A GCE-BEM iterative channel estimation method, system, device and medium based on sub-block transmission

Abstract

A GCE-BEM iterative channel estimation method, system, device and medium based on sub-block transmission, the method comprising: using a sub-block-based TDKD pilot cluster as a transmission frame structure, encoding bit information to form a complex signal, completing a framing operation, modulating the transmission frame signal to a high-frequency carrier signal for transmission, using the signal at the pilot position at the receiving end, using an LS channel estimation method to obtain initial channel information, obtaining soft decision information through a linear frequency domain equalization module and a soft decision module, using an LMMSE channel estimation method for known symbols to obtain the entire channel matrix, sending the result of the second channel estimation to the frequency domain equalization module, obtaining soft decision information of the equalized output, and using it for the known symbols of the next iterative channel estimation; the system, device and medium are used to implement a GCE-BEM iterative channel estimation method based on sub-block transmission; the present invention effectively improves the accuracy of channel estimation, reduces the complexity of channel estimation, and improves the bit error rate performance of the system.

Description

Technical Field

[0001] The present invention belongs to the technical field of dual-selection channel wireless communication, and in particular relates to a GCE-BEM iterative channel estimation method, system, device and medium based on sub-block transmission. Background Art

[0002] For broadband communication systems with high data rate transmission, when the bandwidth becomes larger, the sampling period will be smaller than the delay spread of the channel, especially in multipath transmission, which leads to frequency selective fading. The relative mobility between the transmitter and the receiver introduces Doppler spread, which will cause time variation of the channel, especially at high speeds. The corresponding fading is called time selectivity. A channel where both frequency selective fading and time selective fading exist is called a doubly selective channel (DSC). The doubly selective channel model generally adopts the wide-sense stationary-uncorrelated scattering model (WSSUS). In block transmission systems, frequency selective fading and time selective fading will cause inter-code interference and inter-carrier interference, which greatly reduces the system performance. In order to mitigate these fading effects, it becomes very important to perform reliable channel estimation on doubly selective channels.

[0003] Block transmission techniques provide efficient and computationally affordable solutions. Their advantage is that inter-block interference can be eliminated by adding a cyclic prefix (CP) at the top of the block or adding zero padding (ZP) at the end of the block, such as Orthogonal Frequency Division Multiplexing (OFDM) technology. In a block transmission system, data is transmitted in blocks, which facilitates channel equalization and symbol detection. For block transmission with coherent detection, channel estimation will play an important role, especially in dual-selection channels.

[0004] In order to track channel changes, Dong et al. gave the optimal placement of training symbols for block transmission systems in time-varying flat-fading channels in the literature (Dong, M., Tong, L., & Sadler, BM (2004). Optimal insertion of pilot symbols for transmissions over time-varying flat-fading channels. IEEE Transactions on Signal Processing, 52(5), 1403–1418.), and Ma et al. gave the optimal placement of training symbols over doubly selective channels in the literature (Ma, X., Giannakis, GB, & Ohno, S. (2003). Optimal training for block transmissions over doubly selective wireless channels. IEEE Transactions on Signal Processing, 51(5), 1351–1366.). Rousseaux et al. (Rousseaux, O., Leus, G., & Moonen, M. (2006). Estimation and equalization of doubly selective channels using known symbol padding. IEEE Transactions on Signal Processing, 54 (3), 979–990.) studied the channel estimation and equalization problems based on block transmission in dual-selective channels using known symbol padding transmission technology. By using BEM-based methods, many scholars have proposed different channel estimation and pilot designs. Based on complex exponential BEM, Kannu and Schniter (Kannu, AP, & Schniter, P. (2008). "Design and analysis of MMSE pilot-aided cyclic-prefixed block transmissions for doubly selective channels". IEEE Trans. Signal Process., 56 (3), 1148–1160.) gave the design and analysis of pilot-aided transmission of block transmission system on single-antenna dual-selective channels.Tugnait and He in the literature (Tugnait, JK, & He, S. (2007). Doubly-selective channel estimation using data-dependent superimposed training and exponential basis models. IEEE Transactions on Wireless Communications, 6 (11), 3877–3883.) proposed a channel estimation method based on windowing technology, as well as a windowed least squares estimator and optimal pilot design for doubly selective channels based on oversampled BEM for doubly selective channels. For different BEMs, Whitworth et al. gave the optimization of training sequences and BEM parameters for channel estimation of doubly-selective channels in block transmission systems in the literature (Whitworth, T., Ghogho, M., & McLernon, D. (2009). Optimized training and basis expansion model parameters for doubly-selective channel estimation. IEEE Transactions on Wireless Communications, 8(3), 1490–1498.).

[0005] These research advances have greatly enriched the BEM-based channel estimation methods for dual-selection channels in block transmission systems and improved system performance. However, the above literature basically considers channel estimation within the entire block period. When the block length N is large or the maximum normalized Doppler frequency f nom When it is larger, the number of BEM coefficients required will become more, which greatly increases the computational complexity of the estimation. Especially in the case of fast fading, the CE-BEM model cannot reflect the real channel well and the error of channel estimation will increase. When oversampled BEM (GCE-BEM) is used to model the channel, the number of basis coefficients will become larger, so it is necessary to invert the high-dimensional matrix to perform linear minimum mean square error (LMMSE) estimation or equalization, and the calculation will be more complicated. Summary of the invention

[0006] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to propose a GCE-BEM iterative channel estimation method, system, device and medium based on sub-block transmission, by dividing a long data block into several short data sub-blocks, and performing channel estimation on each sub-block respectively, and then using these estimated values ​​to reconstruct the real channel, and iteratively estimate through the equalized soft symbol information, continuously updating the channel estimation information, effectively improving the accuracy of channel estimation, reducing the complexity of channel estimation, and improving the bit error rate performance of the system.

