self-adaptive multimode transmission method of MIMO-SCFDE (Multiple Input Multiple Output Single Carrier Frequency Domain Equilibrium) system - Google Patents

Abstract

The invention relates to a low-return self-adaptive multimode transmission method of an MIMO-SCFDE (Multiple Input Multiple Output Single Carrier Frequency Domain Equilibrium) system, which comprises the following steps of: (1) determining the rank information R of each subchannel by a receiving end, and sending the R to a sending end through a feedback channel; (2) forming the time domain data frames of M modulated symbols by the sending end, converting the time domain data frames into M-dimensional equivalent frequency domain data frames, generating frequency domain channel mark information D, mapping the M-dimensional frequency domain data frames according to the D, expanding into a row of N*NT-dimensional frequency domain data frames, generating NT rows of N-dimensional frequency domain data frames through serial/parallel conversion, converting the NT parallel N-dimensional frequency domain data frames back into time domain data frames, and respectively sending the time domain data frames by NT sending antennas after adding CP (Code Pulse); and (3) equalizing the data frames received by NR receiving antennas by the receiving end after removing the CP, recovering into one row of N*NT-dimensional data frames through parallel and serial conversion, selecting M usable signals, converting back into time domain data frames, and judging the symbols. The invention has the characteristics of high system frequency spectrum efficiency and low return information amount.

Description

The MIMO-SCFDE system self-adaption multimode transmission method of low passback

Technical field

The present invention relates to a kind of many antennas broadband wireless communications transmission method, belong to the broadband wireless communication technique field.

Background technology

Along with development of internet technology, people also improve constantly the requirement of access network, enter the Internet at a high speed whenever and wherever possible and become increasing people's important need, broadband wireless communication technique is to satisfy the main support technology of people's the demand, and therefore broadband wireless communication technique has obtained fast development in recent years.Increase along with transmission rate, the multipath that the electromagnetic wave radio transmission causes is more and more serious to systematic influence, generally speaking the frequency selective fading that exists multipath transmisstion to cause inevitably in the system of broadband wireless communication, frequency selective fading once were one of principal elements of restriction performance in wireless communication systems.The piecemeal transmission technology based on Cyclic Prefix (hereinafter to be referred as CP:Cyclic Prefix) of being got up by OFDM (hereinafter to be referred as OFDM:Orthogonal Frequency Division Multiplexing) technical development (mainly comprises OFDM, single carrier frequency domain equalization (hereinafter to be referred as SCFDE:Single Carrier with Frequency Domain Equalization) etc.) be the simple and highly effective technology that tackles the frequency selective fading that multipath transmisstion causes in the broadband wireless communications, so OFDM and SCFDE become the mainstream technology of present broadband wireless communications.Spectrum efficiency is the research emphasis of wireless communication technology always, and multiple-input and multiple-output (hereinafter to be referred as MIMO:Multiple-input Multiple-output) is subjected to extensive concern with its conventional single-antenna spectrum efficiency that technology is beyond one's reach in recent years.MIMO and become the key technology of future wireless in conjunction with MIMO-OFDM and MIMO-SCFDE based on the piecemeal transmission technology of CP.

MIMO utilizes in the wireless propagation environment with rich multipath the uncorrelated characteristic of channel between the different antennae, obtains the high channel capacity, thereby improves the availability of frequency spectrum and reliability.OFDM based on the piecemeal transmission can resist multipath fading effectively, because the subcarrier spectrum main lobe is overlapping, has higher spectrum efficiency; CP can well absorb inter-frame-interference; And can take simple frequency-domain equilibrium method to eliminate because the channel disturbance that the time delay expansion is introduced; The baseband modulation process of OFDM can be finished with invert fast fourier transformation (hereinafter to be referred as IFFT:Inverse Fast Fourier Transform), and the base band demodulating process can be with fast fourier transform (hereinafter to be referred as FFT:Fast Fourier Transform)) finish.Single carrier frequency domain equalization (hereinafter to be referred as SCFDE:Single Carrier with Frequency Domain Equalization), be similar to the piecemeal transmission technology of OFDM, though the high peak-to-average power ratio (PAPR that can effectively resist multipath fading equally and not have OFDM to make a start, Peak-to-average Power Ratio), still be subjected to degree of concern far away from OFDM.

The subject matter that MIMO broadband wireless communications communication (mainly being MIMO-OFDM and MIMO-SCFDE) faces is: the time, frequently, the channel fading that causes of empty selectivity.This decline shows as scarce order of local channel matrix or the ill-condition that receiving terminal removes a series of arrowbands mimo channel that obtains behind the CP, and this scarce order or ill-condition cause very big influence to the detection of MIMO signal, is the principal element of restriction wideband MIMO systematic function.

