CN104113393A - Superposition coded modulation method based on subcarrier index modulation (SIM)-orthogonal frequency division multiplexing (OFDM) - Google Patents

Abstract

The invention relates to the technical field of communication anti-jamming, in particular to a superposition coded modulation (SCM) technology communication system, subcarrier index modulation (SIM) and related time domain detecting technologies. According to an SCM method based on SIM-orthogonal frequency division multiplexing (OFDM), ideas of SIM-OFDM and spatial modulation (SM) are used as reference, a part of a string of bit data is used for transmission, the other part of the string of bit data is used as indexing bits, and the indexing bits are respectively coded. During transmission in an OFDM system, particular subcarrier is selected for transmitting data according to the bit data after coding of the indexing bits. A maximum likelihood method is used at a receiving end to obtain soft information of the indexing bits, and the indexing bits are decoded. Compared with prior SCM methods, the SCM method based on the SIM-OFDM has the advantages that the SCM superposition layers are reduced, the complexity is reduced, and error rate performances can be maintained or even improved to a certain degree. Meanwhile, the SCM is used for replacing quadrature amplitude modulation (QAM), and performances can be improved under the same transmission efficiency.

Description

A SIM-OFDM-based Superposition Coding Modulation Method

technical field

The invention belongs to the technical field of communication anti-jamming, and in particular relates to a superposition coded modulation (Superposition Coded Modulation) communication system, a subcarrier index modulation (Subcarrier Index Modulation, SIM) and related time domain detection technology.

Background technique

In noisy channels, coded modulation can be efficiently and reliably transmitted, such as early trellis coded modulation (TCM) and multilevel coded modulation, which can get better performance gains when combined with binary coded multi-ary signals. And bit-interleaved coded modulation (BICM) can obtain better performance with a relatively simple method. The Superposition Coded Modulation (SCM) can be regarded as a special case of BICM. The detection complexity of BICM system increases exponentially with the number of transmitted bits, while the complexity of SCM increases with the increase of the number of layers, and the performance of multi-layer SCM is obviously better than that of BICM.

Recently, an OFDM system based on Subcarrier Index Modulation (SIM) has been proposed as a new multi-carrier communication method. The SIM method uses index bits to select different carriers to transmit data, which makes SIM OFDM have better performance than traditional OFDM. However, the original SIM scheme may cause error bit propagation to cause error concentration, and it is difficult to determine a suitable threshold for demodulation when the receiving end detects. In response to the above problems, some researches have improved the SIM OFDM method and proposed a simple detection method.

In SCM, the more layers are superimposed, the lower the power allocated to each layer, the more serious the interference between layers, and the more difficult it is for the receiving end to distinguish the signals of each layer. If the information of each layer cannot be distinguished , it is impossible to divide and combine, so the performance is less and less obvious as the number of stacked layers increases.

Contents of the invention

The present invention learns from the idea of SIM-OFDM and the idea of spatial modulation SM, uses a part of a string of bit data for transmission, and the other part as index bits, and encodes them respectively. During transmission in the OFDM system, a specific subcarrier is selected to transmit data according to bit data coded by index bits. The maximum likelihood method is adopted at the receiving end to obtain the soft information of the index bits, and then decode the index bits.

A kind of superposition coding modulation method based on SIM-OFDM, specifically as follows:

S1. Generate M ₁ +M ₂ ×L information bits from a digital source, and pass through a serial-to-parallel converter. The first M ₁ bits are used as index bits, and the last M ₂ ×L bits are used as data bits, and the data bits are sent to another Serial-to-parallel converter;

S2. Index bit encoding and interleaving: encode the M ₁ bits described in S1, and interleave the encoded index bits to obtain index data {u _j }, wherein the encoding method is Turbo code, and the encoding rate is R , j=1,2,...,M ₁ /R;

S3, perform SCM modulation: after the M ₂ ×L bits described in S1 pass through a serial-to-parallel converter, the data is divided into L layers for transmission, and the data of the first layer is After a spreader with a spreading length of S, interleaving, and BPSK modulation, we get Each layer is weighted by the weighted value {ω ^(l) ,ω ⁽²⁾ ,…,ω ^(L) }, and the weighted data are superimposed to obtain Send to the SIM modulator, waiting to be sent, where m=1,2,...,M ₂ , j=1,2,...,M ₂ ×S;