[0007] In order to achieve the above object, the technical solution adopted by the present invention is:

[0008] A GCE-BEM iterative channel estimation method based on sub-block transmission includes the following steps:

[0009] Step 1: Use sub-block-based time domain Kronecker TDKD pilot cluster as the transmission frame structure. Each single-carrier frequency domain equalization SC-FDE block has M sub-blocks. Each sub-block contains P TDKD structure pilot clusters. Each pilot cluster consists of a data part and a pilot part. Only the middle position of the pilot part is a non-zero pilot symbol, and both sides are filled with zeros of a certain length as a protection interval. The pilot symbol and position are known to the transmitter and the receiver.

[0010] Step 2: Encode the bit information at the transmitting end, and then modulate the encoded information to different constellation points through the binary baseband signal to form the complex signal required for transmission. Divide the data into blocks according to the frame structure of step 1, insert the local pilot, complete the framing operation, form a transmission frame signal, and finally modulate the signal to the high-frequency carrier signal and send it out through the antenna;

[0011] Step 3: The receiving antenna performs timing synchronization according to the signal received in step 2 to obtain the received signal required for channel estimation, selects the signal at the pilot position, and uses the least squares LS channel estimation method based on GCE-BEM to perform the first channel estimation to obtain the basis coefficients corresponding to the sub-blocks, and then estimates the basis coefficients of the entire data block according to the oversampled complex exponential basis function, and obtains the initial channel information of the entire data block through the basis coefficients;

[0012] Step 4, according to the received signal and the estimated channel information in step 3, the equalized soft decision information is obtained through a linear frequency domain equalization module and a soft decision module for the known symbols of the next channel estimation;

[0013] Step 5: According to the known symbols obtained in step 4, the second channel estimation adopts the minimum mean square error LMMSE channel estimation method based on GCE-BEM to obtain the basis coefficients of the sub-blocks in the SC-FDE block, and then obtain the basis coefficients of the entire data block, and finally fit the channel matrix H of the entire data block through the oversampled complex exponential basis function;

[0014] Step 6, according to the result of the second channel estimation obtained in step 5, it is sent to the frequency domain equalization module to obtain the soft decision information of the equalization output, which is used for the known symbols of the next iterative channel estimation; when the number of iterations is over, the channel information obtained from the last iteration is used to eliminate the inter-code interference of the received signal and the frequency domain equalization module is performed in sequence to obtain the final required bit information.

[0015] In step 1, the sub-block-based time domain Kronecker TDKD pilot cluster is used as the transmission frame structure, and the specific process is as follows:

[0016] Step 1.1: For each single carrier frequency domain equalization SC-FDE data block, the length is N = N s +N t , where N s is the information sequence length, N t is the training sequence length, which is divided into M equal-length sub-blocks, with each sub-block having a length of N0 = N / M, where the choice of M depends on the maximum Doppler frequency and the length of the block;

[0017] Step 1.2: For each sub-block, the information symbol and training sequence are placed according to the time domain Kronecker TDKD pilot placement method. First, each sub-block is divided into P pilot clusters, where m∈{1,…,M} and p∈{1,…,P} are the indices of the mth sub-block and the pth small sub-block respectively. For each pilot cluster, the length of the information symbol is N s,b =N s / (MP), the length of the training sequence is N t,b =N t / (MP)=2L+1;

[0018] Step 1.3: Add L guard zeros to the end of each sub-block by setting the last L elements of each training sequence to zero, where L is the number of multipath channels;

[0019] When the single carrier frequency domain equalization SC-FDE data block x reaches the receiving end through the dual-selection channel, the nth received symbol can be expressed as:

[0020]

[0021] In the formula, z(n) has a mean of 0 and a variance of Additive white Gaussian noise AWGN;

[0022] The above formula is written in matrix form:

[0023] y=Hx+z

[0024] In the formula, z=[z(0),z(1),…,z(N-1)] T , y=[y(0),y(1),…,y(N-1)] T , H is an N×N dimensional matrix, as shown below:

[0025]

[0026] Step 1.3: Using GCE-BEM, the channel matrix H can be expressed as

[0027]

[0028] In the formula, C q is a Toeplitz matrix, as shown below, and

[0029]

[0030] In step 3, the least squares LS channel estimation method based on GCE-BEM is used to perform the first channel estimation, and the specific process is as follows:

[0031] Step 3.1: When the number of iterations i = 0, the data corresponding to the pilot position of the received signal and the basis function matrix are known, and the basis coefficients of each path are obtained by the following formula:

[0032]

[0033] In the formula, y t,l =[y(n 0,l ),…,y(n P-1,l )] T ,n p,l =n p +l,l=0,1,…L;

[0034] Step 3.2: The basis function matrix corresponding to the pilot position is expressed as:

[0035]

[0036] The basis coefficient matrix of the sub-block is obtained from the above And the base coefficient correlation matrix Through the GCE-BEM channel model, the initial channel information of the entire data block can be obtained.

