CN103312652B - A kind of space-frequency coding SFBC MIMO-OFDM system based on F matrix carries out the method for selected mapping method SLM - Google Patents

Abstract

The invention discloses an SLM method for an SFBCMIMO-OFDM system based on an F matrix, which solves the problems of peak-to-average power ratio and high computational complexity of SFBCMIMO-OFDM signals in the MIMO-OFDM system. By studying the orthogonality of signals on antennas after SFBC coding in MIMO systems, an algorithm to reduce system peak-to-average ratio and computational complexity is proposed. The present invention uses the proposed phase sequence group F matrix to process the signal after space-frequency coding SFBC, and the obtained optimal phase factor will be orthogonally coded and applied to the signal of each antenna after space-frequency coding SFBC, so that each The signal to be transmitted on the root antenna can avoid multiple IFFT operations, thereby reducing the computational complexity of the system. At the same time, the use of the F matrix also enables the MIMO-OFDM system to obtain good peak-to-average ratio performance.

Description

A Method of Selective Mapping SLM Based on F Matrix Space-Frequency Coding SFBC MIMO-OFDM System

technical field

The invention relates to the field of mobile communication, in particular to a selective mapping (SLM) method for reducing the peak-to-average ratio of a multiple-input multiple-output-orthogonal frequency division multiplexing (MIMO-OFDM) system based on space-frequency coding (SFBC).

Background technique

Orthogonal Frequency Division Multiplexing (OFDM) is widely used in Digital Audio Broadcasting (DAB), Digital Video Broadcasting (DVB), and Wireless Local Area Network (WLAN). Its advantages of high spectrum utilization and anti-multipath interference can effectively eliminate signal interference. , so it has received more and more attention. Multiple-Input Multiple-Output (MIMO) can improve channel transmission rate and obtain higher diversity gain or capacity gain by configuring multiple antennas on both sides of the transceiver. The combination of the two technologies in wireless communication is MIMO-OFDM (Multiple-Input Multiple-Output Orthogonal Frequency Division Multiplexing), in which OFDM can convert frequency-selective MIMO channels into parallel flat MIMO channels, and use multipath fading to achieve robust transmission of high-speed data. Therefore, MIMO-OFDM has become a very promising alternative in the fourth generation mobile communication system.

However, the disadvantage of the peak-to-average power ratio of MIMO-OFDM signals forces high-power amplifiers (HPAs) to have large backoffs, which reduces the efficiency of the HPA. The MIMO-OFDM signal produces in-band distortion and out-of-band noise, and then spectrum spread interference and signal distortion will lead to serious degradation of the overall system performance. The peak-to-average ratio has become a technical obstacle of MIMO-OFDM, so it is necessary to find a suitable method to reduce the peak-to-average ratio and improve the fidelity of the signal.

For OFDM systems, techniques for reducing PAPR can be roughly divided into three categories: one is coding techniques, document "Chen, HLiang, HPAPR Reduction of OFDM Signal Using Partial Transmit Sequences and Reed-Muller Codes, IEEE Communications Letters, vol.11, no.6, pp.528-530, Sep.2007 ", the algorithm idea is to use different encoding methods to avoid the occurrence of symbols that may generate higher PAPR, but the encoding process is more complicated. The second category is signal predistortion technology, including clipping method and companding method. This technology is the simplest and most direct nonlinear method to reduce the peak-to-average ratio, but the in-band distortion and out-of-band radiation introduced by the former will seriously reduce the system performance. performance. The latter uses the companding function and the inverse transformation function to achieve the reduction of the peak-to-average ratio, the literature "X.B.Wang, T.T.Tjhung, and C.S.Ng, Reduction of peak-to average power ratio of OFDM system using a companding technique, IEEETrans.Broadcast., vol.45, no.3, pp. 303-307, Sep.1999" proposed a method based on speech processing to make the μ-law companding scheme have good performance, but the reduced PAPR is at the cost of increasing the average power. The third category is scrambling technology, including selective mapping SLM and partial transmission sequence PTS. This technology uses different scrambling sequences to weight OFDM symbols. By setting the PAPR threshold condition, select the minimum PAPR from the sequence to be transmitted. One group is used for transmission, which significantly reduces the probability of high peak power signals, but due to the use of too many IFFTs, the computational complexity increases sharply, and the transmission of side information also causes the loss of data transmission rate.