S4, carry out SIM modulation: obtain for S3 SIM modulation is performed by selecting 1 from 2 subcarriers, that is, the M ₂ ×S bit data obtained by S3 is transmitted through M ₂ ×S×2 subcarriers;

S5. Perform IFFT transformation on the M ₂ ×S×2-bit data obtained in S4 to the time domain to the said After adding the cyclic prefix, add noise through the channel, perform FFT transformation to the frequency domain, and remove the cyclic prefix to obtain Among them, N=2×M ₂ ×S, K represents the number of subcarriers actually having data, is the channel parameter value corresponding to the subcarrier, the is unknown and needs to be obtained by ML method, Represents the bits of layer l at the jth moment, n _j is Gaussian noise with zero mean and variance of δ ² =N ₀ /2;

S6. The ML criterion detects the index bit to obtain the channel parameter, specifically as follows:

S61. Use and Indicates that when the jth bit is 1 and 0, the likelihood ratio information of the index bit is calculated by using the ML method $L\left(u_j\right)=\log \{\underset{x^{\′}\∈x}{\underset{u^{\′}=u_j^{\left(1\right)}}{\mathrm{\Σ}}}{e}^{\frac{{\left|\right|\mathrm{the\; y}-h_{u^{\′}}{x}^{\′}\left|\right|}^2}{-{\mathrm{\δ}}^2}}/\underset{x^{\′}\∈x}{\underset{u^{\′}=u_j^{\left(1\right)}}{\mathrm{\Σ}}}{e}^{\frac{{\left|\right|\mathrm{the\; y}-h_{u^{\′}}{x}^{\′}\left|\right|}^2}{-{\mathrm{\δ}}^2}}\},$ Wherein, X represents the set of all transmitted symbols, h _u' represents the channel parameter of the corresponding subcarrier, and the h _u' is determined according to the index bit u';

S62, obtain according to the L(u _j ) described in S61 value, if L(u _j )≥0, then the is the channel parameter value on the 2jth subcarrier, if L(u _j )<0, then the is the channel parameter value on the 2j-1th subcarrier;

S63. According to S62, obtain the channel parameter H={h _j }, where j=M ₂ ×S;

S7. Send the channel parameter information H obtained in S6 to the detection and decoding of the SCM, deinterleave the likelihood ratio information L(u _j ) of the index bits described in S61 and send it to the decoder for decoding. After the decoding is completed, Obtain the detected index bit data, wherein the decoding method corresponds to the encoder at the sending end;

S8, SCM detection, specifically as follows:

S81. In the GA module, Gaussian approximation detection is adopted, and the weighting coefficient is W={w ⁽¹⁾ , w ⁽²⁾ ,..., w ^(l) }, and the received information at the jth moment is $r_j=h_j\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{w}^{\left(l\right)}{c}_j^{\left(l\right)}+{\mathrm{no}}_j;$

S82. Send the received information to the GA module, as described in S3 can get ${\mathrm{\&zeta;}}_j^{\left(l\right)}=r_j-h_j{w}^{\left(l\right)}{c}_j^{\left(l\right)},\mathrm{E.}\left(r_j\right)=h_j\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{w}^{\left(l\right)}\mathrm{E.}\left(c_j^{\left(l\right)}\right),\mathrm{Var}\left(r_j\right)=h_j\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{w}^{\left(l\right)}\mathrm{Var}\left(c_j^{\left(l\right)}\right)+{\mathrm{\&delta;}}^2,$ in, for interference;

S83, initialization, Get the mean and variance of the received signal r _j ;

S84. Get mean of $\mathrm{E.}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)=\mathrm{E.}\left(r_j\right)-h_j{w}^{\left(l\right)}\mathrm{E.}\left(c_j^{\left(l\right)}\right)$ and variance $\mathrm{Var}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)=\mathrm{Var}\left(r_j\right)-\left(h_j{w}^{\left(l\right)}{c}_j^{\left(l\right)}\right);$

