CN102710567A - Part judgment method in interference elimination technology for multiple-input multiple-output (MIMO) wireless communication receiver - Google Patents

Abstract

The invention provides a part judgment method in an interference elimination technology for a multiple-input multiple-output (MIMO) wireless communication receiver. The part judgment method comprises the following steps of (1) setting a threshold, performing part judgment on a judgment statistical magnitude, and thus obtaining a part judgment signal; and (2) reconstructing the part judgment signal obtained in the step (1), and eliminating interference according to the reconstructed signal, wherein in the process of reconstructing the part judgment signal, the signal is reconstructed according to the part judgment signal by adopting the conventional hard judgment reconstruction method; and in the process of eliminating the interference of the reconstructed signal, the interference is eliminated according to the signal reconstructed from the part judgment signal by adopting the conventional hard judgment interference elimination method. The MIMO interference elimination is improved; slight complexity is increased; the advantages of simplicity and feasibility of the conventional MIMO interference elimination method are kept; and under the condition that slight complexity and slight calculation amount are increased, the performance of the MIMO wireless communication receiver is also improved.

Description

Part decision method in the mimo wireless communication receiver interference cancellation techniques

Technical field

The present invention relates to a kind of multi-antenna wireless communication transmission method, belong to wireless communication technology 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; Wireless communication technology 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.Spectrum efficiency is the research emphasis of wireless communication technology always; Multiple-input and multiple-output (Multiple-Input Multiple-Output is hereinafter to be referred as the MIMO) technology that adopts multi-antenna technology based on transmitting-receiving two-end in recent years with its conventional single-antenna technology be beyond one's reach spectrum efficiency and received extensive concern.MIMO and based on Cyclic Prefix (Cyclic Prefix; CP) OFDM (the Orthogonal Frequency DivisionMultiplexing of technology; OFDM) and single carrier frequency domain equalization (Single Carrier with Frequency DomainEqualization, SC-FDE) MIMO-OFDM that combine to occur of technology is technological with the main support that MIMO-SCFDE becomes the future wireless physical layer transmission.

MIMO utilizes in the wireless propagation environment of rich multipath the uncorrelated characteristic of channel gain between the different antennae, obtains the high channel capacity, thereby improves the availability of frequency spectrum and the reliability of whole system.

In the MIMO receiver, the receiver that adopts the linear equalization mode is to remove multiply by received signal vector with a balanced matrix to accomplish equilibrium to received signal.Linear equalization mode commonly used has two kinds, promptly compels zero (Zero Forcing, ZF) balanced and least mean-square error (Minimum Mean Square Error, MMSE) equilibrium.

QAM (Quadrature Amplitude Modulation, quadrature amplitude modulation) modulation system is used widely in a lot of standards, like IEEE802.11,3GPP-LTE, WLAN etc.Mimo system comprises MIMO narrowband systems, MIMO-OFDM system and MIMO-SCFDE system, all extensively adopts qam mode to realize modulation and demodulation.In the QAM demodulating process; Adopt the implementation method of hard decision to be: in-phase component and the quadrature component of treating decision signal are adjudicated respectively; Generally, according to the planisphere bit mapping mode information bit is transformed into decision signal again through repeatedly in-phase component or quadrature component and decision threshold relatively being obtained transmission information bit.Be demodulated into example with 16QAM, establish and treat that decision signal (being also referred to as the decision statistic amount) is a+bj, wherein

Obtain 4 bit b after the judgement ₁b ₂b ₃b ₄, wherein in-phase component a adjudicates the corresponding b in back ₁b ₂, the quadrature component b judgement corresponding b in back ₃b ₄Planisphere is shown in accompanying drawing 2.In-phase component is carried out twice judgement altogether:

Decision threshold is D for the first time ₁, D ₁=0, decision method is:

Decision threshold is D for the second time ₂, D ₂=2, decision method is:

Quadrature component is identical with the in-phase component decision method, obtains b after the judgement ₃b ₄For example, the planisphere bit mapping mode does shown in the accompanying drawing 2, when information bit is 0101, it is transformed to decision signal-1-j.