[0037] The step 4 of eliminating inter-symbol interference of the received signal and obtaining the equalized soft decision information for the known symbol of the next channel estimation is specifically: performing soft decision on the equalized output to obtain the symbol soft information for the next iteration, then the soft information corresponding to the nth symbol in the pth small sub-block of the mth sub-block is obtained by the following formula:

[0038]

[0039] In the formula, and Represents the log-likelihood ratio information of the equalized output.

[0040] The step 5 adopts the LMMSE channel estimation method based on GCE-BEM, and the specific process is as follows:

[0041] Step 5.1: When the number of iterations i>0, the symbol soft information obtained in the previous iteration is inserted into the local pilot as a known symbol to perform LMMSE channel estimation based on GCE-BEM;

[0042] Step 5.2: The basis coefficient of the mth sub-block is obtained by the following formula:

[0043]

[0044] In the formula, is the basis coefficient correlation matrix calculated from the last iteration, and y(m) is the received signal.

[0045]

[0046] U q,p (m) is U q The submatrix corresponding to the training symbol in the p-th small sub-block in (m);

[0047]

[0048] Divided into two parts p,n (m) and t p,n (m) is the pilot sequence inserted again by the receiving end, It consists of the symbol soft information output by the equalization module in the previous iteration;

[0049] Step 5.3: The channel vector of the lth path of the mth sub-block is obtained by the following formula:

[0050]

[0051] In the formula, is the basis coefficient vector of the lth path, is the basis function matrix corresponding to the mth sub-block, with a dimension of N0×(Q0+1), where

[0052] Step 5.4: Obtain the basis coefficients of the entire data block through the M channel vectors obtained according to step 5.3, and the vector dimension is N×1:

[0053]

[0054] The basis coefficient vector corresponding to the lth path of the entire data block can be obtained by the following formula:

[0055]

[0056] In the formula, U = [b 0 ,b 1 ,…,b Q ] is the basis function matrix corresponding to a SC-FDE data block period, and its dimension is N×(Q+1), where

[0057] Step 5.5: Based on Get the estimated base coefficient Then we further get the estimated basis coefficient matrix

[0058] The estimated channel matrix H is obtained by the following formula:

[0059]

[0060] In step 6, iteration starts from the second channel estimation, and the channel estimation method in the iteration adopts the minimum mean square error LMMSE channel estimation method based on GCE-BEM to iteratively update the channel information.

[0061] A GCE-BEM iterative channel estimation system based on sub-block transmission, specifically comprising:

[0062] The channel coding module is used to encode the original bit information and increase the reliability of the information during transmission by adding redundancy;

[0063] A constellation mapping module is used to modulate the binary baseband signal to different constellation points to form a complex signal required for transmission;

[0064] The framing module is used to form a specific frame structure required for sending, which is convenient for the receiving end to process;

[0065] The RF module is used to modulate the transmission signal onto a high-frequency carrier signal and send it out through the antenna;

[0066] A least squares LS channel estimation module is used to provide initial channel information;

[0067] Minimum mean square error LMMSE channel estimation module, used to provide iteratively updated channel information;

[0068] A frequency domain equalization module is used to eliminate inter-symbol interference of the received signal through the output of the channel estimation module;

[0069] A soft decision module, used to provide an input signal to a channel estimation module based on GCE-BEM and soft information input by a decoding module;

[0070] The channel decoding module is used to extract the final useful bit information from the soft information.

[0071] A GCE-BEM iterative channel estimation device based on sub-block transmission, comprising:

[0072] Memory: used for storing a computer program for implementing a GCE-BEM iterative channel estimation method based on sub-block transmission;

[0073] Processor: used to implement the GCE-BEM iterative channel estimation method based on sub-block transmission when executing the computer program.

[0074] A computer-readable storage medium comprising:

[0075] The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, a GCE-BEM iterative channel estimation method based on sub-block transmission can be implemented.

[0076] Compared with the prior art, the beneficial effects of the present invention are:

[0077] (1) In the SC-FDE system, the present invention combines the GCE-BEM-based channel estimation with the equalized soft decision information. The equalized soft decision information is used as feedback information, which not only eliminates inter-symbol interference (ISI) and inter-frequency interference (IFI), but also is used together with the local training sequence as a known sequence to update the channel state information. By iteratively updating the channel estimation information, the SC-FDE system in the dual-selection channel has a higher channel estimation performance.

[0078] (2) The present invention uses the pilot sequence to adopt the LS channel estimation method based on GCE-BEM to obtain the initial channel information, and then enters the linear frequency domain equalization module to obtain the soft decision information of the initial symbol; in the next iteration, the symbol soft decision information obtained in the previous iteration is inserted into the LMMSE channel estimation module based on GCE-BEM of the pilot feedback as a known sequence, and the channel estimation value is iteratively updated. At the same time, the LMMSE channel estimation adopted needs to update the statistical characteristic information of the channel according to the information of the previous iteration, so as to improve the channel estimation performance through continuous iteration.

[0079] (3) The present invention performs block processing on the SC-FDE data block again, and uses fewer basis coefficients, thereby reducing the complexity of the LMMSE channel estimation based on GCE-BEM during the iteration process, and improving the system performance.