S85, obtain the logarithmic likelihood ratio of the lth bit transmitted at the jth moment according to S84 as $L\left(c_j^{\left(l\right)}\right)=\log \left(\frac{\mathrm{PR}(c_j^{\left(l\right)}=+1|r_j)}{\mathrm{PR}(c_j^{\left(l\right)}=-1|r_j)}\right)=\frac{2h_j(r_j-\mathrm{E.}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)}{\mathrm{Var}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)};$

S86. According to S85, obtain the likelihood ratio information L(c) corresponding to the transmitted data of all users in the L layer;

S9, at each layer in the SCM, perform deinterleaving according to L(c) described in S86 to obtain L(a), use Indicates the i-th bit of layer l, the Send it to the despreading module for despreading to obtain the log likelihood ratio of the first bit of the l layer where S is the spreading length in the transmitter, is the spreading bit;

S10, obtain the described in S9 The extra-bit information after deinterleaving $\mathrm{ext}\left(a_j^{\left(l\right)}\right)={\mathrm{the\; s}}_j^{\left(l\right)}L\left(d_1^{\left(l\right)}\right)-L\left(a_j^{\left(l\right)}\right);$

S11. Interleave the extra-bit information described in S10 to obtain will be described into the GA module again, for The mean and variance of are updated: $\mathrm{E.}\left(c_j^{\left(l\right)}\right)=\tanh \left(\frac{\mathrm{ext}\left(c_j^{\left(l\right)}\right)}{2}\right),\mathrm{Var}\left(c_j^{\left(l\right)}\right)=1-\mathrm{E.}\left(c_j^{\left(l\right)}\right);$

S12. Complete the update of the mean value and variance of all data in the L layer, substitute the mean value and variance of the updated sent data into S8-S9, iterate according to the set number of iterations CH, and obtain the value obtained when the iteration reaches the CHth time available in, Indicates the estimate of the mth bit of the kth layer;

S13, the S12 said The index bit data detected in step S7 is sent to the serial-to-parallel converter to obtain the final output bit.

Further, the initialization in S83 is performed during the first iteration.

Further, in S12, the number of iterations CH≥5.

The beneficial effects of the present invention are:

The present invention provides a SIM-OFDM-based SCM technology. The technology is based on the SCM technology and combined with the method of SIM. Part of the data is used for SIM modulation as an index, and part of the data is used for SCM transmission. Compared with the previous SCM, the performance of the whole method can not only reduce the number of superimposed layers of SCM and reduce the complexity, but also maintain or even improve the bit error rate performance to a certain extent. At the same time, using SCM modulation instead of QAM modulation can also improve performance under the same transmission efficiency. And adopting the method of SIM-OFDM can make its anti-ICI effect better, and when changing the number of effective subcarriers of IFFT, the spectral efficiency can also be adjusted flexibly.

Description of drawings

Fig. 1 is a flowchart of the transmitter scheme proposed by the present invention.

Fig. 2 is a flow chart of reception detection proposed by the present invention.

Detailed ways

The present invention combines superposition coding modulation with SIM-OFDM, so that the sending end of the SCM can select a certain subcarrier to send data in each time slot according to certain rules, so that the receiving end can use a certain method to compare the index bits Decoding is performed to find effective subcarriers and corresponding channel parameter values. Therefore, at the receiving end of the SCM, only the Gaussian approximation method needs to be used for detection. As long as an appropriate method is adopted for decoding the index bits, the combination of SCM and SIM-OFDM can obtain better performance gains than the combination of other modulation methods and SIM-OFDM. details as follows:

S1. Generate M ₁ +M ₂ ×L information bits from a digital source, and pass through a serial-to-parallel converter. The first M ₁ bits are used as index bits, and the last M ₂ ×L bits are used as data bits, and the data bits are sent to another Serial-to-parallel converter;

S2. Index bit encoding and interleaving: encode the M ₁ bits described in S1, and interleave the encoded index bits to obtain index data {u _j }, wherein the encoding method is Turbo code, and the encoding rate is R , j=1,2,...,M ₁ /R;