Disturb inhibition (Successive Inference Cancelation based on order; SIC) receiver; Owing to adopted good interference mitigation technology, make that the interference between different layers alleviates greatly, performance is general remarkable only to be superior to the decorrelation receiver based on linear equalization.Typical case's representative based on the MIMO receiver of SIC is BLAST (the Bell Laboratories Layered Space-Time Architecture) receiver that Bell laboratory G Foschini proposes.

Testing process based on the V-BLAST detection algorithm of pseudoinverse progressively is following: before the detection signal; Treat each layer ordering of detection signal according to certain criterion; Principle of ordering be by signal to noise ratio (Signal to Noise Ratio, SNR) or Signal to Interference plus Noise Ratio (Signal toInterference andNoise Ratio, the ordering of SINR) size; Detect the big layer of SNR (or SINR) earlier; Every detection one layer signal, other each layer signal are all handled as disturbing, and remove the current layer signal again to the not interference of detection signal layer.The signal that the signal that detects earlier detects the back is influential, so can cause certain error code diffusion.Sort through certain criterion and can reduce error code diffusion.Removal current layer signal can adopt ZF equilibrium or MMSE equilibrium to the mode of the not interference of detection signal layer.

Introduce below based on the implementation procedure of the V-BLAST detection algorithm of pseudoinverse progressively, in this process, remove the current layer signal and adopt ZF balanced the mode of the not interference of detection signal layer:

H ₁＝H

fori＝1,…,N _T

(1)

Wherein $G_i={[g_i^{\left(1\right)},g_i^{\left(2\right)},\·\·\·,g_i^{\left(N_T\right)}]}^T$

（2） $k_i=\arg \underset{j}{\min }{\left|\right|g_i^{\left(j\right)}\left|\right|}^2$

（3） $y_{k_i}=g_i^{\left(k_i\right)}\·x_i$

（4） ${\hat{c}}_{k_i}=D\left[y_{k_i}\right]$

(5) $x_{i+1}=x_i-h_i^{\left(k_i\right)}\·{\hat{c}}_{k_i},$ Wherein $H_i=[h_i^{\left(1\right)},h_i^{\left(2\right)},\·\·\·,h_i^{\left(N_T\right)}]$

(6) with H _iK _iRow zero setting obtains upgrading matrix H _I+1

end

Wherein, i representes iterations, () _iRepresent the matrix that the i time iteration obtains, () ^(j)Capable or the j row of the j of representing matrix, () ^TThe transposition of representing matrix,

The pseudoinverse of representing matrix,

Represent k _iLayer sends the estimated value (just by the immediate constellation point signal of Euclidean distance) of signal, and D [] representes by the conversion process of decision statistic amount to decision signal.Step (2) is selected layer to be detected according to certain criterion; Step (3) (4) has realized treating the process of detection layers symbol detection, and wherein step (4) is a quantizing process, is used to eliminate noise jamming; Step (5) has realized eliminating current layer and has received the process of detection layers symbol-interference, and wherein

realizes the reconstruct of detection layers symbol.

In the MIMO-OFDM system detection algorithm,, can the V-BLAST detection algorithm of arrowband mimo system be applied in the input of MIMO-OFDM system by subchannel because each subchannel of MIMO-OFDM system can be regarded the arrowband mimo channel as.In the MIMO-SCFDE system detection algorithm, to the principle that the layering of MIMO signal detects, the piece V-BLAST detection algorithm of MIMO-SCFDE system has been proposed at present based on V-BLAST; Method is following: at the MIMO receiving terminal; Criterion of choosing by certain is successively carried out frequency domain equalization and detection to sending signal, after the detection of every completion to a layer signal, transforms the signal into frequency domain; Be multiplied by corresponding channel matrix again; Again it is eliminated from signal to be detected as disturbing, detect next layer signal according to identical method then, all detect end until all layer signals.