[0080] (4) Step 1 of the present invention adopts a sub-block-based time-domain Kronecker TDKD pilot cluster as a transmission frame structure. The time-domain Kronecker TDKD pilot cluster structure can decouple information symbols from training symbols, eliminate inter-block interference, and is better suitable for dual-selection channel communication. It divides the data blocks in the data frame into sub-blocks, which can reduce the number of basis coefficients that need to be estimated in channel estimation and has the characteristics of low complexity.

[0081] (5) Step 3 of the present invention uses LS estimation to perform the first channel estimation, which can provide initial channel information for subsequent iterations and has the characteristics of simple implementation and low complexity.

[0082] (6) In step 4 of the present invention, frequency domain equalization is used and the frequency domain equalization output is subjected to soft decision operation. Compared with time domain equalization, frequency domain equalization can reduce the complexity of the equalization algorithm and has the characteristics of low complexity. Soft decision can retain all information of the received signal and has the characteristics of improving channel estimation and decoding performance.

[0083] (7) Step 5 of the present invention uses LMMSE estimation to perform iterative channel estimation. Compared with the first LS channel estimation, the channel information can be estimated more accurately due to the consideration of noise influence. Compared with the existing LMMSE channel estimation, the complexity of the algorithm can be effectively reduced due to the use of a sub-block-based transmission method. The method has the characteristics of low complexity and good performance.

[0084] (8) Step 6 of the present invention adopts an iterative channel estimation processing method, which can effectively improve the accuracy of the system's channel estimation through multiple iterations, and has the characteristic of improving performance.

[0085] In summary, compared with the prior art, the present invention can improve the accuracy of channel estimation, reduce calculation complexity, and has the characteristics of higher performance and lower complexity. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1 It is a flow chart of the method provided by an embodiment of the present invention.

[0087] Figure 2 It is a schematic diagram of the frame structure of the SC-FDE system provided in an embodiment of the present invention.

[0088] Figure 3 It is a flow chart of an iterative channel estimation system based on GCE-BEM provided in an embodiment of the present invention.

[0089] Figure 4 It is a schematic diagram of BER performance simulation comparison provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0090] The present invention is further described in detail below with reference to the accompanying drawings.

[0091] like Figure 1 As shown, the method provided by the embodiment of the present invention includes the following steps:

[0092] Step 1: Use sub-block-based time domain Kronecker TDKD pilot cluster as the transmission frame structure. Each single-carrier frequency domain equalization SC-FDE block has M sub-blocks. Each sub-block contains P TDKD structure pilot clusters. Each pilot cluster consists of a data part and a pilot part. Only the middle position of the pilot part is a non-zero pilot symbol, and both sides are filled with zeros of a certain length as a protection interval. The pilot symbol and position are known to the transmitter and the receiver.

[0093] Step 2: Encode the bit information at the transmitting end, and then modulate the encoded information to different constellation points through the binary baseband signal to form the complex signal required for transmission. Divide the data into blocks according to the frame structure of step 1, insert the local pilot, complete the framing operation, form a transmission frame signal, and finally modulate the signal to the high-frequency carrier signal and send it out through the antenna;

[0094] Step 3: The receiving antenna performs timing synchronization according to the signal received in step 2 to obtain the received signal required for channel estimation, selects the signal at the pilot position, and uses the least squares LS channel estimation method based on GCE-BEM to perform the first channel estimation to obtain the basis coefficients corresponding to the sub-blocks, and then estimates the basis coefficients of the entire data block according to the oversampled complex exponential basis function, and obtains the initial channel information of the entire data block through the basis coefficients;

[0095] Step 4, according to the received signal and the estimated channel information in step 3, the equalized soft decision information is obtained through a linear frequency domain equalization module and a soft decision module for the known symbols of the next channel estimation;

[0096] Step 5: According to the known symbols obtained in step 4, the second channel estimation adopts the minimum mean square error LMMSE channel estimation method based on GCE-BEM to obtain the basis coefficients of the sub-blocks in the SC-FDE block, and then obtain the basis coefficients of the entire data block, and finally fit the channel matrix H of the entire data block through the oversampled complex exponential basis function;

[0097] Step 6, according to the result of the second channel estimation obtained in step 5, it is sent to the frequency domain equalization module to obtain the soft decision information of the equalization output, which is used for the known symbols of the next iterative channel estimation; when the number of iterations is over, the channel information obtained from the last iteration is used to eliminate the inter-code interference of the received signal and the frequency domain equalization module is performed in sequence to obtain the final required bit information.

[0098] like Figure 2 As shown in the figure, the frame structure diagram of the SC-FDE system, for each SC-FDE data block, the length is N = N s +N t , where N s is the information sequence length, N t is the length of the training sequence. We first divide it into M equal-length sub-blocks, each of which has a length of N0 = N / M. The choice of M depends on the maximum Doppler frequency and the length of the block.

[0099] For each sub-block, the information symbol and training sequence are placed according to the TDKD pilot placement method. First, each sub-block is divided into P pilot clusters (consisting of their own data part and pilot part), where m∈{1,…,M} and p∈{1,…,P} are the indices of the mth sub-block and the pth small sub-block respectively. For each pilot cluster, the length of the information symbol is N s,b =N s / (MP), the length of the training sequence is N t,b =N t / (MP) = 2L + 1. In order to eliminate inter-block interference (IBI), L guard zeros need to be added to the end of each (sub)block, which is achieved by setting the last L elements of each training sequence to zero, where L is the number of multipath channels.

