CN101202607A - Enhanced multi-aerial signal interfere counteract device with code merged retransmission system - Google Patents

Abstract

An enhanced multi-antenna signal interruption canceller of a retransmission system with code combination includes an equivalent signal noise ratio confirming module (101) which calculates the signal noise ratio of a retransmitted data flow; a detecting sequence confirming module (102) used for confirming the sequence of a detecting data flow according to the size of the equivalent signal noise ratio of each data flow; a space time detecting module (103) used for detecting parallel data flows; a retransmission data combining module(104) used for combining the retransmitted data; a decoding module (105) used for carrying through channel decoding on the detected data; a hard judging module (106) used for carrying through hard judging on the data after being decoded; a coding module (107) used for carrying through channel coding and debugging on the data after being detected and judged; and an interruption canceling module (108) used for canceling the detected data out of the received signal. The invention fully utilizes retransmission information and achieves superior performance under the situation that the realizing complexity is increased very little.

Description

Have yard multi-aerial signal interfere counteract device of the enhancing of the retransmission system that merges

Technical field

The present invention relates to the field of information transmission in the multiple antenna communication, particularly have yard multi-aerial signal interfere counteract device of the enhancing of the retransmission system that merges.

Background technology

Utilize space multiplexing technique, promptly the BLAST technology can improve the transmission rate of data effectively under the limited frequency spectrum resources condition.The detection algorithm of existing BLAST technology mainly contains ZF detection algorithm (ZF), least mean-square error detection algorithm (MMSE), and ZF adds Interference Cancellation detection algorithm (ZF SIC), and least mean-square error adds Interference Cancellation detection algorithm (MMSE SIC).

The following describes these detection algorithms.

The linearity test algorithm

If received signal is:

r＝Hs+n

In the formula, H is N * M channel matrix, and s is that the M dimension sends signal phasor, and r is a N dimension received signal vector, and n is that N ties up independent white Gaussian noise vector.

For zero forcing algorithm,

${\hat{s}}_{\mathrm{ZF}}={\left(H^{H}H\right)}^{-1}{H}^{H}r$

$=s+{\left(H^{H}H\right)}^{-1}{H}^{H}n$

For least-mean-square error algorithm,

${\hat{s}}_{\mathrm{MMSE}}=(H^{H}H+{\mathrm{\σ}}^{2}I{)}^{-1}{H}^{H}r$

$=s+(H^{H}H+{\mathrm{\σ}}^{2}I{)}^{-1}{H}^{H}n$

The ZF linear algorithm has been eliminated many interference between antennas fully, and the ZF linear algorithm has strengthened noise power when having eliminated many antenna interference fully; And least-mean-square error algorithm has been taken all factors into consideration interference between antennas and noise power increases, and its performance is better than the ZF linear algorithm.

The serial interference elimination algorithm

More than zero forcing algorithm of Jie Shaoing and least-mean-square error algorithm all are linear algorithms, are used for many antennas and disturb inhibition.

Linear technique is feasible, if but use nonlinear technology will obtain more performance.A kind of non-linear preferably alternative method is interference eliminated (Interference Canceller) algorithm.So-called serial interference elimination is exactly will come from the interference that has detected composition to remove when detecting current demand signal, thereby reduces to disturb the influence to less smnr data, is similar to decision feedback equalization.Fig. 1 has described the testing process that every day, the line coded interference was offset.

For serial interference elimination, testing process is:

For the ZF-SIC detector

G ₁＝H ⁺＝(H ^HH) ^-1H ^H

For the MMSE-SIC detector

G ₁＝H ⁺＝(H ^HH+σ ²I) ^-1H ^H

k ₁＝arg?min||(G ₁) _j|| ²

k ₁, k ₂..., k _MOrdering for testing process.

$w_{k_i}=(G_i{)}_{k_i}$

$y_{k_i}=w_{k_i}^T{r}_i$

${\hat{a}}_{k_i}=Q\left(y_{k_i}\right)$

So just ruled out a signal.Then its influence is deducted from received signal, and determine new pseudo-inverse matrix, determine new judgement order.

$r_{i+1}=r_i-{\hat{a}}_{k_i}{\left(H\right)}_{k_i}$

$G_{i+1}=H_{k_i}^{\±}$

$k_{i+1}=\underset{j\&NotElement;\{{\mathrm{<figure-callout\; id="1"\; label="k"\; filenames="A20061016589100091.png,A20061016589100111.png"\; state="\{\{state\}\}">k</figure-callout>}}_1\&CenterDot;\&CenterDot;\&CenterDot;k_i\}}{\arg \min }{\left|\right|{\left({\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="A20061016589100091.png,A20061016589100111.png"\; state="\{\{state\}\}">G</figure-callout>}}_{1+i}\right)}_j\left|\right|}^2$

i←i+1

Formed a cyclic process like this, until i=M, after all signals were all adjudicated and finished, cyclic process finished, and this process as shown in Figure 2.

We know that determining of judgement order is very big to the performance impact of Interference Cancellation, cause very serious error code transmission situation if the order of offsetting is determined improper meeting.