S3, perform SCM modulation: after the M ₂ ×L bits described in S1 pass through a serial-to-parallel converter, the data is divided into L layers for transmission, and the data of the first layer is After a spreader with a spreading length of S, interleaving, and BPSK modulation, we get Each layer is weighted by the weighted value {ω ^(l) ,ω ⁽²⁾ ,…,ω ^(L) }, and the weighted data are superimposed to obtain Send to the SIM modulator, waiting to be sent, where m=1,2,...,M ₂ , j=1,2,...,M ₂ ×S;

S4, carry out SIM modulation: obtain for S3 SIM modulation is performed by selecting 1 from 2 subcarriers, that is, the M ₂ ×S bit data obtained in S3 is transmitted through M ₂ ×S×2 subcarriers. Therefore, the obtained M ₂ ×S bit data is transmitted through 2×M ₂ ×S subcarriers, and every two adjacent subcarriers transmit one data bit, and which one to transmit depends on the index bit, for example, the first If the index bit u ₁ is 0, then x ₁ is transmitted on the first subcarrier; if u ₁ is 1, then x ₁ is transmitted on the second subcarrier; if the second index bit u ₂ is 1, then the second data on the 4th subcarrier, and so on. It can be known from the above that the index bits {u _j , j=1, 2, ..., M ₁ /R} satisfy M ₁ /R=M ₂ ×S.

S5. Perform IFFT transformation on the M ₂ ×S×2-bit data obtained in S4 to the time domain to the said After adding the cyclic prefix, add noise through the channel, perform FFT transformation to the frequency domain, and remove the cyclic prefix to obtain Among them, N=2×M ₂ ×S, K represents the number of subcarriers actually having data, is the channel parameter value corresponding to the subcarrier, the is unknown and needs to be obtained by ML method, Represents the bits of layer l at the jth moment, n _j is Gaussian noise with zero mean and variance of δ ² =N ₀ /2.

S6. The ML criterion detects the index bit to obtain the channel parameter, specifically as follows:

S61. Use and Indicates that when the jth bit is 1 and 0, the likelihood ratio information of the index bit is calculated by using the ML method $L\left(u_j\right)=\log \{\underset{x^{\&prime;}\&Element;x}{\underset{u^{\&prime;}=u_j^{\left(1\right)}}{\mathrm{\&Sigma;}}}{e}^{\frac{{\left|\right|\mathrm{the\; y}-h_{u^{\&prime;}}{x}^{\&prime;}\left|\right|}^2}{-{\mathrm{\&delta;}}^2}}/\underset{x^{\&prime;}\&Element;x}{\underset{u^{\&prime;}=u_j^{\left(1\right)}}{\mathrm{\&Sigma;}}}{e}^{\frac{{\left|\right|\mathrm{the\; y}-h_{u^{\&prime;}}{x}^{\&prime;}\left|\right|}^2}{-{\mathrm{\&delta;}}^2}}\},$ Wherein, X represents the set of all transmitted symbols, h _u' represents the channel parameter of the corresponding subcarrier, and the h _u' is determined according to the index bit u';

S62, obtain according to the L(u _j ) described in S61 value, if L(u _j )≥0, then the is the channel parameter value on the 2jth subcarrier, if L(u _j )<0, then the is the channel parameter value on the 2j-1th subcarrier;

S63. According to S62, obtain the channel parameter H={h _j }, where j=M ₂ ×S;

S7. Send the channel parameter information H obtained in S6 to the detection and decoding of the SCM, deinterleave the likelihood ratio information L(u _j ) of the index bits described in S61 and send it to the decoder for decoding. After the decoding is completed, Obtain the detected index bit data, wherein the decoding method corresponds to the encoder at the sending end;

S8, SCM detection, specifically as follows:

S81. In the GA module, Gaussian approximation detection is adopted, and the weighting coefficient is W={w ⁽¹⁾ , w ⁽²⁾ ,..., w ^(l) }, and the received information at the jth moment is $r_j=h_j\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{w}^{\left(l\right)}{c}_j^{\left(l\right)}+{\mathrm{no}}_j;$

S82. Send the acceptance information to the GA module, as described in S3 can get ${\mathrm{\&zeta;}}_j^{\left(l\right)}=r_j-h_j{w}^{\left(l\right)}{c}_j^{\left(l\right)},\mathrm{E.}\left(r_j\right)=h_j\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{w}^{\left(l\right)}\mathrm{E.}\left(c_j^{\left(l\right)}\right),\mathrm{Var}\left(r_j\right)=h_j\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{w}^{\left(l\right)}\mathrm{Var}\left(c_j^{\left(l\right)}\right)+{\mathrm{\&delta;}}^2,$ in, for interference;