However, there are few studies on PAPR reduction algorithms for MIMO-OFDM systems. The methods to solve the PAPR problem in MIMO-OFDM systems are summarized in two aspects: one is to directly transplant the methods in OFDM systems to MIMO-OFDM systems. Each antenna, such as the literature "ByungMooLee, RuiJ.P.deFigueiredo.SideInformationPowerAllocationForMIMO-OFDMPAPRReductionBySelectedMapping, IEEE, InternationConferenceonAcoustics, SpeechandSignalProcessing, 2007." can effectively reduce the system PAPR but the complexity is slightly higher; on the other hand, consider the MIMO-OFDM system itself characteristics, adopt a flexible method to deal with it. Literature "Joo-HeeMoon, Young-HwanYou, Won-GiJeon, Ki-WonKwon, Hyoung-Kyusong. Peak-to-AveragepowercontrolforMultiple-AntennaHIPEPLAN/2andIEEE802.11asystems.IEEETrans.OnConsumerElectronics, 2003,49(4):1078-1083." The proposed simplified PTS and SLM algorithms use the same sideband information for each antenna. Compared with the independent PTS and SLM algorithms, the PAPR reduction effect is slightly reduced, but the bit error rate performance is improved. The patent "Zhang Chaoyang et al., A Peak-to-Average Ratio Control Method, Receiver and Transmitter" describes a method for reducing the PAPR of the system through the combination of linear and nonlinear methods, but the algorithm process is too complicated. Patent "CN102075222A, Jiang Tao, Li Si, Qu Daiming, a method for reducing the peak-to-average power ratio of space-frequency coded MIMO-OFDM signals", although the sideband information does not need to be sent, the algorithm has a large amount of calculation. The CARI and its improved algorithm proposed by the patent "CN101073217A, TANM, ZORANLZ, YEHESKELBN.STBCMIMO-OFDMPeak-to-AveragePowerRatioReduction by Cross-AntennaRotationandInversion" do not have complex multiplication operations in each iteration, and the computational complexity is reduced. However, if the algorithm is directly extended to more In systems with transmitting antennas, the additional degrees of freedom provided by adding antennas cannot be fully utilized.

In order to further improve the PAPR performance of the MIMO-OFDM signal and reduce the computational complexity, the present invention proposes an SLM method for the SFBC MIMO-OFDM system based on the F matrix for the algorithm for reducing the PAPR of the above MIMO-OFDM system. By setting an F matrix as the phase factor, the SLM algorithm is used to scramble the signal before SFBC encoding, and after calculating the symbol peak-to-average ratio, the selected optimal phase factor sequence and its orthogonal encoding sequence are used as the optimal phase after SFBC encoding factor. Therefore, each antenna is only used for one IFFT in one symbol period, which greatly reduces the computational complexity, and the use of the F matrix also greatly improves the peak-to-average ratio of the system.

Contents of the invention

In order to more effectively overcome the above-mentioned defects in the MIMO-OFDM system, the purpose of the present invention is to provide a method that can reduce the peak-to-average power ratio in the MIMO-OFDM system and can be more effectively applied to practical communication systems.

The innovation of the present invention lies in proposing an effective phase factor for reducing the peak-to-average ratio of the MIMO-OFDM system, called F matrix.

The innovation of the present invention is that by proposing the F matrix as the phase factor, the SLM algorithm is used to scramble the signal before the SFBC encoding, and after calculating the symbol peak-to-average ratio, the selected optimal phase factor sequence and its orthogonal encoding sequence are used as the SFBC The optimal phase factor of the signal to be transmitted on each antenna after encoding. When OFDM modulation is performed on each antenna, only one IFFT is used in one symbol period, which greatly reduces the computational complexity.

The present invention is a kind of SLM method of SFBCMIMO-OFDM system based on F matrix, and transmitting antenna is T root (T>0), and the specific process of described method comprises the following steps:

Step 1 Input the binary data bit stream, modulate to obtain the mapping signal, and obtain the X _i frequency domain signal after serial-to-parallel conversion, where Xi represents the _i -th frequency domain symbol:

X _i =[X ₀ ,X ₁ ,…X _N-1 ] ^T (1)

Wherein, N represents the number of subcarriers, the upper limit of i is set to p, P>0, and [ ] ^T represents the transposition of the matrix;

Step 2 uses the F matrix in formula (2) to generate the initial phase factor group:

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; k</span>}}\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; N</span>}}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

Among them, I(N) represents an N*N unit matrix, k represents the attenuation factor, rand(*) represents an N*N square matrix composed of uniformly distributed numbers in the open interval (0,1); F matrix It is an N*N square matrix, k>0, and the PAPR performance of the system decreases with the increase of the value of k, and the value range of k is related to the number of subcarriers and the number of antennas;