The HARQ technology

The HARQ technology is a kind of scheme of effective raising throughput of system, the Chase-Combining technology be exactly the data message of the 1st transmission add redundancy encoding and the data message that retransmits later on to add redundancy encoding be identical, and the IR technology be exactly the 1st time the transmission be data message and check information, if it is incorrect, in re-transmission, increase redundant coded information again.The present invention considers is situation with the two combination, i.e. the information of data is all transmitted in the 1st time and later re-transmission, and the redundancy of each transmission is different.As shown in Figure 3.

In the system that combines of spatial reuse of existing C ode combining HARQ and line coding every day, in definite detection order, do not consider the situation whether re-transmission is arranged, so the optimal detection order definite according to the current transmission channel situation, in fact not optimum, as shown in Figure 4.

What need improvement is exactly in the detection of determining spatial reuse and interference cancellation order the re-transmission factor of data flow to be taken into account, and makes full use of various channel informations, and under the very little situation of implementation complexity increase, it is optimum that performance reaches.

Summary of the invention

The purpose of this invention is to provide a kind of yard multi-aerial signal interfere counteract device of the enhancing of the retransmission system that merges that has.

For achieving the above object, a kind of yard multi-aerial signal interfere counteract device of the enhancing of the retransmission system that merges that has comprises:

Signal to noise ratio determination module of equal value is with its signal to noise ratio of equal value of data-flow computation that retransmits;

Detection order determination module is according to the definite order that detects data flow of the size of each data flow signal to noise ratio of equal value;

Detection module when empty detects parallel data flow;

Data retransmission merges module, and the data flow that retransmits is merged;

Decoding module carries out channel decoding to detected data;

The hard decision module is carried out hard decision to the data after the decoding;

Coding module carries out chnnel coding and modulation to the data after detecting, adjudicating;

The Interference Cancellation module balances out detected data from the signal that receives.

The present invention makes full use of retransmission information, and under the very little situation of implementation complexity increase, performance has reached optimum.

Description of drawings

Fig. 1 is that the receiving terminal of Code combining HARQ of the present invention and the multiplexing combination of line space encoder every day detects and the Interference Cancellation schematic diagram;

Fig. 2 is the detection principle of the multiplexing reception counteracting serial interference of line space encoder every day;

Fig. 3 is the transmitting terminal schematic diagram of Code combining HARQ and the multiplexing combination of line space encoder every day;

Figure is that the receiving terminal of 4Code combining HARQ and the multiplexing combination of line space encoder every day detects and the Interference Cancellation schematic diagram;

Fig. 5 is that the receiving terminal of Code combining HARQ of the present invention and the multiplexing combination of line space encoder every day detects and the Interference Cancellation flow chart;

Fig. 6 Code combining of the present invention HARQ and the multiplexing combination of line space encoder every day and not improvement project performance comparison (4 transmitting antennas, 4 reception antennas).

Embodiment

The present invention combines counteracting serial interference and Code combining HARQ, and testing process is as follows:

For the ZF-SIC detector

G ₁＝H ⁺＝(H ^HH) ^-1H ^H

For the MMSE-SIC detector

G ₁＝H ⁺＝(H ^HH+σ ²I) ^-1H ^H

Check whether the data flow of re-transmission is arranged,,, try to achieve new G according to the signal to noise ratio of data retransmission stream and the signal to noise ratio of the same data flow that sent in the past if having ₁

${\mathrm{SNR}}_{\mathrm{equv}}=2^{\underset{b=1}{\overset{B}{\mathrm{\&Sigma;}}}\left(\frac{L_b}{L}\underset{n=1}{\overset{N}{\mathrm{\&Sigma;}}}{\log }_2(1+{\mathrm{SNR}}_{\mathrm{nb}})\right)}-1$

G ₁＝SNR ₀/SNR _equv

If not data retransmission stream, then remain unchanged,

${\mathrm{<figure-callout\; id="1"\; label="k"\; filenames="A20061016589100091.png,A20061016589100111.png"\; state="\{\{state\}\}">k</figure-callout>}}_1=\underset{j}{\arg }\min {\left|\right|(G_1{)}_j\left|\right|}^2$

k ₁, k ₂..., k _MOrdering for testing process.

$w_{k_i}={\left(G_i\right)}_{k_i}$

$y_{k_i}=w_{k_i}^T{r}_i$

Event data stream k _iBe data retransmission stream, then the data flow of this detection delivered to decoder with the data flow of former detection and decipher, quantize to obtain y ' then _K1,, then the data flow of this detection is delivered in the decoder and is deciphered if not data retransmission stream.