When utilizing the adaptive technique of channel condition information (hereinafter to be referred as CSI:Channel State Information) to resist effectively, frequently, empty selectivity, thereby more reliably more effectively communicate.Precoding is the adaptive technique that extensively adopts in the present wideband MIMO system, it can be according to the situation of the channel matrix of arrowband mimo channel, the bit number of this arrowband mimo channel transmission of the precoding adaptively modifying by transmitting terminal, can greatly improve the efficient of each scarce order or ill-condition arrowband mimo channel transmission, thereby improve the efficient of whole wideband MIMO system.Existing MIMO-OFDM method for precoding mainly is the pre-coding matrix of other available CSI formation of sum of ranks according to each (can divide into groups) arrowband mimo channel, and transmitting terminal multiply by corresponding pre-coding matrix with data vector to be sent on each arrowband MIMO subchannel and forms the transmission vector.This linear pre-coding method is fairly simple, and when selectable pre-coding matrix quantity (being called code book quantity) is big, efficiency of transmission is also than higher, shortcoming is: when code book quantity is less, the spectrum efficiency of system is relatively poor, and code book quantity is when big, though the spectrum efficiency of system can improve, but corresponding back information amount and system complexity all can increase substantially, and promptly have contradiction between the spectrum efficiency of back information amount, complexity and system.

Summary of the invention

The present invention is directed between the spectrum efficiency of back information amount, complexity and system that pre existing coded mimo-OFDM technology exists and have contradictory problems, a kind of efficient low MIMO-SCFDE system self-adaption multimode transmission method that returns is provided.

The MIMO-SCFDE system self-adaption multimode transmission method of low passback of the present invention may further comprise the steps:

(1) for N _T* N _RAdaptive MIMO systems, N _TThe expression number of transmit antennas, N _RExpression reception antenna number, receiving end is determined each subchannel order information R according to channel condition information, the modulation system that is adopted and desired system errored bit performance, R is sent to via feedback channel make a start then;

(2) suppose each subchannel order and for M, make a start according to the modulation system that is adopted, form the time domain data frame of M modulation symbol, through M point FFT, become M to tie up equivalent frequency domain data frame the time domain data frame transform, utilize each the subchannel order information R that obtains by feedback channel to generate frequency domain channel beacon information D then, with this M dimension frequency domain data frame mapping, expand into a row N * N according to channel beacon information D _TDimension frequency domain data frame again by serial/parallel conversion, generates N _TRow N dimension frequency domain data frame utilizes N point IFFT with N then _TIndividual parallel N dimension frequency domain data frame changes back to time domain, adds behind the CP simultaneously respectively by N _TIndividual transmitting antenna sends;

(3) receiving end is with N _RAfter the Frame that individual reception antenna receives removed CP, FFT transformed to frequency domain through the N point, and be N this moment _RRow N dimension frequency domain vector generates balanced matrix according to each frequency domain subchannel order information R and channel condition information and carries out equilibrium, and balanced back is N _TRow N dimension data frame then by parallel serial conversion, reverts to a row N * N _TThe dimension data frame generates channel beacon information D by each subchannel order information R again, selects M useful signal according to channel beacon information D, and conversion is carried out symbol judgement after returning time domain.

For the ease of better expression, symbol used among the present invention is carried out the part explanation.Lowercase is represented time domain and equivalent time domain symbol, and capitalization is represented frequency domain and equivalent frequency domain symbol; Represent the n moment or m equivalence constantly for the subscript of lowercase (being time domain and equivalent time domain symbol), n=0,1 ..., N-1, m=0,1 ..., M-1; The subscript of capitalization (being frequency domain and equivalent frequency domain symbol) is represented k frequency domain subchannel or q equivalent frequency domain subchannel, k=0, and 1 ..., N-1, q=0,1 ..., M-1; Capital M in the subscript represents that this vector is the symbol of equivalent time domain or equivalent frequency domain, and N represents that this vector is the symbol of time domain or frequency domain; Lowercase l, i in the subscript represent l root reception antenna and i transmit antennas respectively, if l and i occur simultaneously, l is preceding, i after, expression by i transmitting antenna to l reception antenna, i=1,2 ... N _T, l=1,2 ..., N _RMaking a start,

Expression M ties up original equivalent time domain Frame, corresponding original equivalent frequency domain data frame

M point FFT, wherein

Average power for information symbol.Generate a row N * N according to the frequency domain channel beacon information _TThe frequency domain channel mapping matrix of dimension is tieed up equivalent frequency domain data frame with M

Be mapped to N * N _TDimension frequency domain data frame

Generate N by serial/parallel conversion then _TThe frequency domain data frame of row N dimension

I=1,2 ..., N _T, carry out IFFT again and transform to time domain, respectively by N _TTransmit antennas sends, and i root antenna time domain data frame to be sent is

I=1,2 ..., N _T

It is the symbolic vector to be sent on k the frequency domain subchannel.