S83, initialization, Get the mean and variance of the received signal r _j ;

S84. Get mean of $\mathrm{E.}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)=\mathrm{E.}\left(r_j\right)-h_j{w}^{\left(l\right)}\mathrm{E.}\left(c_j^{\left(l\right)}\right)$ and variance $\mathrm{Var}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)=\mathrm{Var}\left(r_j\right)-\left(h_j{w}^{\left(l\right)}{c}_j^{\left(l\right)}\right);$

S85, obtain the logarithmic likelihood ratio of the lth bit transmitted at the jth moment according to S84 as $L\left(c_j^{\left(l\right)}\right)=\log \left(\frac{\mathrm{PR}(c_j^{\left(l\right)}=+1|r_j)}{\mathrm{PR}(c_j^{\left(l\right)}=-1|r_j)}\right)=\frac{2h_j(r_j-\mathrm{E.}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)}{\mathrm{Var}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)};$

S86. According to S85, obtain the likelihood ratio information L(c) corresponding to the transmitted data of all users in the L layer;

S9, at each layer in the SCM, perform deinterleaving according to L(c) described in S86 to obtain L(a), use Indicates the i-th bit of layer l, the Send it to the despreading module module for despreading to obtain the log likelihood ratio of the first bit of the l layer where S is the spreading length in the transmitter, is the spreading bit;

S10, obtain the described in S9 The extra-bit information after deinterleaving $\mathrm{ext}\left(a_j^{\left(l\right)}\right)={\mathrm{the\; s}}_j^{\left(l\right)}L\left(d_1^{\left(l\right)}\right)-L\left(a_j^{\left(l\right)}\right);$

S11. Interleave the extra-bit information described in S10 to obtain will be described into the GA module again, for The mean and variance of are updated: $\mathrm{E.}\left(c_j^{\left(l\right)}\right)=\tanh \left(\frac{\mathrm{ext}\left(c_j^{\left(l\right)}\right)}{2}\right),\mathrm{Var}\left(c_j^{\left(l\right)}\right)=1-\mathrm{E.}\left(c_j^{\left(l\right)}\right);$

S12. Complete the update of the mean value and variance of all data in the L layer, substitute the mean value and variance of the updated sent data into S8-S9, iterate according to the set number of iterations CH, and obtain the value obtained when the iteration reaches the CHth time available in, Indicates the estimate of the mth bit of the kth layer, CH≥5. The greater the number of iterations, the more accurate the decoding, but the higher the complexity.

S13, the S12 said The index bit data detected in step S7 is sent to the serial-to-parallel converter to obtain the final output bit.

The technical solution of the present invention will be described in detail below in combination with the embodiments and the accompanying drawings.

It is assumed that the number of layers of the SCM is 4, so there are 4 weighting coefficients, W={w ₁ , w ₂ , w ₃ , w ₄ ,}, and the spreading length is 4.

Step 1: Suppose the sequence sent is: {1,1,0,1,0,0,1,1}. The previous bit 1101 is used as an index bit, and the index bit is coded and interleaved, assuming that the coded and interleaved index bit is {1,1,0,1,0,0,1,1,1,1,0,1, 0,0,1,1}.

Step 2: After serial-to-parallel conversion, the following 0011 is mapped to the 4 layers of SCM, the first layer is 0, the second layer is 0, the third layer is 1, and the fourth layer is 1. After BPSK modulation, assuming that the spreading sequences of the four layers are all {-1,1,-1,1}, the data of the first layer becomes {1,-1,1,-1}, the second The layer is also {1,-1,1,-1}, and the third and fourth layers are {-1,1,-1,1}. Assuming that after interleaving is completed, layers 1 to 4, the data of each layer are {1,-1,-1,1}, {1,1,-1,-1}, {1,-1,1,-1 },{-1,-1,1,1}: . Then the data sent at the first moment is x ₁ =1·w ₁ +1·w ₂ +1·w ₃ −1·w ₄ .

