CN101022325A - Maximum likelihood simplified detecting method for multi inputting and multi outputting antenna system space division multiplexing - Google Patents

Abstract

A ML simplified detecting method of space division multiplexing in multi-input/multi-output antenna system includes finding out a sending antenna with maximum modulation system when channel parameter is known, carrying out ML exhaustion search on modulation constellation of the other sending antenna then subtracting influence of these antenna signals in received signal and using IF algorithm to demodulate antenna with maximum modulation system, corresponding each ML exhaustion search to one IF demodulation result, finding out an optimum combination and using it as demodulation signal of sending antenna.

Description

The maximum likelihood simplification detection method of multi-input multi-output antenna system space division multiplexing

Technical field

The present invention relates to the detection method of the mobile communication link of a kind of multiple-input and multiple-output (MIMO) antenna data multiplexing transmission pattern.

Background technology

Along with development of Communication Technique and the people requirement to high-speed data communication, the MIMO technology is applied.Because in MIMO, a plurality of different data flow are gone out from different antenna transmission, and the many antennas of receiving terminal receive and these data of demodulation, thisly spectrum efficiency can be improved exponentially based on multiplexing space division multiple access mode.

The multiplexing reception detection method of MIMO has linearity test, maximum likelihood method (ML) detection etc., the linearity test mode has modes such as ZF (ZF) detection, MMSE detection again, for ML detects, these detection modes implement simpler, but its performance can not show a candle to ML and detects, and when transmitting antenna was identical with the reception antenna number, its space diversity degree only was 1.ML detects can obtain the smaller error rate, and when receiving fate greater than number of transmit antennas, its space diversity degree is the reception antenna number; In addition also have some nonlinear detection modes, detect such as globular decoding, lattice are simplified auxiliary (LRA) and are detected, continuously Interference Cancellation detection modes such as (SUC), for ML detection and linearity test, these methods are in complexity and aspect of performance embodiment compromise to a certain degree.It is simple relatively that wherein SUC detects realization, and its performance is better than, and ZF detects and MMSE detects.

Summary of the invention

Technical problem: the objective of the invention is to provide the maximum likelihood simplification detection method of a kind of multiple-input and multiple-output (MIMO) antenna system space division multiplexing, this method can effectively reduce the complexity of system, and does not reduce systematic function.

Technical scheme: the invention provides the multiplexing detection method of a kind of many antenna mobile communication systems MIMO.This method is in multi-input/output antenna is multiplexing, transmitting terminal has a plurality of transmitting antennas, the data that different transmitting antennas sends are inequality, adopt the detection signal of a certain transmitting antenna to detect with ZF " ZF ", and the detection of other transmitting antenna adopts maximum likelihood " ML " to detect exhaustive search, can improve the transmission data transfer rate like this; The signal of each each antenna transmission of the moment is to concentrate from limited complex signal to select an element, and the real part of this element and imaginary part be the carrier wave in-phase component and the quadrature component of corresponding band modulation respectively; If transmitting terminal has M antenna, the transmission signal set of i transmitting antenna is C _i={ c _I1, c _I2... c _Isi, s _iIt is the transmission signal set number of elements of i transmitting antenna;

When receiving terminal received with maximum likelihood detection method, the received signal vector was y, and channel gain matrix is H, and H estimates with channel estimation methods at receiving terminal;

Described detection method is as follows:

1) the transmitting antenna j that chooses peak signal set element number is used to be ZF and detects, promptly

$J=\arg \underset{i}{\max }\{s_i,i=\mathrm{1,2},...M\},$

2) except that j transmitting antenna, carry out maximum likelihood ML exhaustive search to sending signal on other transmitting antenna, getting these transmission signals is { x ₁, x ₂... x _J-1, x _J+1... x _M), x wherein _i∈ C _i,

3) in received signal, remove the influence of other antenna transmission signal, that is: $y_t=y-\underset{i=1,i\≠j}{\overset{M}{\mathrm{\Σ}}}{h}_i{x}_i$

H wherein _iBe the i row of H,

4) use y _tJ antenna carried out ZF detect, that is: $x_j=Q(\left(h_j^H{y}_t\right)/\left(h_j^H{h}_j\right))$ , wherein Q () is a decision function, its return value is that independent variable parameter and constellation are concentrated nearest constellation point; The conjugate transpose of subscript H representing matrix or vector,

5) make x={x ₁, x ₂... x _M), calculate d=‖ y-Hx ‖, wherein " ‖ ‖ " is the 2-norm of vector, and preserves hypothesis transmission vector x and 2-norm d,

6) if { x ₁, x ₂... x _J-1, x _J+1... x _M) do not have to have searched for the combination that might occur, then select a new combination, and turn back to 3) recomputate, newer d and old d size if new d is less, then are updated to x new combination; Otherwise look for one group of new combination to return 3 again) continue to handle,

7) repeat said process,, obtain detecting data x at last until searched for all combinations.