The piece V-BLAST detection algorithm of the V-BLAST detection algorithm of arrowband mimo system, the BLAST detection algorithm of MIMO-OFDM system and MIMO-SCFDE system all is that detected signals layer is carried out hard decision; The signal that hard decision is obtained carries out reconstruct and obtains reconstruction signal, further utilizes reconstruction signal to carry out interference eliminated.

The BLAST receiver that Bell laboratory G Foschini proposes is though its V-BLAST receives the extensive concern of academia, because complexity is too high and to the sensitiveness of channel measurement error, still do not accepted extensively by industrial quarters so far.

Chinese patent document CN102006250A discloses a kind of " the Turbo Enhancement Method of MIMO-SCFDE wireless communication receiver "; This method makes this decorrelation receiver can keep the advantage of easy realization simple in structure, and MIMO-SCFDE wireless communication receiver performance is got a promotion.

In the MIMO communication system, claim that traditionally the signal of transmit antennas emission is one deck, each layer signal has N symbol, can use a N * 1 dimension matrix notation; Transmitting of the corresponding different layers of different transmit antennas, the signal that the i transmit antennas sends is called the i layer.

The Turbo Enhancement Method step of MIMO-SCFDE wireless communication receiver is following:

(1) frequency domain baseband signal R before the buffer memory equilibrium; Take out buffered signal and it is carried out linear equalization; Become the signal after the equilibrium again time domain, this time-domain signal is called the decision statistic amount, and the decision statistic amount is carried out hard decision; Obtain the information bit of each layer, and further obtain the frequency domain estimated signal of each equivalent layer symbol

Wherein

Be the frequency domain estimated value of i layer, i=1,2 ..., N _T,

Be the frequency domain estimated value of k frequency domain subchannel, k=0,1 ..., N-1.

(2) the frequency domain estimated value of each layer symbol is carried out Turbo and strengthen, the frequency domain estimated value of each layer symbol is once strengthened to be called take turns Turbo and strengthen,, carry out taking turns the Turbo enhancing at least according to requirement to receiver performance and complexity aspect.Wherein, it is following the frequency domain estimated value of each layer symbol to be carried out taking turns the concrete grammar that Turbo strengthens:

Be (1 ..., N _T) any arrangement, N _TThe expression number of transmit antennas; The frequency domain estimated value that obtains from step (1)

K=0,1 ..., take out k among the N-1 ₁The frequency domain estimated value of other each layer symbols beyond the layer is used for the k that the reconstruct receiver receives ₁The frequency-region signal that other each layers beyond the layer transmit,

K=0,1 ..., N-1; I ∈ 1,2 ..., N _TBe the reconstruct of frequency-region signal that the i layer that receiver receives is transmitted, () ^TThe transposition of representing matrix or vector; K=0,1 ..., N-1; I ∈ 1,2 ..., N _TBe the N except that the i layer that receiver is received _TThe reconstruct of-1 layer of frequency-region signal that transmits; Take out frequency domain baseband signal R before buffer memory balanced then, with the signal of buffer memory deduct that receiver receives except that K ₁Other N beyond the layer _T-1 layer of reconstructed frequency domain signals that transmits, promptly

K=0,1 ..., N-1; With the baseband signal that obtains

Premultiplication

Obtain

K=0,1 ..., N-1; K ₁∈ 1,2 ..., N _T; Then with k ₁Layer baseband signal transforms to time domain through N point IFFT, obtains the decision statistic amount after the interference eliminated, again this decision statistic amount carried out hard decision, obtains k ₁The output information bit vectors of layer Will

Again carry out changing back to frequency domain after the sign map by transmitting terminal sign map mode, upgrade former frequency domain estimated value with current frequency domain estimated value

In

Method with identical is handled k ₂Layer baseband signal, until

When layer baseband signal, frequency-region signal that other each layers beyond the current layer that each reconstruct receiver receives transmit, use the frequency domain estimated value of latest update

Upgrade.