[0100] like Figure 3 As shown, the flowchart of the iterative channel estimation system based on GCE-BEM.

[0101] The original bit information is generated into a complex signal through the channel coding and constellation mapping module at the transmitting end, and then transmitted through Figure 2 The frame structure is framed, and further sent out through the radio frequency module via the antenna.

[0102] When an SC-FDE data block x reaches the receiving end through the dual-selection channel, an effective received signal is obtained through ideal timing synchronization, where the nth received symbol can be expressed as:

[0103]

[0104] In the formula, z(n) has a mean of 0 and a variance of Additive White Gaussian Noise (AWGN).

[0105] The above formula is written in matrix form:

[0106] y=Hx+z

[0107] In the formula, z=[z(0),z(1),…,z(N-1)] T , y=[y(0),y(1),…,y(N-1)] T , H is an N×N dimensional matrix, as shown below:

[0108]

[0109] First, using GCE-BEM, the channel matrix H can be expressed as

[0110]

[0111] In the formula, C q is a Toeplitz matrix, as shown below, and

[0112]

[0113] For the iterative channel estimation part, the following iterations are performed:

[0114] First iteration channel estimation:

[0115] LS channel estimation does not require any channel statistics information, so when the number of iterations i = 0, the receiver uses LS channel estimation for processing. Given the data corresponding to the pilot position of the received signal and the basis function matrix, we can obtain the basis coefficients of each path by the following formula:

[0116]

[0117] In the formula, y t,l =[y(n 0,l ),…,y(n P-1,l )] T ,n p,l =n p +l,l=0,1,…L.

[0118] The basis function matrix corresponding to the pilot position is expressed as:

[0119]

[0120] Therefore, the basis coefficient matrix of the SC-FDE data block is obtained And the base coefficient correlation matrix As the channel statistical characteristics required for the next LMMSE iteration estimation.

[0121] The initial channel estimation information can be obtained through the GCE-BEM channel model.

[0122] Next, the linear frequency domain equalizer is used for equalization processing, and the equalization output is soft-determined to obtain the symbol soft information for the next iteration. Then, the soft information corresponding to the nth symbol in the pth small sub-block of the mth sub-block is obtained by the following formula:

[0123]

[0124] In the formula, and Represents the log-likelihood ratio information of the equalized output.

[0125] The i-th iteration channel estimation:

[0126] Next, when the number of iterations i>0, we insert the symbol soft information obtained in the previous iteration into the local pilot as a known symbol to perform LMMSE channel estimation based on GCE-BEM.

[0127] We also use the GCE-BEM-based channel estimation method, so the basis coefficient of the mth small sub-block is obtained by the following formula:

[0128]

[0129] In the formula, is the basis coefficient correlation matrix calculated from the last iteration, and y(m) is the received signal.

[0130]

[0131] U q,p (m) is U q The submatrix corresponding to the training symbol in the p-th small sub-block in (m).

[0132]

[0133] Divided into two parts p,n (m) and t p,n (m) is the pilot sequence inserted again by the receiving end, It consists of the soft information of the symbols output by the equalization module in the previous iteration.

[0134] By the following formula, we can estimate the channel vector of the lth path of the mth sub-block:

[0135]

[0136] In the formula, is the basis coefficient vector of the lth path, is the basis function matrix corresponding to the mth sub-block, with a dimension of N0×(Q0+1), where

[0137] Furthermore, through the M channel vectors estimated above, we reconstruct the channel vector of the entire data block, and the vector dimension is N×1:

[0138]

[0139] Then, the basis coefficient vector corresponding to the lth path of the entire data block can be obtained by the following formula:

[0140]

[0141] In the formula, U = [b 0 ,b 1 ,…,b Q ] is the basis function matrix corresponding to a SC-FDE data block period, and its dimension is N×(Q+1), where

[0142] based on , we can get the estimated base coefficient from the above formula Then we further get the estimated basis coefficient matrix

[0143] Finally, the estimated channel matrix H is obtained by the following formula:

[0144]

[0145] When the iterative channel estimation value is updated, the soft decision information output after equalization can be more accurate. After inserting the pilot, it is fed back to the channel estimation part as a new known pilot sequence, and LMMSE channel estimation is performed to obtain a new channel update. This cycle is iterated until the iteration termination condition is met.

[0146] The application principle of the present invention is further described below in conjunction with specific embodiments.

[0147] In the embodiment, the SC-FDE system adopts a single-transmit and single-receive system. The channel model adopts a dual-selection channel in a time-varying underwater acoustic channel scenario, and the carrier frequency is f c =12kHz, the propagation speed of sound waves in the ocean is about 1500m / s, and the symbol period is T s =250μs. Normalized Doppler frequency shift f nom Characterizes how fast the channel changes, f nom =0.005 represents a slow time-varying channel with a speed of 10 km / h, f nom =0.033 represents a fast time-varying channel with a speed of 60 km / h. The number of iterations is set to 3, and iterative channel estimation is performed on the received signal. The bit error rate (BER) of the method of the present invention is simulated for different signal-to-noise ratios (SNRs), and the system performance is compared under traditional non-iterative channel estimation (hereinafter referred to as the existing method) and ideal channel conditions.

[0148] As shown in the following table, a schematic diagram of the complexity comparison of LMMSE channel estimation methods based on GCE-BEM is shown.