Be k frequency domain subchannel gains between i transmit antennas and l root reception antenna; A ^kBe the channel matrix of k frequency domain subchannel, wherein the capable i column element of l is

I=1,2 ..., N _T, l=1,2 ..., N _R, k=0,1 ..., N-1.After receiving terminal removed CP, l root reception antenna received L=1,2 ..., N _R, wherein Be the useful signal part, be N point FFT and obtain frequency domain form and be

For receiving noise vector, corresponding frequency domain form is

Wherein

Be noise variance.On k the frequency domain subchannel, signal component is

Wherein Noise component(s) is

K=0,1 ..., N-1; According to each frequency domain subchannel order information, obtain the equivalent channel matrix of a frequency domain subchannel Q wherein ^kBe the pre-coding matrix that obtains by k frequency domain subchannel order r (k)

The signal that is subjected to noise pollution that receives is carried out ZF (ZF) equilibrium, be equivalent to

M-P contrary

The difference premultiplication

With

Obtain

With

By the parallel/serial row N * N that is transformed into _TBe frequency domain symbol vectors, select useful signal according to the frequency domain channel beacon information then, becoming time domain again by M point IFFT again is x _MWith

Carry out symbol judgement.

For the adaptive M IMIO-SCFDE system that uses the ZF equilibrium, signal to noise ratio before the time domain equalization of n sampling time

As follows:

${\mathrm{SNR}}_{\mathrm{pre}}^n\stackrel{\·}{=}\frac{E\left({\left|\right|y_N^n\left|\right|}_2^2\right)}{E\left({\left|\right|w_N^n\left|\right|}_2^2\right)}=\frac{\frac{{\mathrm{\σ}}_x^2}{N}\underset{q=0}{\overset{M-1}{\mathrm{\Σ}}}\mathrm{tr}({\tilde{A}}^k\·{\left({\tilde{A}}^k\right)}^H)}{N_R{\mathrm{\σ}}_w^2}$

And can calculate signal to noise ratio after m the equivalence equivalent time domain equilibrium constantly

As follows:

${\mathrm{SNR}}_{\mathrm{post}}^m\stackrel{\·}{=}\frac{E\left[{\left|\right|x_M^m\left|\right|}_2^2\right]}{E\left[{\left|\right|v_M^m\left|\right|}_2^2\right]}=\frac{{\mathrm{\σ}}_x^2}{\frac{{\mathrm{\σ}}_w^2}{M}\underset{q=0}{\overset{M-1}{\mathrm{\Σ}}}\mathrm{tr}(G^k\·{\left(G^k\right)}^H)}$

In fact, signal to noise ratio is the signal to noise ratio of n sampling time in the balanced preceding frame before the time domain equalization, and signal to noise ratio can be regarded as m equivalent judgement signal to noise ratio constantly in the frame of balanced back after the equivalent time domain equilibrium.

The specific implementation method of above steps is as follows:

Step (1) is for N _T* N _RAdaptive MIMO systems, N _TThe expression number of transmit antennas, N _RExpression reception antenna number, receiving end is according to channel condition information, and modulation system that is adopted and desired system errored bit performance are determined each subchannel order information R, R is sent to via feedback channel make a start then.

Wherein, determine that the implementation method of each subchannel order information is as follows:

After receiving-transmitting sides was set up communication, receiving end channel estimating mode was according to a preconcerted arrangement obtained the channel condition information of current channel, and wherein channel condition information is the channel matrix of all frequency domain subchannels, the channel matrix A on each frequency domain subchannel ^k, k=0 wherein, 1 ..., N-1; According to the modulation system that adopts, under the desired errored bit performance of system, transmit information symbol as much as possible, promptly require channel capacity C big as much as possible, suppose that each subchannel order information is R={R _k, k=0,1 ..., N-1}, wherein R _k∈ 0,1 ..., N _T, represent the order of k frequency domain subchannel; Each value of initialization all is 0, i.e. R ₀=R ₁=...=R _N-1=0, calculate the channel capacity of each subchannel this moment then by mountain farming channel capacity formula, and each subchannel order all adds the channel capacity after 1, calculate the added value that each subchannel order adds 1 back channel capacity again, select order to add 1 back channel capacity and increase maximum subchannels, its order is added 1, and remaining keeps order constant, and judges whether to satisfy condition Wherein: M be each subchannel order and, M ∈ 0,1 ..., N * N _T, k ∈ 0,1 ..., N-1},

Be the average power of information symbol,

Be noise variance, SNR _ReqFor adopting certain modulation system, signal to noise ratio after the pairing equivalent time domain equilibrium of desired system errored bit performance; If satisfy, repeat this step, find out the maximum M value that satisfies condition, record each subchannel order information R=(R at this moment ₀, R ₁..., R _N-1) be each subchannel order information.