Wherein, H ^kBe the channel matrix of k frequency domain subchannel, be expressed as

$H^k=\left[\begin{array}{ccc}{\mathrm{<figure-callout\; id="1"\; label="H"\; filenames="HDA00001615252000012.png,HDA00001615252000021.png"\; state="\{\{state\}\}">H</figure-callout>}}_{1,1}^{\mathrm{<figure-callout\; id="1"\; label="k\; H"\; filenames="HDA00001615252000012.png,HDA00001615252000021.png"\; state="\{\{state\}\}">k</figure-callout>}}& H_{}^{}{}_{\mathrm{1,2}}^{\mathrm{<figure-callout\; id="1"\; label="k\; ...\; H"\; filenames="HDA00001615252000012.png,HDA00001615252000021.png"\; state="\{\{state\}\}">k</figure-callout>}}& ...& H_{}^{}{}_{1,N_{\mathrm{<figure-callout\; id="2"\; label="T\; k\; H"\; filenames="HDA00001615252000021.png"\; state="\{\{state\}\}">T</figure-callout>}}}^k& \\ H_{}^{}{}_{\mathrm{2,1}}^{\mathrm{<figure-callout\; id="2"\; label="k\; H"\; filenames="HDA00001615252000021.png"\; state="\{\{state\}\}">k</figure-callout>}}& H_{}^{}{}_{\mathrm{2,2}}^{\mathrm{<figure-callout\; id="2"\; label="k\; ...\; H"\; filenames="HDA00001615252000021.png"\; state="\{\{state\}\}">k</figure-callout>}}& ...& H_{}^{}{}_{2,N_T}^k\\ ...& ...& ...& ...\\ H_{N_R,1}^k& H_{N_R,2}^k& ...& H_{N_R,N_T}^k\end{array}\right]$

Wherein, k ∈ 0,1 ..., N-1}.K frequency domain subchannel complex gain of channel between

j transmitting antenna of expression and i reception antenna.

No matter the Turbo Enhancement Method still is the BLAST method; When detected signals layer carries out interference eliminated to signals layer to be detected; All be to use the hard decision result of detected signals layer to carry out signal reconstruction; The error code of this hard decision can influence the quality of reconstruction signal, thereby influences the effect of interference eliminated.

Summary of the invention

The present invention is directed to the error code diffusion problem that the hard decision that exists in the mimo wireless communication interference elimination method causes; Part decision method in a kind of mimo wireless communication receiver interference cancellation techniques has been proposed; This method is on the former hard decision interference elimination method of MIMO, to improve; Under the situation that increases less complexity and amount of calculation, the performance of mimo wireless communication receiver is further got a promotion.

Part decision method in the mimo wireless communication receiver interference cancellation techniques of the present invention may further comprise the steps:

(1) thresholding is set, the decision statistic amount is carried out the part judgement, obtain the part decision signal; Detailed process is:

Thresholding ε is set, and the thresholding span is 0＜ε＜1; If the decision statistic scale is shown r, decision statistic amount r is carried out the part judgement, obtain the part decision signal

In the part judging process, need carry out the part judgement respectively to in-phase component and the quadrature component of decision statistic amount r, the process that the in-phase component of decision statistic amount r is carried out the part judgement is following: establishing k decision threshold of decision statistic amount r in-phase component is D _k, k=1 ..., K, K are the sum of decision threshold, if the in-phase component of decision statistic amount r drops on (D _k-ε, D _k+ ε) in the zone, do not adjudicate, keep initial value; If the in-phase component of decision statistic amount drops on (D _k-ε, D _k+ ε) outside the zone, then by the nearest principle of Euclidean distance, adjudicating is the coordinate figure of respective constellation point in-phase component correspondence, the part judging process of the quadrature component of decision statistic amount r is identical with the part judging process of in-phase component; After accomplishing the judgement of above-mentioned in-phase component and quadrature component, promptly obtain the part decision signal

(2) the part decision signal that step (1) is obtained carries out signal reconstruction, utilizes reconstruction signal to carry out interference eliminated; Wherein, the part decision signal carries out the process of signal reconstruction, is to utilize the part decision signal to adopt former hard decision reconstructing method to carry out signal reconstruction; Reconstruction signal carries out the process of interference eliminated, is that the signal that utilizes the reconstruct of part decision signal to obtain adopts the method for former hard decision interference eliminated to carry out interference eliminated.