Step 3: According to the previous index bits {1,1,0,1,0,0,1,1,1,1,0,1,0,0,1,1}, according to the corresponding one-of-two method, Since the first bit is 1, put x ₁ on the first subcarrier. By analogy, the data is placed on the corresponding subcarriers according to the index bits. Then add noise through the channel.

Step 4: For the detection of index bit soft information, the method of maximum likelihood is adopted, $L\left(u_1\right)=\log \{\underset{x^{\&prime;}\&Element;x}{\underset{u^{\&prime;}=1}{\mathrm{\&Sigma;}}}{e}^{\frac{{\left|\right|\mathrm{the\; y}-h_{u^{\&prime;}}{x}^{\&prime;}\left|\right|}^2}{-{\mathrm{\&delta;}}^2}}/\underset{x^{\&prime;}\&Element;x}{\underset{u^{\&prime;}=0}{\mathrm{\&Sigma;}}}{e}^{\frac{{\left|\right|\mathrm{the\; y}-h_{u^{\&prime;}}{x}^{\&prime;}\left|\right|}^2}{-{\mathrm{\&delta;}}^2}}\}.$ If w ₁ , w ₂ , w ₃ , and w ₄ are different, then X has 2 ⁴ =16 elements. If L(u ₁ )≥0, then h ₁ takes the channel parameter value of the second subcarrier, otherwise, takes the channel parameter value of the first subcarrier, thus the channel parameter information H={h _j ,j= M ₂ ×S}.

Step 5: Bring H={h _j , j=M ₂ ×S} into the bit detection of the SCM. In the first time slot, the received signal is: $r_1=h_1\underset{l=1}{\overset{4}{\mathrm{\&Sigma;}}}{w}^{\left(l\right)}{c}_1^{\left(l\right)}+{\mathrm{no}}_1,r_1=h_1{w}^{\left(1\right)}{c}_1^{\left(1\right)}+{\mathrm{\&zeta;}}_1^{\left(1\right)},$ in, for interference.

${\mathrm{<span\; class="notranslate">\; \&zeta;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; .</span>$

$\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 4</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span>$

$\mathrm{<span\; class="notranslate">\; Var</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 4</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}\mathrm{<span\; class="notranslate">\; Var</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&delta;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}$

During the first iteration, initialization can be done: Therefore, the mean and variance of the received signal r ₁ can be obtained to obtain mean of $\mathrm{E.}\left({\mathrm{\&zeta;}}_1^{\left(1\right)}\right)=\mathrm{E.}\left(r_1\right)-h_1{w}_1^{\left(1\right)}\mathrm{E.}\left(c_1^{\left(1\right)}\right)$ and variance $\mathrm{Var}\left({\mathrm{\&zeta;}}_1^{\left(1\right)}\right)=\mathrm{Var}\left(r_1\right)-\left(h_1{w}_1^{\left(1\right)}{c}_1^{\left(1\right)}\right).$

Through the above derivation, the log likelihood ratio of the first bit of the first layer can be obtained as $L\left(c_1^{\left(1\right)}\right)=\log \left(\frac{\mathrm{PR}(c_1^{\left(1\right)}=+1|r_1)}{\mathrm{PR}(c_1^{\left(1\right)}=-1|r_1)}\right)=\frac{2h_1(r_1-\mathrm{E.}\left({\mathrm{\&zeta;}}_1^{\left(1\right)}\right)}{\mathrm{Var}\left({\mathrm{\&zeta;}}_1^{\left(1\right)}\right)}$

In this way it can be calculated The logarithmic likelihood ratio of the first time slot can be obtained in the same way, and the logarithmic likelihood ratio of the bits of other layers. Since H={h _j ,j=M ₂ ×S}, according to the channel parameters of different time slots, the soft information of each layer of each time slot can be obtained in the same way.

Step 6: Then deinterleave the likelihood ratio information to get Despreading: where 4 is the spreading length in the transmitter, is the i-th spreading bit corresponding to layer l. Get the extrinsic information of the deinterleaved bits $\mathrm{ext}\left(b_j^{\left(l\right)}\right)={\mathrm{the\; s}}_j^{\left(l\right)}L\left(d_1^{\left(l\right)}\right)-L\left(b_j^{\left(l\right)}\right),$ in, Indicates the jth bit of layer l.