[0149] Complexity analysis of LMMSE channel estimation method based on GCE-BEM in iterative process

[0150]

[0151] The complexity of the improved LMMSE channel estimation method based on GCE-BEM in the iterative process of the present invention is compared with the complexity of the existing LMMSE channel estimation method based on BEM. For long data blocks, without losing the channel estimation performance, compared with directly estimating the channel of the entire data block by BEM, we divide the data block into sub-blocks and use fewer basis coefficients, thereby reducing the complexity of the LMMSE channel estimation based on GCE-BEM in the iterative process.

[0152] Figure 4 The figure is a comparison diagram of the SNR performance of the iterative channel estimation method based on sub-block GCE-BEM in the SC-FDE system using the method of the present invention and the existing channel estimation method, wherein the horizontal axis represents the signal-to-noise ratio SNR, and the vertical axis represents the bit error probability BER. For the slow time-varying channel, the present invention can obtain better performance than the existing method. As the degree of channel time variation increases, the BER performance of the method of the present invention decreases slightly, but it is still better than the existing method. The BER performance of the traditional method decreases significantly under the fast time-varying channel. Therefore, the method proposed in the present invention uses iterative channel estimation to effectively improve the bit error rate performance of the SC-FDE system under the fading conditions of the time-varying channel.

[0153] It can be seen from the embodiments that the channel estimation method of the present invention can significantly improve the accuracy of channel estimation of the SC-FDE system. By dividing the data block into sub-blocks, the present invention can also effectively reduce the complexity of LMMSE channel estimation based on GCE-BEM during the iteration process.

[0154] A signal detection device of a GCE-BEM iterative channel estimation system based on sub-block transmission, comprising:

[0155] Memory: used for storing a computer program for implementing a GCE-BEM iterative channel estimation method based on sub-block transmission;

[0156] A processor is used to implement the GCE-BEM iterative channel estimation method based on sub-block transmission when executing the computer program.

[0157] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc. The processor is the control center of the GCE-BEM iterative channel estimation device based on sub-block transmission, and uses various interfaces and lines to connect various parts of the entire GCE-BEM iterative channel estimation device based on sub-block transmission.

[0158] When the processor executes the computer program, the steps of the signal detection method of the GCE-BEM iterative channel estimation system based on sub-block transmission are implemented, for example: a sub-block-based TDKD pilot cluster is used as a transmission frame structure, each SC-FDE block has M sub-blocks, each sub-block contains P TDKD structure pilot clusters, the pilot symbol and the position are known to the transmitting end and the receiving end, the bit information is encoded to form a complex signal required for transmission, the framing operation is completed, the transmission frame signal is formed, and the high-frequency carrier signal is modulated and sent. The receiving end uses the signal of the pilot position and adopts the GCE-BE The LS channel estimation method of M is used to obtain the initial channel information of the entire data block. The received signal and the estimated channel information are passed through a linear frequency domain equalization module and a soft decision module to obtain the equalized soft decision information for the known symbols of the next channel estimation. The LMMSE channel estimation method based on GCE-BEM is used for the known symbols to obtain the channel matrix of the entire data block. The result of the second channel estimation is sent to the frequency domain equalization module to obtain the soft decision information of the equalization output, which is used for the known symbols of the next iterative channel estimation. A GCE-BEM iterative channel estimation method based on sub-block transmission is implemented.

[0159] Alternatively, the processor implements the functions of each module in the above system when executing the computer program, for example: a channel coding module, which is used to encode the original bit information and enhance the reliability of the information during transmission by increasing redundancy; a constellation mapping module, which is used to modulate the binary baseband signal to different constellation points to form a complex signal required for transmission; a framing module, which is used to form a specific frame structure required for transmission to facilitate processing at the receiving end; a radio frequency module, which is used to modulate the transmission signal to a high-frequency carrier signal and send it out through an antenna; a least squares LS channel estimation module, which is used to provide initial channel information; a minimum mean square error LMMSE channel estimation module, which is used to provide iteratively updated channel information; a frequency domain equalization module, which is used to eliminate inter-code interference of the received signal through the output of the channel estimation module; a soft decision module, which is used to provide an input signal of a channel estimation module based on GCE-BEM and soft information input by a decoding module; a channel decoding module, which is used to extract the final useful bit information from the soft information; and an output to obtain the signal detection result of the GCE-BEM iterative channel estimation system based on sub-block transmission.

[0160] Exemplarily, the computer program may be divided into one or more modules / units, the one or more modules / units are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of completing preset functions, and the instruction segments are used to describe the execution process of the computer program in the device for signal detection of the GCE-BEM iterative channel estimation system based on sub-block transmission. For example, the computer program can be divided into a channel coding module, which is used to encode the original bit information and enhance the reliability of the information during transmission by increasing redundancy; a constellation mapping module, which is used to modulate the binary baseband signal to different constellation points to form a complex signal required for transmission; a framing module, which is used to form a specific frame structure required for transmission to facilitate processing at the receiving end; a radio frequency module, which is used to modulate the transmission signal to a high-frequency carrier signal and send it out through an antenna; a least squares LS channel estimation module, which is used to provide initial channel information; a minimum mean square error LMMSE channel estimation module, which is used to provide iteratively updated channel information; a frequency domain equalization module, which is used to eliminate inter-code interference of the received signal through the output of the channel estimation module; a soft decision module, which is used to provide an input signal of a channel estimation module based on GCE-BEM, and soft information input by a decoding module; a channel decoding module, which is used to extract the final useful bit information from the soft information and output the signal detection result of the GCE-BEM iterative channel estimation system based on sub-block transmission.