Step (2), that calculate each subchannel order by step (1) and M, make a start according to the modulation system that is adopted, form the time domain data frame of M modulation symbol, through M point FFT, become M to tie up equivalent frequency domain data frame the time domain data frame transform, utilize each the subchannel order information R that obtains by feedback channel to generate frequency domain channel beacon information D then, with this M dimension frequency domain data frame mapping, expand into a row N * N according to channel beacon information D _TDimension frequency domain data frame again by serial/parallel conversion, generates N _TRow N dimension frequency domain data frame utilizes N point IFFT with N then _TIndividual parallel N dimension frequency domain data frame changes back to time domain, adds behind the CP simultaneously respectively by N _TIndividual transmitting antenna sends.

Wherein, it is as follows to generate frequency domain channel beacon information implementation method by each subchannel order information:

Making a start obtains each subchannel order information R by feedback channel, and the frequency domain channel beacon information form that generate is D={D (k), k=0,1 ..., N-1}, wherein D (k)={ D (k _s), s=0,1 ... N _T-1}, the value of D (k) is by the order R of k subchannel _kDecision, the preceding R of D (k) _kIndividual value is 1, the back be 0; The D of Sheng Chenging is a row N * N like this _T Dimension 0,1 sequence, D (k _sChannel matrix A on k frequency domain subchannel of)=1 expression ^kS row available, D (k _sChannel matrix A on k frequency domain subchannel of)=0 expression ^kS row unavailable.

Wherein, according to the channel beacon information that the implementation method of equivalent frequency domain data frame mapping is as follows:

Suppose each subchannel order and be M, make a start and carry out sign map according to the modulation system that is adopted, form M waiting for transmission equivalent time domain symbol

FFT obtains frequency domain symbol by the M point

Generate a row N * N _TDimension frequency domain data frame

According to frequency domain channel beacon information D with D (k _s) value is that 1 position correspondence is inserted above-mentioned M dimension frequency domain symbol, D (k successively _s) be 0 position zero padding.

Wherein, the implementation method of serial/parallel conversion has multiple, for example:

Method one: a row N * N that will obtain _TDimension frequency domain data frame

The top n data symbol

As first row, N+1 to 2 * N data symbol

As secondary series, by that analogy, will

Be transformed into N _TRow N dimension frequency domain data frame

I=1,2 ..., N _T

Method two: a row N * N that will obtain _TDimension frequency domain data frame

Preceding N _TIndividual data symbol

As first row, N _T+ 1 to 2 * N _TIndividual data symbol

As second row, by that analogy, will

Be transformed into N _TRow N dimension frequency domain data frame

I=1,2 ..., N _T

Again every row Frame is carried out the IFFT conversion, change to time domain and be

I=1,2 ..., N _T, add behind the CP by N _TTransmit antennas sends simultaneously.

Step (3), receiving end is with N _RAfter the Frame that individual reception antenna receives removed CP, FFT transformed to frequency domain through the N point, and be N this moment _RRow N dimension frequency domain vector generates balanced matrix according to each frequency domain subchannel order information R and channel condition information and carries out equilibrium, and balanced back is N _TRow N dimension data frame then by parallel serial conversion, reverts to a row N * N _TThe dimension data frame generates channel beacon information D by each subchannel order information R again, selects M useful signal according to channel beacon information D, and conversion is carried out symbol judgement after returning time domain.

Wherein, it is as follows to generate balanced matrix and balanced process (is example with the ZF equilibrium) implementation method:

After receiving terminal removed CP, l root reception antenna received useful signal and partly is

Noise section is

Being N point FFT obtains frequency domain form and is respectively

With Signal component is on k the frequency domain subchannel

Wherein Noise component(s) is

K=0,1 ..., N-1; According to each subchannel order information of each frequency domain, obtain the equivalent channel matrix of each frequency domain subchannel

Q wherein ^kBe the pre-coding matrix that obtains by k frequency domain subchannel order r (k)

The signal that is subjected to noise pollution that receives is carried out the ZF equilibrium, use

M-P contrary

The difference premultiplication

With Obtain the useful signal part

And noise section $V_N^k=(V_{N,1}^k,V_{N,2}^k,...,V_{N,N_T}^k)=G^k\·W_N^k.$

Wherein, generated in channel beacon information implementation method and the step (2) identically by each subchannel order information, parallel/serial conversion process according to the serial/parallel method of selecting for use in the step (2), is implemented corresponding inverse transformation, obtaining N after the equilibrium _TRow N dimension data symbol is transformed into a row N * N _TThe dimension frequency domain symbol.