Step 7: After interleaving the obtained external information, you can get into the GA module again, for The mean and variance of are updated:

$\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; tanh</span>}<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; ext</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span>$

$\mathrm{<span\; class="notranslate">\; Var</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; .</span>$

In this way, the update of the mean and variance of all data in the L layer is completed. Feedback the updated data to the GA module, and repeat steps 5 and 6.

Step 8: Send {L(u _j ),j=1,2,...,M ₁ /R} to the decoder for decoding after passing through the deinterleaver of index bits, and then all initial index values can be obtained .

Step 9: Perform serial-to-parallel conversion on the index bits and the bits obtained through SCM decoding to obtain the original data.

Claims (3)

1. a method of superposition coding modulation based on SIM-OFDM, is characterized in that, comprises the following steps: S1. Generate M ₁ +M ₂ ×L information bits from a digital source, and pass through a serial-to-parallel converter. The first M ₁ bits are used as index bits, and the last M ₂ ×L bits are used as data bits, and the data bits are sent to another Serial-to-parallel converter; S2. Index bit encoding and interleaving: encode the M ₁ bits described in S1, and interleave the encoded index bits to obtain index data {u _j }, wherein the encoding method is Turbo code, and the encoding rate is R , j=1,2,...,M ₁ /R; S3, perform SCM modulation: after the M ₂ ×L bits described in S1 pass through a serial-to-parallel converter, the data is divided into L layers for transmission, and the data of the first layer is After a spreader with a spreading length of S, interleaving, and BPSK modulation, we get Each layer is weighted by the weighted value {ω ^(l) ,ω ⁽²⁾ ,…,ω ^(L) }, and the weighted data are superimposed to obtain Send to the SIM modulator, waiting to be sent, where m=1,2,...,M ₂ , j=1,2,...,M ₂ ×S; S4, carry out SIM modulation: obtain for S3 SIM modulation is performed by selecting 1 from 2 subcarriers, that is, the M ₂ ×S bit data obtained by S3 is transmitted through M ₂ ×S×2 subcarriers; S5. Perform IFFT transformation on the M ₂ ×S×2-bit data obtained in S4 to the time domain to the said After adding the cyclic prefix, add noise through the channel, perform FFT transformation to the frequency domain, and remove the cyclic prefix to obtain Among them, N=2×M ₂ ×S, K represents the number of subcarriers actually having data, is the channel parameter value corresponding to the subcarrier, the is unknown and needs to be obtained by ML method, Represents the bits of layer l at the jth moment, n _j is Gaussian noise with zero mean and variance of δ ² =N ₀ /2; S6. The ML criterion detects the index bit to obtain the channel parameter, which is as follows: S61. Use and Indicates that when the jth bit is 1 and 0, the likelihood ratio information of the index bit is calculated by using the ML method $L\left(u_j\right)=\log \{\underset{x^{\&prime;}\&Element;x}{\underset{u^{\&prime;}=u_j^{\left(1\right)}}{\mathrm{\&Sigma;}}}{e}^{\frac{{\left|\right|\mathrm{the\; y}-h_{u^{\&prime;}}{x}^{\&prime;}\left|\right|}^2}{-{\mathrm{\&delta;}}^2}}/\underset{x^{\&prime;}\&Element;x}{\underset{u^{\&prime;}=u_j^{\left(1\right)}}{\mathrm{\&Sigma;}}}{e}^{\frac{{\left|\right|\mathrm{the\; y}-h_{u^{\&prime;}}{x}^{\&prime;}\left|\right|}^2}{-{\mathrm{\&delta;}}^2}}\},$ Wherein, X represents the set of all transmitted symbols, h _u' represents the channel parameter of the corresponding subcarrier, and the h _u' is determined according to the index bit u'; S62, obtain according to the L(u _j ) described in S61 value, if L(u _j )≥0, then the is the channel parameter value on the 2jth subcarrier, if L(u _j )<0, then the is the channel parameter value on the 2j-1th subcarrier; S63. According to S62, obtain the channel parameter