Step 3: Carry out orthogonal coding on the X _i signal by using space-frequency coding to obtain a signal that can be transmitted on T antennas;

Step 4 multiplies X _i with the initial phase factor group generated by the F matrix, and calculates the PAPR after passing the weighted signal through IFFT, and then selects the optimal phase factor that makes the symbol have the minimum PAPR, denoted as F _m :

F _m ＝[p _0, p _1, ......, p _2m, p _2m+1 ] ^T (3)

in, $m\&Element;\&lsqb;0,\frac{N}{2}-1\&rsqb;;$

Step 5 Take the conjugate of the optimal phase factor F _m and denote it as F _m ^* , and then perform orthogonal coding on F _m and F _m ^* to obtain T phase factors; Step 6 combine the T phase factors obtained above with the corresponding T root The signal transmitted on the antenna is multiplied, and the respective PAPR values are calculated after IFFT, and the signal with max(PAPR ₁ , PAPR ₂ ...PAPR _T ) is selected as the signal to be sent After adding the guard interval, the signal is x~ _i , and the guard interval is obtained by the cyclic extension of the OFDM symbol;

Step 7 output signal Then read in the next signal X _i+1 , judge whether i+1 is equal to the upper limit p, P>0, if not, go to step 2, if yes, the loop ends.

The beneficial effect of the present invention is that a phase factor group that effectively reduces the peak-to-average ratio of the MIMO-OFDM system is proposed, which is called an F matrix. By using the F matrix to scramble the symbols before space-frequency coding, the optimal phase factor with the optimal PAPR can be selected, and the optimal phase factor is orthogonally coded, respectively acting on the corresponding signal after space-frequency coding , so that the signal to be transmitted on each antenna can avoid performing multiple IFFT operations, thereby reducing the computational complexity of the system. At the same time, the use of the F matrix also enables the MIMO-OFDM system to obtain good peak-to-average ratio performance.

Description of drawings

Figure 1 The basic block diagram of the transmitting end of the MIMO-OFDM system,

In the figure, the mapped symbols need to be space-time/frequency coded, then modulated by OFDM, and finally sent by multiple antennas;

Fig. 2 is a block diagram of SLM method of SFBC MIMO-OFDM system based on F matrix,

In the figure, the figure shows the algorithm block diagram under two antennas, it is proposed that the mapped symbols must be space-frequency coded SFBC, and finally the proposed SLM method is used to reduce the peak-to-average ratio of the MIMO-OFDM system;

The CCDF curve diagram of Fig. 3 different phase factor group numbers V (same attenuation factor K=5),

Among the figure, when the algorithm of the present invention adopts attenuation factor K=5, compare with original algorithm, independent SLM algorithm CCDF curve under the situation of different phase factor group number, wherein abscissa represents different peak-to-average ratio PAPR (dB) value, vertical The coordinates represent the complementary cumulative function CCDF value;

The CCDF curve diagram of different attenuation factors K (same phase factor group number V=16) of Fig. 4,

In the figure, the comparison of the CCDF curves of the algorithm of the present invention with the original algorithm and the independent SLM algorithm under the same phase factor group number, and the algorithm of the present invention has also changed the attenuation factor K, and carried out with the CCDF curve of the original algorithm and the independent SLM algorithm Comparison, where the abscissa represents different peak-to-average ratio PAPR (dB) values, and the ordinate represents the complementary cumulative function CCDF value; the CCDF curve figure when Fig. 5 attenuation factor K=16,

Among the figure, when the attenuation factor K=16 is set, the CCDF curves of the algorithm of the present invention and the original algorithm and the independent SLM algorithm are compared wherein the abscissa represents different peak-to-average ratio PAPR (dB) values, and the ordinate represents the complementary cumulative function CCDF value.

detailed description

Provide the specific implementation method of the present invention below, take sending antenna T=2 as example:

Step 1: Input the binary data bit stream, modulate with PSK/QAM to obtain the mapped signal, and obtain the X _i frequency domain signal after serial-to-parallel conversion, where X _i represents the i-th frequency domain symbol:

X _i =[X ₀ ,X ₁ ,......,X _N-1 ] ^T (1)

Wherein, N represents the number of subcarriers, the upper limit of i is set to P (P>0), and [ ] ^T represents the transposition of the matrix;

Step 2 uses the F matrix in formula (2) to generate the initial phase factor group:

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; k</span>}}\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; N</span>}}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

Among them, I(N) represents an N*N unit matrix, k represents the attenuation factor, and rand(*) represents an N*N square matrix composed of random numbers within the range of the open interval (0,1);

In step 3, _SFBC is used to perform orthogonal coding on the Xi signal, and the signals that can be transmitted on the two antennas a and b are obtained, which are respectively Xi _{, a} ′, Xi _{, b} ′, and the expressions are as follows:

Antenna a: ${x_{i,a}}^{\&prime;}={\&lsqb;x_i\left(0\right),-x_i^{*}\left(1\right),x_i\left(2\right),-x_i^{*}\left(3\right),\mathrm{ \dots },x_i\left(2m\right),-x_i^{*}(2m+1)\&rsqb;}^T---\left(3\right)$

Antenna b:

in, () ^* indicates the conjugate of the signal in brackets;

Step 4 multiplies X _i with the initial phase factor group generated by the F matrix, calculates PAPR after IFFT of the weighted signal, and selects the optimal phase factor that makes the symbol have the minimum PAPR, denoted as F _m :

F _m ＝[p _0, p _1, ......, p _2m, p _2m+1 ] ^T (5)

in, $m\&Element;\&lsqb;0,\frac{N}{2}-1\&rsqb;;$

Step 5 Take the conjugate of the optimal phase factor F _m and denote it as F _m ^* , and then perform orthogonal encoding on the two vectors to obtain two phase factors f _m,a and f _m,b , the expressions are as follows:

f _m,a =[p ₀ ,-p ₁ ^* ,......,p _2m ,-p _2m+1 ^* ] ^T (6)

f _m,b =[p ₁ ,p ₀ ^* ,......,p _2m+1 ,p _2m ^* ] ^T (7)

Among them, the phase factors f _{m, a} , f _{m, b} correspond to the signals X _{i, a} ′, X _{i, b} ′ respectively;

Step 6 Multiply the phase factor obtained above with the corresponding signal, calculate the respective PAPR values after IFFT and record them as PAPR _a and PAPR _b , and select the signal with max(PAPR _a , PAPR _b ) as the signal to be sent After adding the guard interval the signal is The guard interval is obtained by cyclic extension of OFDM symbols;

Step 7 output signal Then read the next signal X _i+1 , judge whether i+1 is equal to the upper limit value, if not, go to step 2, if yes, the loop ends.

The accompanying drawings will be further described in conjunction with the above specific implementation manners.

FIG. 2 is a system block diagram of an SLM method for an F-matrix-based SFBCMIMO-OFDM system. The block diagram describes in detail the specific implementation steps of the system of the present invention when the number of antennas is T=2. "Selecting the largest PAPR" in the figure corresponds to step 6, and steps 1 to 6 indicate that the processing of one OFDM symbol is completed, that is, the entire content of the system block diagram in Fig. 2 is completed.

Fig. 3 is a CCDF curve diagram of different phase factor group numbers V (the same attenuation factor K=5). The parameters of the simulation process are set as follows: Alamouti scheme SFBC coded MIMO system is adopted, the number of transmitting antennas is 2, the number of subcarriers is N=128, the system adopts oversampling coefficient L=4, and the binary signal adopts PSK modulation. This figure shows the comparison of the CCDF curves of the algorithm of the present invention with the original algorithm and the independent SLM algorithm at different numbers of phase factor groups when the attenuation factor K=5 is adopted. Seen from the CCDF curve in the figure, when V=4, the PAPR of the algorithm of the present invention has the improvement of about 6.5dB at ^10-3 than the original algorithm, and compared with the independent SLM algorithm at ^10-3 PAPR has about 4.5dB improvement. When V=16, the algorithm of the present invention has about 6.6dB improvement in PAPR at 10 ^-3 compared with the original algorithm, and about 2.8dB improvement in PAPR at 10 ^-3 compared with the independent SLM algorithm. And the increase of the number of phase factor groups makes the independent SLM algorithm improve by 1.6dB, but hardly affects the PAPR of the algorithm of the present invention, which shows that the F matrix is used to reduce the PAPR and has strong stability.