Can also choose N transmitting antenna and carry out maximum likelihood ML detection, remaining M-N transmitting antenna adopts ZF ZF to detect, wherein N＜M.

In the method, ZF detects the linearity test mode that adopts Minimum Mean Square Error MMSE, counteracting serial interference SUC, ordering counteracting serial interference OSUC.Maximum likelihood ML detects the globular decoding detection mode.

Beneficial effect: the maximum likelihood simplification detection method of multi-input multi-output antenna system space division multiplexing of the present invention has proposed a kind of maximum likelihood (ML) and ZF (ZF) detects the algorithm-simplification maximum likelihood algorithm (RML) that combines, this algorithm is identical with the maximum likelihood algorithm performance, but its complexity detects low than ML.This method can be simplified detection algorithm increases communication throughput, reduces receiver complexity (also having reduced receiver volume, power consumption etc. simultaneously).

Description of drawings

Fig. 1 is that multiaerial system is simplified ML testing process figure.

Embodiment

In MIMO was multiplexing, transmitting terminal had a plurality of transmitting antennas, and the data that different transmitting antennas sends are inequality, can make like this to send the data rate raising, but when receiving terminal received with maximum likelihood, complexity will the exponentially form increase.The method applied in the present invention is input ZF (ZF) method of a certain transmitting antenna, and the input of other transmitting antenna adopts maximum likelihood method.Through theoretical and to experimental results show that this detection method identical with the performance of ML detection method, and complexity is than order of magnitude of ML detection method reduction.The signal of each each antenna transmission of the moment is to concentrate from limited complex signal to select an element, and the real part of this element and imaginary part be the carrier wave in-phase component and the quadrature component of corresponding band modulation respectively.If transmitting terminal has M antenna, the transmission signal set of i transmitting antenna is C _j={ c _I1, c _I2... c _Isi), s _iBe the transmission signal set number of elements of i transmitting antenna, the received signal vector is y, and channel gain matrix is H, and H estimates with channel estimation methods at receiving terminal.Described detection method is as follows:

1) the transmitting antenna j that chooses peak signal set element number is used to do ZF and detects, promptly

$j=\arg \underset{i}{\max }\{s_i,i=\mathrm{1,2},...M\}$

2) except that j transmitting antenna, carry out the ML exhaustive search to sending signal on other transmitting antenna, getting these transmission signals is { x ₁, x ₂... x _J-1, x _J+1... x _M), x wherein _i∈ C _i

3) in received signal, remove the influence of other antenna transmission signal.That is: $y_t=y-\underset{i=1,i\≠j}{\overset{M}{\mathrm{\Σ}}}{h}_i{x}_i$

H wherein _iI row for H.

4) use y _tJ antenna carried out ZF detect, promptly $:x_j=Q(\left(h_j^H{y}_t\right)/\left(h_j^H{h}_j\right))$ , wherein Q () is a decision function, its return value is that independent variable parameter and constellation are concentrated nearest constellation point.The conjugate transpose of subscript H representing matrix or vector.

5) make { x ₁, x ₂... x _M, calculate d=‖ y-Hx ‖, wherein ‖ ‖ is the 2-norm of vector, and preserves x and d.

6) if { x ₁, x ₂... x _J-1, x _J+1... x _MDo not have to have searched for the combination that might occur, then select a new combination, and turn back to 3) recomputate, newer d and old d size if new d is less, then are updated to x new combination.Otherwise look for one group of new combination to return 3 again) continue to handle.

7) repeat said process,, obtain detecting data x at last until searched for all combinations.

With 3 transmitting antennas is example, and 1 the present invention is described further in conjunction with the accompanying drawings for we.

The modulation system of supposing three transmitting antennas identical (QPSK also can adopt different modulation systems and modulation constellation), optionally sending signal set is C={c ₁, c ₂, c ₃, c ₄}={ 1, j ,-1 ,-j}, wherein $j=\sqrt{-1},$ Be imaginary unit.

If receiving terminal is a N antenna, be x=(x when sending signal ₁, x ₂, x ₃) ^TThe time, channel matrix is:

$H={[h_1,h_2,...h_N]}^H={\left[\begin{array}{cccc}{h}_{11}& h_{21}& ...& h_{N1}\\ h_{12}& h_{22}& ...& h_{N2}\\ h_{13}& h_{23}& ...& h_{N3}\end{array}\right]}^H$

H wherein _IjRepresent the matrix between i transmitting antenna and j the reception antenna.

The signal of receiving at receiving terminal is y=(y ₁, y ₂... y _N) ^T

Because the modulation format of each transmitting antenna is identical, thus each transmitting antenna optionally to send the set element number all be 4, we select arbitrary transmitting antenna to be ZF to detect, other transmitting antenna is then searched for (promptly search for all may send signal combination).Suppose to select the 3rd transmitting antenna to carry out ZF and detect, and the 1st, 2 two the enterprising line search of transmitting antenna, testing result is x _d=(x _D1, x _D2, x _D3) ^TThe flow process of its detection as shown in Figure 1.