H={h _j }, where j=M ₂ ×S; S7. Send the channel parameter information H obtained in S6 to the detection and decoding of the SCM, deinterleave the likelihood ratio information L(u _j ) of the index bits described in S61 and send it to the decoder for decoding. After the decoding is completed, Obtain the detected index bit data, wherein the decoding method corresponds to the encoder at the sending end; S8, SCM detection, specifically as follows: S81. In the GA module, Gaussian approximation detection is adopted, and the weighting coefficient is W={w ⁽¹⁾ , w ⁽²⁾ ,..., w ^(l) }, and the received information at the jth moment is $r_j=h_j\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{w}^{\left(l\right)}{c}_j^{\left(l\right)}+{\mathrm{no}}_j;$ S82. Send the received information to the GA module, as described in S3 can get ${\mathrm{\&zeta;}}_j^{\left(l\right)}=r_j-h_j{w}^{\left(l\right)}{c}_j^{\left(l\right)},\mathrm{E.}\left(r_j\right)=h_j\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{w}^{\left(l\right)}\mathrm{E.}\left(c_j^{\left(l\right)}\right),\mathrm{Var}\left(r_j\right)=h_j\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{w}^{\left(l\right)}\mathrm{Var}\left(c_j^{\left(l\right)}\right)+{\mathrm{\&delta;}}^2,$ in, for interference; S83, initialization, Get the mean and variance of the received signal r _j ; S84. Get mean of $\mathrm{E.}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)=\mathrm{E.}\left(r_j\right)-h_j{w}^{\left(l\right)}\mathrm{E.}\left(c_j^{\left(l\right)}\right)$ and variance $\mathrm{Var}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)=\mathrm{Var}\left(r_j\right)-\left(h_j{w}^{\left(l\right)}{c}_j^{\left(l\right)}\right);$ S85, obtain the logarithmic likelihood ratio of the lth bit transmitted at the jth moment according to S84 as $L\left(c_j^{\left(l\right)}\right)=\log \left(\frac{\mathrm{PR}(c_j^{\left(l\right)}=+1|r_j)}{\mathrm{PR}(c_j^{\left(l\right)}=-1|r_j)}\right)=\frac{2h_j(r_j-\mathrm{E.}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)}{\mathrm{Var}\left({\mathrm{\&zeta;}}_j^{\left(l\right)}\right)};$ S86. According to S85, obtain the likelihood ratio information L(c) corresponding to the transmitted data of all users in the L layer; S9, at each layer in the SCM, perform deinterleaving according to L(c) described in S86 to obtain L(a), use Indicates the i-th bit of layer l, the Send it to the despreading module for despreading to obtain the log likelihood ratio of the first bit of the l layer where S is the spreading length in the transmitter, is the spreading bit; S10, obtain the described in S9 The extra-bit information after deinterleaving $\mathrm{ext}\left(a_j^{\left(l\right)}\right)={\mathrm{the\; s}}_j^{\left(l\right)}L\left(d_1^{\left(l\right)}\right)-L\left(a_j^{\left(l\right)}\right);$ S11. Interleave the extra-bit information described in S10 to obtain will be described into the GA module again, for The mean and variance of are updated: $\mathrm{E.}\left(c_j^{\left(l\right)}\right)=\tanh \left(\frac{\mathrm{ext}\left(c_j^{\left(l\right)}\right)}{2}\right),\mathrm{Var}\left(c_j^{\left(l\right)}\right)=1-\mathrm{E.}\left(c_j^{\left(l\right)}\right);$ S12. Complete the update of the mean value and variance of all data in the L layer, substitute the mean value and variance of the updated sent data into S8-S9, iterate according to the set number of iterations CH, and obtain the value obtained when the iteration reaches the CHth time available in, Indicates the estimate of the mth bit of the kth layer; S13, the S12 said The index bit data detected in step S7 is sent to the serial-to-parallel converter to obtain the final output bit. 2. A SIM-OFDM-based superposition coding and modulation method according to claim 1, characterized in that: the initialization in S83 is carried out in the first iteration. 3. A SIM-OFDM-based superposition coding and modulation method according to claim 1, characterized in that: the number of iterations CH≥5 in S12.