${\hat{a}}_{k_i}=Q\left(y_{k_i}^{\&prime;}\right)$

So just ruled out a signal.Then its influence is deducted from received signal, and the pseudo-inverse matrix of looking for novelty

$r_{i+1}=r_i-{\hat{a}}_{k_i}{\left(H\right)}_{k_i}$

$G_{i+1}=H_{k_i}^{\&PlusMinus;}$

Check in the data flow that does not detect whether the data flow of re-transmission is arranged,,, try to achieve new G according to the signal to noise ratio of data retransmission stream and the signal to noise ratio of the same data flow that sent in the past if having _I+1

${\mathrm{SNR}}_{\mathrm{equv}}=2^{\underset{b=1}{\overset{B}{\mathrm{\&Sigma;}}}\left(\frac{L_b}{L}\underset{n=1}{\overset{N}{\mathrm{\&Sigma;}}}{\log }_2(1+{\mathrm{SNR}}_{\mathrm{nb}})\right)}-1$

G _i+1＝SNR ₀/SNR _equv

Wherein: SNR _EquvBe that data flow retransmits the signal to noise ratio of equal value after merging, N is the number of times that data flow is transmitted altogether, and L is the length of whole data block, L _bBe the length of b data sub-block, B is the number of the sub-piece of transmission that is divided into of each data block, SNR _NbIt is the signal to noise ratio of the n time transmission of b data block.

If not data retransmission stream, then remain unchanged,

$k_{i+1}=\underset{j\&NotElement;\{{\mathrm{<figure-callout\; id="1"\; label="k"\; filenames="A20061016589100091.png,A20061016589100111.png"\; state="\{\{state\}\}">k</figure-callout>}}_1\&CenterDot;\&CenterDot;\&CenterDot;k_i\}}{\arg \min }{\left|\right|{\left(G_{i+1}\right)}_j\left|\right|}^2$

k _I+1For detecting the sequence number of data flow at present.

i←i+1

Formed a cyclic process like this, until i=M, after all signals were all adjudicated and finished, cyclic process finished.Its process as shown in Figure 5.

From this process we as can be seen, the method by this paper proposes combines HARQ and Interference Cancellation.When calculating detection and interference cancellation order, the factor that retransmits is taken into account, so just can be really according to signal to noise ratio sequence detection and Interference Cancellation data flow from big to small.Can improve the reliability of interference-cancelled signals, reduce the error propagation of Interference Cancellation, thereby improve total reliability of data transmission.

Describe multi-antenna signal below in conjunction with Fig. 1 and detect and interference offset device, comprising:

Signal to noise ratio determination module of equal value is with its signal to noise ratio of equal value of data-flow computation that retransmits;

Detection order determination module is according to the definite order that detects data flow of the size of each data flow signal to noise ratio of equal value;

Detection module when empty detects parallel data flow;

Data retransmission merges module, and the data flow that retransmits is merged;

Decoding module carries out channel decoding to detected data;

The hard decision module is carried out hard decision to the data after the decoding;

Coding module carries out chnnel coding and modulation to the data after detecting, adjudicating;

The Interference Cancellation module balances out detected data from the signal that receives.

Present embodiment has adopted four transmitting antennas and four reception antennas.The BPSK modulation, channel is the flat Rayleigh fading channel.Adopt Code combining HARQ technology, maximum retransmission is 3 times.

We see that the original interference cancellation algorithm of improved algorithm throughput ratio improves significantly from Fig. 6, and abscissa is the Eb/NO of the noise of the gross power of transmitting terminal and receiving terminal.Ordinate is a throughput.

Claims (4)

1. one kind has yard multi-aerial signal interfere counteract device of the enhancing of the retransmission system that merges, and comprising:

Signal to noise ratio determination module (101) of equal value is with its signal to noise ratio of equal value of data-flow computation that retransmits;

Detection order determination module (102) is according to the definite order that detects data flow of the size of each data flow signal to noise ratio of equal value;

Detection module (103) when empty detects parallel data flow;

Data retransmission merges module (104), and the data flow that retransmits is merged;

Decoding module (105) carries out channel decoding to detected data;

Hard decision module (106) is carried out hard decision to the data after the decoding;

Coding module (107) carries out chnnel coding and modulation to the data after detecting, adjudicating;

Interference Cancellation module (108) balances out detected data from the signal that receives.

2. by the described Canceller of claim 1, it is characterized in that described antenna is the MIMO antenna.

3. by the described Canceller of claim 1, it is characterized in that described when empty detection module comprise ZFSIC detector or MMSE-SIC detector.

4. by the described Canceller of claim 1, it is characterized in that the signal to noise ratio of equal value of each data flow is:

${\mathrm{SNR}}_{\mathrm{equv}}=2^{\underset{b=1}{\overset{B}{\mathrm{\&Sigma;}}}\left(\frac{L_b}{L}\underset{n=1}{\overset{N}{\mathrm{\&Sigma;}}}{\log }_2(1+{\mathrm{SNR}}_{\mathrm{nb}})\right)}-1$

Wherein: SNR _EquvBe that data flow retransmits the signal to noise ratio of equal value after merging, N is the number of times that data flow is transmitted altogether, and L is the length of whole data block, L _bBe the length of b data sub-block, B is the number of the sub-piece of transmission that is divided into of each data block, SNR _NbIt is the signal to noise ratio of the n time transmission of b data block.

Images