Wherein, by inverse mapping, the implementation method of selecting useful signal is as follows:

Obtain a row N * N _TBehind the dimension frequency domain symbol,, select corresponding beacon information D (k according to frequency domain channel beacon information D _s) be 1 frequency domain symbol, abandon beacon information and be the frequency domain symbol of 0 carrier frequency point, obtain a row M dimension useful signal, become time domain again by M point IFFT again, carry out symbol judgement.

The contradiction between system spectral efficiency and back information amount and the system complexity has been alleviated in the present invention, has system spectral efficiency height, characteristics that the back information amount is low.

Description of drawings

Fig. 1 is a MIMO-SCFDE system block diagram of realizing self adaptation multimode transmission method of the present invention.

Fig. 2 is the errored bit curve chart that adopts the MIMO-SCFDE system of the self adaptation multimode transmission method that the present invention proposes.

Among the figure: 1, information source module, 2, the sign map module, 3, FFT module (M point), 4, the signal spectrum conversion, 5, serial/parallel module, 6, IFFT module (N point), 7, add Cyclic Prefix (CP) module, 8, the D/A module, 9, intermediate frequency and rf modulations module, 10, mimo channel, 11, radio frequency and intermediate frequency demodulation module, 12, the A/D module, 13, go the CP module, 14, FFT module (N point), 15, balance module, 16 parallel/serial modules, 17, the signal spectrum inverse transform block, 18, IFFT module (M point), 19, judgement and symbol inverse mapping module, 20, synchronization module, 21, channel estimation module, 22, each frequency domain subchannel order information module, 23, frequency domain channel beacon information module, 24, feedback channel.

Embodiment

What embodiment provided is the base band simulation result, does not consider the influence of synchronous error, and does not consider channel estimation errors, and promptly channel estimating is desirable.

Fig. 1 has provided the MIMO-SCFDE system block diagram of realizing self adaptation multimode transmission method of the present invention, and each module effect is as follows:

Information source module 1: the data that generation will be transmitted.

Sign map module 2: the data based modulation system that adopts that information source is produced is mapped on the planisphere corresponding points.

M point FFT module 3: the individual mapping signal of every frame M is transformed to frequency domain, obtain the M point frequency-region signal of signal.

Signal spectrum conversion module 4: send the channel beacon information of returning according to feedback channel, the M point frequency-region signal of module 3 outputs is mapped on M the available frequency domain bearing point, and forbids the zero setting of frequency domain bearing point, or fill non-information data, just obtain a frame N * N _TThe dimension frequency domain vector.This module need be programmed according to the method that invention is introduced, and is realized by the general digital signal processing chip.

Serial/parallel module 5: N * N that module 4 is obtained _TThe dimension frequency domain vector is transformed into N _TIndividual N dimension frequency domain vector.

N point IFFT module 6: will obtain frequency-region signal and transform to time domain again.

Add CP module 7: every frame data that will obtain add Cyclic Prefix.

D/A module 8: digital signal conversion is become analog signal.

Intermediate frequency and rf modulations module 9: signal is modulated to and carries out the intermediate frequency amplification on the intermediate frequency, does rf modulations again, at last modulated signal is launched by antenna.

Mimo channel 10: the frequency domain selectivity mimo channel of transmission signals.

Radio frequency and intermediate frequency demodulation module 11: the frequency spectrum that reception antenna is received signal is moved low frequency from radio frequency or intermediate frequency.Before demodulation, need the frequency deviation that causes with in the Frequency Synchronization data correction signal transmission course.

A/D module 12: analog signal after the demodulation is transformed to digital signal.The A/D conversion need be sampled to analog signal, provides the crystal oscillator of clock signal need follow the crystal oscillator frequency of transmitter D/A module identical, otherwise will cause the sampling rate error.Therefore it is synchronous to carry out sampling rate before the A/D conversion.

Go CP module 13: Cyclic Prefix is removed.At this moment just have the problem of judging which frame data begin from, then going needs to do regularly synchronously before the CP.

N point FFT module 14: the time-domain signal that will remove CP transforms to frequency domain.