[0161] The device of the signal detection system of the GCE-BEM iterative channel estimation system based on sub-block transmission can be a computing device such as a desktop computer, a notebook, a PDA and a cloud server. The signal detection device of the GCE-BEM iterative channel estimation system based on sub-block transmission may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that the above is an example of a device for signal detection of a GCE-BEM iterative channel estimation system based on sub-block transmission, and does not constitute a limitation on a device for signal detection of a GCE-BEM iterative channel estimation system based on sub-block transmission, and may include more components than the above, or a combination of certain components, or different components. For example, the device for signal detection of a GCE-BEM iterative channel estimation system based on sub-block transmission may also include input and output devices, network access devices, buses, etc.

[0162] The memory can be used to store the computer program and / or module, and the processor implements various functions of the signal detection device of the GCE-BEM iterative channel estimation system based on sub-block transmission by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory.

[0163] The memory may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (SmartMediaCard, SMC), a secure digital (SecureDigital, SD) card, a flash card (FlashCard), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0164] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the GCE-BEM iterative channel estimation method based on sub-block transmission are implemented.

[0165] The module / unit integrated in the signal detection system of the GCE-BEM iterative channel estimation system based on sub-block transmission can be stored in a computer-readable storage medium if it is implemented in the form of a software functional unit and sold or used as an independent product.

[0166] The present invention implements all or part of the processes in the above-mentioned GCE-BEM iterative channel estimation method based on sub-block transmission, and can also be completed by instructing related hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and when the computer program is executed by a processor, the steps of the above-mentioned GCE-BEM iterative channel estimation method based on sub-block transmission can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or preset intermediate form, etc.

[0167] The computer-readable storage medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0168] It should be noted that the content contained in the computer-readable storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable storage media do not include electrical carrier signals and telecommunication signals.

[0169] It should be noted that the embodiments of the present invention can be implemented by hardware, software or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware.

[0170] It will be appreciated by those skilled in the art that the above-mentioned devices and methods may be implemented using computer executable instructions and / or contained in processor control codes, such as providing such codes on a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention may be implemented by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., may also be implemented by software executed by various types of processors, or may be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

Claims

1. A GCE-BEM iterative channel estimation method based on sub-block transmission, characterized in that: The following steps are involved: Step 1: Use sub-block-based time domain Kronecker TDKD pilot cluster as the transmission frame structure. Each single-carrier frequency domain equalization SC-FDE block has M sub-blocks. Each sub-block contains P TDKD structure pilot clusters. Each pilot cluster consists of a data part and a pilot part. Only the middle position of the pilot part is a non-zero pilot symbol, and both sides are filled with zeros of a certain length as a protection interval. The pilot symbol and position are known to the transmitter and the receiver. Step 2: Encode the bit information at the transmitting end, and then modulate the encoded information to different constellation points through the binary baseband signal to form the complex signal required for transmission. Divide the data into blocks according to the frame structure of step 1, insert the local pilot, complete the framing operation, form a transmission frame signal, and finally modulate the signal to the high-frequency carrier signal and send it out through the antenna; Step 3: The receiving antenna performs timing synchronization according to the signal received in step 2 to obtain the received signal required for channel estimation, selects the signal at the pilot position, and uses the least squares LS channel estimation method based on GCE-BEM to perform the first channel estimation to obtain the basis coefficients corresponding to the sub-blocks, and then estimates the basis coefficients of the entire data block according to the oversampled complex exponential basis function, and obtains the initial channel information of the entire data block through the basis coefficients; The first channel estimation is performed using the least squares LS channel estimation method based on GCE-BEM. The specific process is as follows: Step 3.1: When the number of iterations i = 0, the data corresponding to the pilot position of the received signal and the basis function matrix are known, and the basis coefficients of each path are obtained by the following formula: where y t,l = [y(n 0,l ), …, y(n P-1,l )] T , n p,l = n p + l, l = 0, 1, … L; Step 3.2: The basis function matrix corresponding to the pilot position is expressed as: The basis coefficient matrix of the sub-block is obtained from the above And the base coefficient correlation matrix Through the GCE-BEM channel model, the initial channel information of the entire data block can be obtained; Step 4, according to the received signal and the estimated channel information in step 3, the equalized soft decision information is obtained through a linear frequency domain equalization module and a soft decision module for the known symbols of the next channel estimation; Step 5: According to the known symbols obtained in step 4, the second channel estimation adopts the minimum mean square error LMMSE channel estimation method based on GCE-BEM to obtain the basis coefficients of the sub-blocks in the SC-FDE block, and then obtain the basis coefficients of the entire data block, and finally fit the channel matrix H of the entire data block through the oversampled complex exponential basis function; The LMMSE channel estimation method based on GCE-BEM is adopted. The specific process is as follows: Step 5.1: When the number of iterations i>0, the symbol soft information obtained in the previous iteration is inserted into the local pilot as a known symbol to perform LMMSE channel estimation based on GCE-BEM; Step 5.2: The basis coefficient of the mth sub-block is obtained by the following formula: In the formula, is the basis coefficient correlation matrix calculated from the last iteration, and y(m) is the received signal; U q,p (m) is U q The submatrix corresponding to the training symbol in the p-th small sub-block in (m); Divided into two parts p,n (m) and t p,n (m) is the pilot sequence inserted again by the receiving end, It consists of the symbol soft information output by the equalization module in the previous iteration; Step 5.3: The channel vector of the lth path of the mth sub-block is obtained by the following formula: In the formula, is the basis coefficient vector of the lth path, is the basis function matrix corresponding to the mth sub-block, with a dimension of N0×(Q0+1), where Step 5.4: Obtain the basis coefficients of the entire data block through the M channel vectors obtained according to step 5.3, and the vector dimension is N×1: The basis coefficient vector corresponding to the lth path of the entire data block can be obtained by the following formula: In the formula, U = [b 0 ,b 1 ,…,b Q ] is the basis function matrix corresponding to a SC-FDE data block period, and its dimension is N×(Q+1), where Step 5.5: Based on Get the estimated base coefficient Then we further get the estimated basis coefficient matrix The estimated channel matrix H is obtained by the following formula: Step 6, according to the result of the second channel estimation obtained in step 5, it is sent to the frequency domain equalization module to obtain the soft decision information of the equalization output, which is used for the known symbols of the next iterative channel estimation; when the number of iterations is over, the channel information obtained from the last iteration is used to eliminate the inter-code interference of the received signal and the frequency domain equalization module is performed in sequence to obtain the final required bit information.