The present invention is directed to the error code diffusion problem that hard decision causes; A kind of part decision method has been proposed; This method is on the former hard decision interference elimination method of MIMO, to improve; The advantage that has kept the MIMO interference elimination method to be simple and easy to realize under the situation that increases less complexity and amount of calculation, has been further upgraded the performance of mimo wireless communication receiver.

Description of drawings

Accompanying drawing 1 is that modulation system is the planisphere of 4QAM.

Accompanying drawing 2 is that modulation system is the planisphere of 16QAM.

Accompanying drawing 3 is system block diagrams of realizing the part decision method in the MIMO-SCFDE wireless communication receiver interference cancellation techniques of the present invention.

Accompanying drawing 4 is the errored bit curve comparison diagrams when adopting IMT2000 (Vehicular A) channel MMSE balanced of part decision method and the former hard decision method in the MIMO-SCFDE wireless communication receiver interference cancellation techniques.

Accompanying drawing 5 is the errored bit curve comparison diagrams when adopting COST259 (UTx) channel MMSE balanced of part decision method and the former hard decision method in the MIMO-SCFDE wireless communication receiver interference cancellation techniques.

Wherein: 1, MIMO-SCFDE transmitting terminal processing module, 2, radio frequency, intermediate frequency demodulation and baseband processing module, 3, go the CP module; 4, FFT module (N point); 5, linear equalization module, 6, IFFT module (N point), 7, the part judging module; 8, utilize the part court verdict to carry out the Turbo enforcement module, 9, output module.

Embodiment

What embodiment provided is that the part decision method that utilizes the present invention to propose in the MIMO-SCFDE wireless communication receiver combines simulation result with the Turbo Enhancement Method.

This embodiment simulation parameter:

Simulated environment: MATLAB R2010a,

Subchannel sum: N=1024,

CP length: 256,

Number of transmit antennas: 4,

The reception antenna number: 4,

Sign map mode: 16QAM,

Sampling rate: 20M bps,

Threshold setting: ε=0.2,

Turbo strengthens the wheel number: 4 take turns,

The average received signal to noise ratio scope of emulation: SNR=17～23 (dB).

The system that realizes the part decision method in the MIMO-SCFDE wireless communication receiver interference cancellation techniques of the present invention is shown in accompanying drawing 3, and the effect of each module is described below:

1.MIMO-SCFDE transmitting terminal processing module: accomplish the Base-Band Processing of MIMO-SCFDE signal, after up-conversion, launch.

2. radio frequency, intermediate frequency demodulation and baseband processing module: the signal that receives is carried out down-converted.

3. go the CP module: the deletion Cyclic Prefix.

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

5. linear equalization module: frequency-region signal is adopted modes such as ZF equilibrium or MMSE equilibrium carry out linear equalization.

6.IFFT module (N point): the frequency-region signal after the equilibrium is transformed to time domain.

7. part judging module: the method according to the present invention describes is carried out the part judgement to time-domain signal.

8.Turbo enforcement module: the signal after the part judgement is carried out Turbo strengthen.

9. output module: output signal.

The concrete implementation method of part judgement is following among the embodiment:

With 16QAM is example, and thresholding is divided shown in accompanying drawing 2.If treating decision signal is a+bj; Wherein a is an in-phase component, and b is a quadrature component.If decision threshold is D _k, k=0,1,2.D ₀＝0，D ₁＝2，D ₂＝-2。In-phase component part decision method is:

0＜ε＜1 wherein.Quadrature component is identical with the part decision method of in-phase component.

The thresholding division methods that the present invention proposes extends in the qam modes such as 4QAM, 64QAM.