Balance module 15: each subchannel order information that CSI that sends according to channel estimation module 21 and module 22 are sent generates balanced matrix and carries out equilibrium.Balanced way can be selected one of three kinds of balanced ways: zero forcing equalization, least mean-square error equilibrium, hybrid mode equilibrium.

Parallel/serial module 16: will cross the N that obtains behind the balance module 15 _TIndividual N dimensional vector is transformed to N * N _TDimensional vector.

Signal spectrum inverse transform block 17: select the equivalent frequency-region signal that carries on M the available frequency domain bearing point according to the channel beacon information.This module need be programmed according to the method that the present invention introduces, and is realized by the general digital signal processing chip.

M point IFFT module 18: the equivalent frequency domain symbol after the equilibrium is transformed to equivalent time domain.

Judgement and symbol inverse mapping module 19:, finish the judgement of time-domain signal according to the modulation system that system adopted.

Synchronization module 20: the way by parameter Estimation (for example: blind estimation and based on the estimation of auxiliary data) obtains the various synchrodatas that system needs.Synchronization module is given radio frequency and intermediate frequency demodulation module 11 with the Frequency Synchronization data; Give A/D module 12 with the sampling rate synchrodata; Regularly synchrodata is given CP module 13.At present embodiment, be assumed to be synchronous ideal.

Channel estimation module 21: with syncsort seemingly, also need to obtain CSI blind Channel Estimation that commonly used generally is and based on the channel estimating of auxiliary data by parameter Estimation.At present embodiment, suppose to estimate accurately.

Each frequency domain subchannel order information module 22: the CSI according to channel estimation module 21 obtains, according to the desired errored bit performance of system, determine each subchannel order information.

Frequency domain channel beacon information module 23: each the frequency domain subchannel order information according to module 22 generates generates the channel beacon information.

Backward channel 24: the channel beacon information is returned to transmitting terminal.

This embodiment simulation parameter:

Simulated environment: MATLAB R2010a

Subchannel sum: N=256

CP length: 64

Modulation system: QPSK

Sampling rate: 10M bps

The selected average received signal to noise ratio of emulation scope: SNR=4～15 (dB)

Simulated channel environment: adopt 4 * 4SUI-4 channel respectively, but the SUI-4 channel in the present embodiment is not considered Doppler frequency deviation and Antenna Correlation.(can be with reference to D.S.Baum, " Simulating the SUI channel models, " IEEE 802.16Broad Wireless Access Working Group, 2001, (D.S.Baum, " emulation SUI channel model ", IEEE 802.16 broadband wireless access working groups, 2001))

Do not consider in the emulation that to the influence of system, the error of promptly supposing all synchronization parameters all is 0 to synchronous error (comprising regularly synchronous error of carrier synchronization error, sampling rate synchronous error and frame); Do not consider the propagation delay time of backward channel return path identification information and the influence of transmission error code, suppose that promptly propagation delay time and error code all are 0; Do not consider the influence of other non-ideal factors

(for example device non-linear etc.).

Simulation result:

Fig. 2 has provided the errored bit curve of the MIMO-SCFDE system of the self adaptation multimode transmission method that adopts the present invention's proposition, comparison with the errored bit energy of a kind of multi-mode precoding MIMO-ofdm system that limits passback (64 of wherein equally distributed pilot tones, the code book space that quantizes pre-coding matrix contains 16 pre-coding matrixes) and awgn channel based on interpolation.It should be noted that the SNR that is labeled as square dotted line abscissa no longer represents received signal to noise ratio but the balanced back of equivalence signal to noise ratio, this is to compare with the BER of awgn channel for convenience.A kind of multi-mode precoding MIMO based on interpolation-ofdm system that limits passback adopts 64 equally distributed pilot tones and contains 16 pre-coding matrixes and quantizes the pre-coding matrix code book, and also need the precoding model selection of 2 bits for each carrier wave of system of 4 transmit antennas, so the back information amount is 64 * 4+256 * 2=768 bit.When this system and the MIMO-SCFDE system of self adaptation multimode transmission method have identical average transmit power and identical bit rate, performance is better than self adaptation multimode MIMO-SCFDE under low signal-to-noise ratio, but performance is worse than self adaptation multimode MIMO-SCFDE under the high s/n ratio, and each the subchannel passback of self adaptation multimode MIMO-SCFDE system only needs log ₂(N _T+ 1) bit is got greater than log ₂(4+1) nearest integer is 3, therefore only needs the back information amount of (256-64) * 3=576 bit altogether, and the back information amount is less than the multi-mode precoding MIMO-ofdm system based on interpolation.(can be with reference to N.K.Khaled, B.Mondal, G.Leus, R W.Heath, and F.Petre, " Interpolation-Based Multi-Mode Precoding for MIMO-OFDM system with Limited Feedback, " IEEE Trans.on Wireless Communications, vol.6, no.3, pp.1003-1013, Mar.2007 (N.K.Khaled, B.Mondal, G.Leus, R W.Heath, and F.Petre, " based on the MIMO-OFDM multi-mode precoding of the limited passback of interpolation ", IEEE Trans.on Wireless Communications, the 6th volume, No. three, the 1003-1013 page or leaf, in March, 2007).