Fig. 4 is a CCDF curve diagram of different attenuation factors K (same number of phase factor groups). The parameters of the simulation process are set as follows: Alamouti scheme SFBC coded MIMO system is adopted, the number of transmitting antennas is 2, the number of subcarriers is N=128, the system adopts oversampling coefficient L=4, and the binary signal adopts PSK modulation. This figure simulates the influence of different attenuation coefficients on the algorithm of the present invention when the number of phase factors V=16, that is, the corresponding CCDF curves when K takes different values. It can be seen from the CCDF curve in the figure that as the attenuation coefficient decreases, the PAPR also improves. When k=5, the PAPR at 10 ^-2 is about 3.7dB; when k=2.5, the PAPR at 10 ^-2 is about 2dB; when k=1, the PAPR at 10 ^-2 is about 0.9dB. Therefore, the application of the F matrix improves the peak-to-average ratio of the MIMO-OFDM system.

Table 1 is a comparison table of algorithm calculation amount. As can be seen from the table, as the number of antennas increases, the number of IFFT times on each antenna of the independent SLM algorithm will increase greatly, and the algorithm of the present invention will increase the number of times of IFFT under the condition of multi-antenna multi-modulation in the MIMO-OFDM system less, thus effectively reducing the computational complexity of the system.

Table 1 Comparison table of algorithm calculation amount

algorithm name Algorithm calculation amount (IFFT times) SLM algorithm count1=(M*T)*G patent algorithm count2=(M+T)*G

Where M represents the number of phase factors (the dimension of the F matrix), T represents the number of antennas in the MIMO-OFDM system, and G represents the number of symbols contained in a frame signal. It can be seen that the algorithm of the present invention can reduce the computational complexity under the condition of multi-antenna modulation in the MIMO-OFDM system.

FIG. 5 is a CCDF curve diagram when the attenuation factor K=16. The parameters of the simulation process are set as follows: Alamouti scheme SFBC coded MIMO system is adopted, the number of transmitting antennas is 2, the number of subcarriers is N=128, the system adopts oversampling coefficient L=4, and the binary signal adopts PSK modulation. The original algorithm, the independent SLM algorithm, and the algorithm of the present invention in the figure represent the CCDF curves of the attenuation factor K=16 under the above parameter conditions. After K>16, the performance of the peak-to-average ratio of the present invention will be worse than that of the traditional independent SLM algorithm. Moreover, the selection of the K value is related to the number of antennas and the number of subcarriers.

Claims (1)

1. A method for selective SLM mapping of space-frequency coding SFBCMIMO-OFDM system based on F matrix is characterized in that:

step 1, inputting binary data bit stream, modulating to obtain mapping signal, and obtaining X after serial-to-parallel conversion_iFrequency domain signal, X_iRepresents the ith frequency domain symbol:

X_i＝[X₀,X₁,…X_N-1]^T(1)

wherein N represents the number of subcarriers, and the upper limit value of i is set to P, P > 0 [. cndot. ]]^TRepresentation matrixTransposing;

step 2, generating an initial phase factor set by using the F matrix in the formula (2):

$F=\frac{1}{k}I\left(N\right)+\frac{1}{N}rand\left(N\right)---\left(2\right)$

wherein, i (N) represents an N × N unit matrix, k represents an attenuation factor, and rand (x) represents an N × N square matrix composed of uniformly distributed numbers within an open interval (0, 1); the F matrix is an N x N square matrix, k is larger than 0, the PAPR performance of the system is reduced along with the increase of the k value, and the k value range is related to the number of subcarriers and the number of antennas;

step 3, adopting space-frequency coding to X_iCarrying out orthogonal coding on the signals to obtain signals capable of being transmitted on the T antennas;

step 4 of adding X_iMultiplying the initial phase factor group generated by the F matrix, calculating the PAPR after the weighted signals pass through IFFT, and selecting the optimal phase factor which leads the symbol to have the minimum PAPR from the PAPR, and recording the optimal phase factor as F_m：

F_m＝[p₀,p₁,.......,p_2m,p_2m+1]^T(3)

Wherein, $m\&Element;\&lsqb;0,\frac{N}{2}-1\&rsqb;;$

step 5 for the optimal phase factor F_mTake the conjugation as F_m ^*Then to F_mAnd F_m ^*Carrying out orthogonal coding to obtain T phase factors;

step 6, the obtained T phase factors and signals transmitted on the corresponding T antennas are subjected to complex multiplication, respective PAPR values are calculated after IFFT, and max (PAPR) is selected₁,PAPR₂…PAPR_T) As a signal to be transmittedAfter adding a guard interval signal ofThe guard interval is obtained by cyclic extension of the OFDM symbol;

step 7 output signalLater reading in the next signal X_i+1And judging whether i +1 is equal to the upper limit value P, wherein P is more than 0, if not, going to the step 2, and if so, ending the circulation.