At first the 1st, 2 two transmitting antenna carried out exhaustive search at transmitting terminal.Searching times is 16 times altogether.When each search, the signal of hypothesis transmitting antenna 1,2 is correct earlier, deducts the influence of transmitting antenna 1,2 in the received signal, thereby in remaining signal only remaining the 3rd transmitting antenna and The noise.Signal to the 3rd antenna carries out the ZF detection on its planisphere.Promptly the signal on the 3rd transmitting antenna is relevant with the signal of day the 1st, 2 antenna.Thereby can calculate noise at this moment.After all having traveled through in all possibilities on the antenna 1,2, one group of signal of the noise minimum that selection is calculated (transmitting antenna 3 that transmitting antenna 1,2 and ZF detect) is as the decoding that sends signal.

Only be essential part of the present invention below, should be pointed out that application of the present invention is easy to expand to others.Organize into groups in twos sending symbol as needing in the demodulation to QOSTC (accurate orthogonal coding), then two the twoth symbols are carried out ML and detect, adopt this method earlier one of them symbol to be carried out exhaustive search and then the another one symbol is carried out the ZF detection.For another example, some antenna in the multiple transmit antennas is adopted exhaustive search, and another part antenna (being not only 1 antenna) is adopted ZF, MMSE (least mean-square error), perhaps OSUC detection modes such as (Interference Cancellations of ordering), though some loss on performance like this has a large amount of minimizings is arranged on the amount of calculation.And other expansion on basis of the present invention all should belong to scope of patent protection of the present invention.

Claims (4)

1. the maximum likelihood simplification detection method of a multi-input multi-output antenna system space division multiplexing, it is characterized in that in multi-input/output antenna is multiplexing, transmitting terminal has a plurality of transmitting antennas, the data that different transmitting antennas sends are inequality, adopt the detection signal of a certain transmitting antenna to detect with ZF " ZF ", and the detection of other transmitting antenna adopts maximum likelihood " ML " to detect exhaustive search, can improve the transmission data transfer rate like this; The signal of each each antenna transmission of the moment is to concentrate from limited complex signal to select an element, and the real part of this element and imaginary part be the carrier wave in-phase component and the quadrature component of corresponding band modulation respectively; If transmitting terminal has M antenna, the transmission signal set of i transmitting antenna is $C_i=\{c_{i1},c_{i2},...c_{{\mathrm{is}}_i}\}$ , s _iIt is the transmission signal set number of elements of i transmitting antenna;

When receiving terminal received with maximum likelihood detection method, the received signal vector was y, and channel gain matrix is H, and H estimates with channel estimation methods at receiving terminal;

Described detection method is as follows:

1) the transmitting antenna j that chooses peak signal set element number is used to be ZF and detects, promptly

$j=\arg \underset{i}{\max }\{s_i,i=\mathrm{1,2},...M\},$

2) except that j transmitting antenna, carry out maximum likelihood ML exhaustive search to sending signal on other transmitting antenna, getting these transmission signals is { x ₁, x ₂... x _J-1, x _J+1... x _M, x wherein _i∈ C _i,

3) in received signal, remove the influence of other antenna transmission signal, that is: $y_t=y-\underset{i=1,i\≠j}{\overset{M}{\mathrm{\Σ}}}{h}_i{x}_i$

H wherein _iBe the i row of H,

4) use y _tJ antenna carried out ZF detect, that is: $x_j=Q(\left(h_j^H{y}_t\right)/\left(h_j^H{h}_j\right)),$ Wherein Q () is a decision function, and its return value is that independent variable parameter and constellation are concentrated nearest constellation point; The conjugate transpose of subscript H representing matrix or vector,

5) make x={x ₁, x ₂... x _M), calculate d=‖ y-Hx ‖, wherein " ‖ ‖ is the 2-norm of vector, and preserves hypothesis transmission vector x and 2-norm d,

6) if { x ₁, x ₂... x _J-1, x _J+1... x _MDo not have to have searched for the combination that might occur, then select a new combination, and turn back to 3) recomputate, newer d and old d size if new d is less, then are updated to x new combination; Otherwise look for one group of new combination to return 3 again) continue to handle,

7) repeat said process,, obtain detecting data x at last until searched for all combinations.

2. according to the maximum likelihood simplification detection method of the described multi-input multi-output antenna system space division multiplexing of claim 1, it is characterized in that choosing N transmitting antenna and carry out maximum likelihood ML detection, remaining M-N transmitting antenna adopts ZF ZF to detect, wherein N＜M.

3. according to the maximum likelihood simplification detection method of claim 1 or 2 described multi-input multi-output antenna system space division multiplexings, it is characterized in that ZF detects the linearity test mode that adopts Minimum Mean Square Error MMSE, counteracting serial interference SUC, ordering counteracting serial interference OSUC.

4. according to the maximum likelihood simplification detection method of claim 1 or 2 described multi-input multi-output antenna system space division multiplexings, it is characterized in that maximum likelihood ML detects the globular decoding detection mode.

Images