Following table has provided the comparison that 4 * 4MIMO-SCFDE system adopts the spectrum efficiency of the MIMO-SCFDE system of the self adaptation multimode transmission method that the present invention proposes and self adaptation single mode MIMO-SCFDE system.Wherein this spectrum efficiency is SNR _Req=SNR _PreThe mean value of the spectrum efficiency that fixedly 1000 secondary channel samples obtain under the BER performance of average transmit power and identical system requirements during=10dB.The spectrum efficiency that can find the self adaptation multimode MIMO-SCFDE system that the present invention proposes is significantly improved, and system complexity does not have much increases.Therefore what the present invention proposed self adaptation multimode MIMO-SCFDE system can obviously improve the availability of frequency spectrum.

The comparison of the spectrum efficiency of 4 * 4MIMO-SCFDE system self-adaption multimode and self adaptation single mode

For avoiding confusion, some nouns or the symbol mentioned in this specification are done following explanation:

1, the frequency domain subchannel notion of frequency domain subchannel: MIMO-SCFDE is equal to the notion of the subcarrier among the MIMO-OFDM.

2, equivalent time domain symbol: before the M point FFT and the symbol behind the M point IFFT.

3, equivalent frequency domain symbol: reach the preceding symbol of M point IFFT behind the M point FFT.

4, time-domain signal: before the N point FFT and the symbol behind the N point IFFT.

5, frequency-region signal: reach the preceding symbol of N point IFFT behind the N point FFT.

6,

The M-P of matrix is contrary.

Claims (8)

1. the MIMO-SCFDE system self-adaption multimode transmission method of a low passback is characterized in that, may further comprise the steps:

(1) for N _T* N _RAdaptive MIMO systems, N _TThe expression number of transmit antennas, N _RExpression reception antenna number, receiving end is determined each subchannel order information R according to channel condition information, the modulation system that is adopted and desired system errored bit performance, R is sent to via feedback channel make a start then;

(2) suppose each subchannel order and for M, make a start according to the modulation system that is adopted, form the time domain data frame of M modulation symbol, through M point FFT, become M to tie up equivalent frequency domain data frame the time domain data frame transform, utilize each the subchannel order information R that obtains by feedback channel to generate frequency domain channel beacon information D then, with this M dimension frequency domain data frame mapping, expand into a row N * N according to channel beacon information D _TDimension frequency domain data frame again by serial/parallel conversion, generates N _TRow N dimension frequency domain data frame utilizes N point IFFT with N then _TIndividual parallel N dimension frequency domain data frame changes back to time domain, adds behind the CP simultaneously respectively by N _TIndividual transmitting antenna sends;

(3) receiving end is with N _RAfter the Frame that individual reception antenna receives removed CP, FFT transformed to frequency domain through the N point, and be N this moment _RRow N dimension frequency domain vector generates balanced matrix according to each frequency domain subchannel order information R and channel condition information and carries out equilibrium, and balanced back is N _TRow N dimension data frame then by parallel serial conversion, reverts to a row N * N _TThe dimension data frame generates channel beacon information D by each subchannel order information R again, selects M useful signal according to channel beacon information D, and conversion is carried out symbol judgement after returning time domain.

2. the MIMO-SCFDE system self-adaption multimode transmission method of low passback according to claim 1 is characterized in that, determines in the described step (1) that the implementation method of each subchannel order information is as follows:

After receiving-transmitting sides was set up communication, receiving end channel estimating mode was according to a preconcerted arrangement obtained the channel condition information of current channel, and wherein channel condition information is the channel matrix of all frequency domain subchannels, the channel matrix A on each frequency domain subchannel ^k, k=0 wherein, 1 ..., N-1; According to the modulation system that adopts, under the desired errored bit performance of system, transmit information symbol as much as possible, promptly require channel capacity C big as much as possible, suppose that each subchannel order information is R={R _k, k=0,1 ..., N-1}, wherein R _k∈ 0,1 ..., N _T, represent the order of k frequency domain subchannel; Each value of initialization all is 0, i.e. R ₀=R ₁=...=R _N-1=0, calculate the channel capacity of each subchannel this moment then by mountain farming channel capacity formula, and each subchannel order all adds the channel capacity after 1, calculate the added value that each subchannel order adds 1 back channel capacity again, select order to add 1 back channel capacity and increase maximum subchannels, its order is added 1, and remaining keeps order constant, and judges whether to satisfy condition