2. The GCE-BEM iterative channel estimation method based on sub-block transmission according to claim 1, characterized in that: In step 1, the sub-block-based time domain Kronecker TDKD pilot cluster is used as the transmission frame structure, and the specific process is as follows: Step 1.1: Each single carrier frequency domain equalization SC-FDE data block has a length of N = N s +N t , where N s is the information sequence length, N t is the training sequence length, which is divided into M equal-length sub-blocks, with each sub-block having a length of N0 = N / M, where the choice of M depends on the maximum Doppler frequency and the length of the block; Step 1.2: Each sub-block information symbol and training sequence are placed according to the time domain Kronecker TDKD pilot placement method. First, each sub-block is divided into P pilot clusters, where m∈{1,…,M} and p∈{1,…,P} are the indices of the mth sub-block and the pth small sub-block respectively. For each pilot cluster, the length of the information symbol is N s,b =N s / (MP), the length of the training sequence is N t,b =N t / (MP)=2L+1; Step 1.3: Add L guard zeros to the end of each sub-block by setting the last L elements of each training sequence to zero, where L is the number of multipath channels; When the single carrier frequency domain equalization SC-FDE data block x reaches the receiving end through the dual-selection channel, the nth received symbol can be expressed as: In the formula, z(n) has a mean of 0 and a variance of Additive white Gaussian noise AWGN; The above formula is written in matrix form: y=Hx+z In the formula, z=[z(0),z(1),…,z(N-1)] T , y=[y(0),y(1),…,y(N-1)] T , H is an N×N dimensional matrix, as shown below: Step 1.4: Using GCE-BEM, the channel matrix H can be expressed as: In the formula, C q is a Toeplitz matrix, as shown below, and 3. The GCE-BEM iterative channel estimation method based on sub-block transmission according to claim 1, characterized in that: The soft decision information obtained in step 4 after equalization is used for the known symbols of the next channel estimation. The specific process is as follows: The equalized output is soft-determined to obtain the symbol soft information for the next iteration. Then the soft information corresponding to the nth symbol in the pth small sub-block of the mth sub-block is obtained by the following formula: In the formula, and Represents the log-likelihood ratio information of the equalized output.

4. The GCE-BEM iterative channel estimation method based on sub-block transmission according to claim 1, characterized in that: In step 6, iteration starts from the second channel estimation, and the channel estimation method in the iteration adopts the minimum mean square error LMMSE channel estimation method based on GCE-BEM to iteratively update the channel information.

5. A GCE-BEM iterative channel estimation system based on sub-block transmission based on the method according to any one of claims 1 to 4, characterized in that: Specifically include: The channel coding module is used to encode the original bit information and increase the reliability of the information during transmission by adding redundancy; A constellation mapping module is used to modulate the binary baseband signal to different constellation points to form a complex signal required for transmission; The framing module is used to form a specific frame structure required for sending, which is convenient for the receiving end to process; The RF module is used to modulate the transmission signal onto a high-frequency carrier signal and send it out through the antenna; A least squares LS channel estimation module is used to provide initial channel information; Minimum mean square error LMMSE channel estimation module, used to provide iteratively updated channel information; A frequency domain equalization module is used to eliminate inter-symbol interference of the received signal through the output of the channel estimation module; A soft decision module, used to provide an input signal to a channel estimation module based on GCE-BEM and soft information input by a decoding module; The channel decoding module is used to extract the final useful bit information from the soft information.

6. A GCE-BEM iterative channel estimation device based on sub-block transmission, characterized in that: include: Memory: used to store a computer program for implementing the GCE-BEM iterative channel estimation method based on sub-block transmission as described in claims 1-4; Processor: used to implement the GCE-BEM iterative channel estimation method based on sub-block transmission as described in claims 1-4 when executing the computer program.

7. A computer-readable storage medium, characterized in that: include: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it is capable of implementing the GCE-BEM iterative channel estimation method based on sub-block transmission according to claims 1-4.

Images