Simulation result:

Accompanying drawing 4 has provided in the MIMO-SCFDE wireless communication system; The errored bit curve that the part decision method that adopts the present invention to propose combines with the Turbo Enhancement Method, IMT2000 (Vehicular A) (reference " TSI TR 125996V7.0.0UniversalMobile Telecommunications System (UMTS); Spatial channel model for Multiple InputMultiple Output (MIMO) simulations ", 2007) under the channel, and balanced with MMSE, adopt the bit error rate curve of the Turbo Enhancement Method of former hard decision to compare.Can find out that by accompanying drawing 4 performance that the part decision method of the MIMO-SCFDE wireless communication receiver that the present invention proposes combines with the Turbo Enhancement Method adopts the performance of the Turbo Enhancement Method of former hard decision to improve than MIMO-SCFDE wireless communication receiver.IMT2000 (Vehicular A) channel is 10 ^-1To 10 ^-2Scope in; Adopt one of part decision method to take turns Turbo Enhancement Method ratio and adopt one of former hard decision to take turns the about 0.5dB of performance improvement of Turbo Enhancement Method, that adopts the part decision method takes turns the Turbo Enhancement Method than the about 1dB of the performance improvement of taking turns the Turbo Enhancement Method that adopts former hard decision more more.

Accompanying drawing 5 is to have provided in the MIMO-SCFDE wireless communication system; The errored bit curve that the part decision method that adopts the present invention to propose combines with the Turbo Enhancement Method; In COST259 (UTx) (reference " 3GPP TR 25.943V6.0.03GPP channelmodels for deployment evaluation "; 2004) under the channel, and balanced with MMSE, adopt the bit error rate curve of the Turbo Enhancement Method of former hard decision to compare.Find out COST259 (UTx) channel, 4 * 10 by accompanying drawing 5 ^-2To 4 * 10 ^-3Scope in, adopt one of part decision method take turns the Turbo Enhancement Method than adopt former hard decision one take turns the Turbo Enhancement Method the about 0.5dB of performance improvement, 10 ^-2To 10 ^-3Scope in, adopt the Turbo Enhancement Method of taking turns of part decision method to arrive 1.5dB than the performance improvement of taking turns the Turbo Enhancement Method that adopts former hard decision about 1 more more.

Claims (1)

1. the part decision method in the mimo wireless communication receiver interference cancellation techniques is characterized in that, may further comprise the steps:

(1) thresholding is set, the decision statistic amount is carried out the part judgement, obtain the part decision signal; Detailed process is:

Thresholding ε is set, and the thresholding span is 0＜ε＜1; If the decision statistic scale is shown r, decision statistic amount r is carried out the part judgement, obtain the part decision signal

In the part judging process, need carry out the part judgement respectively to in-phase component and the quadrature component of decision statistic amount r, the process that the in-phase component of decision statistic amount r is carried out the part judgement is following: establishing k decision threshold of decision statistic amount r in-phase component is D _k, k=1 ..., K, K are the sum of decision threshold, if the in-phase component of decision statistic amount r drops on (D _k-ε, D _k+ ε) in the zone, do not adjudicate, keep initial value; If the in-phase component of decision statistic amount drops on (D _k-ε, D _k+ ε) outside the zone, then by the nearest principle of Euclidean distance, adjudicating is the coordinate figure of respective constellation point in-phase component correspondence, the part judging process of the quadrature component of decision statistic amount r is identical with the part judging process of in-phase component; After accomplishing the judgement of above-mentioned in-phase component and quadrature component, promptly obtain the part decision signal

(2) the part decision signal that step (1) is obtained carries out signal reconstruction, utilizes reconstruction signal to carry out interference eliminated; Wherein, the part decision signal carries out the process of signal reconstruction, is to utilize the part decision signal to adopt former hard decision reconstructing method to carry out signal reconstruction; Reconstruction signal carries out the process of interference eliminated, is that the signal that utilizes the reconstruct of part decision signal to obtain adopts the method for former hard decision interference eliminated to carry out interference eliminated.

Images