Wherein: M be each subchannel order and, M ∈ 0,1 ..., N * N _T, k ∈ 0,1 ..., N-1},

Be the average power of information symbol,

Be noise variance, SNR _ReqFor adopting certain modulation system, signal to noise ratio after the pairing equivalent time domain equilibrium of desired system errored bit performance; If satisfy, repeat this step, find out the maximum M value that satisfies condition, record each subchannel order information R=(R at this moment ₀, R ₁..., R _N-1) be each subchannel order information.

3. the MIMO-SCFDE system self-adaption multimode transmission method of low passback according to claim 1 is characterized in that, the implementation method by each subchannel order information generation frequency domain channel beacon information in the described step (2) is as follows:

Making a start obtains each subchannel order information R by feedback channel, and the frequency domain channel beacon information form of generation is D={D (k), k=0,1 ..., N-1}, wherein D (k)={ D (k _s), s=0,1 ... N _T-1}, the value of D (k) is by the order R of k subchannel _kDecision, the preceding R of D (k) _kIndividual value is 1, the back be 0; The D of Sheng Chenging is a row N * N like this _TDimension 0,1 sequence, D (k _sChannel matrix A on k frequency domain subchannel of)=1 expression ^kS row available, D (k _sChannel matrix A on k frequency domain subchannel of)=0 expression ^kS row unavailable.

4. the MIMO-SCFDE system self-adaption multimode transmission method of low passback according to claim 1 is characterized in that, and is according to the channel beacon information that the implementation method of equivalent frequency domain data frame mapping is as follows in the described step (2):

Suppose each subchannel order and be M, make a start and carry out sign map according to the modulation system that is adopted, form M waiting for transmission equivalent time domain symbol

FFT obtains frequency domain symbol by the M point Generate a row N * N _TDimension frequency domain data frame

According to frequency domain channel beacon information D with D (k _s) value is that 1 position correspondence is inserted above-mentioned M dimension frequency domain symbol, D (k successively _s) be 0 position zero padding.

5. the MIMO-SCFDE system self-adaption multimode transmission method of low passback according to claim 1 is characterized in that the implementation method of serial/parallel conversion is in the described step (2):

With a row N * N who obtains _TDimension frequency domain data frame The top n data symbol

As first row, N+1 to 2 * N data symbol

As secondary series, by that analogy, will

Be transformed into N _TRow N dimension frequency domain data frame

I=1,2 ..., N _T

6. the MIMO-SCFDE system self-adaption multimode transmission method of low passback according to claim 1 is characterized in that the implementation method of serial/parallel conversion is in the described step (2):

With a row N * N who obtains _TDimension frequency domain data frame

Preceding N _TIndividual data symbol

As first row, N _T+ 1 to 2 * N _TIndividual data symbol

As second row, by that analogy, will

Be transformed into N _TRow N dimension frequency domain data frame

I=1,2 ..., N _T

7. the MIMO-SCFDE system self-adaption multimode transmission method of low passback according to claim 1 is characterized in that, when adopting ZF balanced, the procedural implementation method that generates balanced matrix and equilibrium is as follows in the described step (3):

After receiving terminal removed CP, l root reception antenna received useful signal and partly is

Noise section is

Being N point FFT obtains frequency domain form and is respectively

With

Signal component is on k the frequency domain subchannel

Wherein Noise component(s) is

K=0,1 ..., N-1; According to each subchannel order information of each frequency domain, obtain the equivalent channel matrix of each frequency domain subchannel Q wherein ^kBe the pre-coding matrix that obtains by k frequency domain subchannel order r (k)

The signal that is subjected to noise pollution that receives is carried out the ZF equilibrium, use

M-P contrary

The difference premultiplication With

Obtain the useful signal part

And noise section $V_N^k=(V_{N,1}^k,V_{N,2}^k,...,V_{N,N_T}^k)=G^k\·W_N^k.$

8. the MIMO-SCFDE system self-adaption multimode transmission method of low passback according to claim 1 is characterized in that, it is as follows to select the implementation method of useful signal by inverse mapping in the described step (3):

Obtain a row N * N _TBehind the dimension frequency domain symbol,, select corresponding beacon information D (k according to frequency domain channel beacon information D _s) be 1 frequency domain symbol, abandon beacon information and be the frequency domain symbol of 0 carrier frequency point, obtain a row M dimension useful signal, become time domain again by M point IFFT again, carry out symbol judgement.

Images