CN101997639B - Iterative receiver method of low density parity check-multi-input/output communication system - Google Patents

Abstract

Translated fromChinese

本发明公开了一种LDPC-MIMO通信系统的迭代接收方法，减少了接收机算法复杂度。本发明方法通过设定内迭代停止准则和外迭代停止准则，来控制LDPC迭代译码次数和MIMO检测与LDPC译码之间Turbo迭代次数，从而在保持迭代LDPC-MIMO接收机性能基本不变的基础上，减少LDPC迭代译码次数以及MIMO检测与LDPC译码之间Turbo迭代次数，进而减少接收机计算复杂度和计算时延。

The invention discloses an iterative receiving method of an LDPC-MIMO communication system, which reduces the algorithm complexity of the receiver. The method of the present invention controls the number of LDPC iterative decoding and the number of Turbo iterations between MIMO detection and LDPC decoding by setting the inner iteration stop criterion and the outer iteration stop criterion, thereby keeping the performance of the iterative LDPC-MIMO receiver basically unchanged On the basis, the number of LDPC iterative decoding and the number of Turbo iterations between MIMO detection and LDPC decoding are reduced, thereby reducing the computational complexity and computational delay of the receiver.

Description

Translated fromChinese

低密度奇偶校验-多输入多输出通信系统的迭代接收方法Iterative Reception Method for LDPC-MIMO Communication System

技术领域technical field

本发明涉及通信技术领域，尤其涉及一种低密度奇偶校验(LDPC)-多输入多输出(MIMO)通信系统的迭代接收方法。The invention relates to the technical field of communication, in particular to an iterative receiving method of a Low Density Parity Check (LDPC)-Multiple Input Multiple Output (MIMO) communication system.

背景技术Background technique

低密度奇偶校验(Low Density Parity Check，LDPC)码是继Turbo码后纠错编码领域的又一重大进展。在目前诸多信道编码技术中，LDPC码是纠错性能最优秀的方法之一。LDPC码性能接近香农限，具有较低的差错平底特性，译码复杂度低于Turbo码，且可以实现完全的并行操作，吞吐量大，具有高速译码的特质。目前，LDPC码在IEEE 802.16、中国移动多媒体广播(CMMB)等标准中获得采纳，在未来无线通信系统标准4G指定过程中，极有可能取代Turbo码，成为数据业务的首选信道编码解决方案。Low Density Parity Check (Low Density Parity Check, LDPC) code is another major progress in the field of error correction coding after Turbo code. Among many current channel coding techniques, LDPC code is one of the methods with the best error correction performance. The performance of LDPC code is close to the Shannon limit, it has a lower error flat floor, the decoding complexity is lower than that of Turbo code, and it can realize complete parallel operation, high throughput, and has the characteristics of high-speed decoding. At present, LDPC codes have been adopted in standards such as IEEE 802.16 and China Mobile Multimedia Broadcasting (CMMB). In the process of specifying 4G standards for future wireless communication systems, LDPC codes are likely to replace Turbo codes and become the preferred channel coding solution for data services.

LDPC码运用迭代译码的方法进行译码。然而其纠错性能在一定迭代次数之后不再随着迭代次数的增加而改善，同时，迭代次数的增大显著增加了LDPC码的译码计算量。因此，有必要采用一种停止准则来减少LDPC译码迭代中不必要的迭代处理。LDPC codes are decoded using an iterative decoding method. However, its error correction performance does not improve with the increase of the iteration number after a certain number of iterations. At the same time, the increase of the iteration number significantly increases the amount of decoding calculation of LDPC codes. Therefore, it is necessary to adopt a stopping criterion to reduce unnecessary iterative processing in LDPC decoding iterations.

多输入多输出(MIMO)技术是当前及未来提高无线通信性能的关键性技术之一，在宽带码分多址(WCDMA)、IEEE802.16、第三代合作伙伴计划(3GPP)长期演进(LTE)等标准协议中广泛采用，成为未来无线通信标准制定过程中不可或缺的技术之一。MIMO技术能够提供分集增益和空间复用增益，在带宽和发射功率不变的前提下，利用MIMO分集增益能够有效改善信息数据传输可靠性，利用MIMO空间复用增益能够有效提高数据传输速率。Multiple-input multiple-output (MIMO) technology is one of the key technologies to improve wireless communication performance at present and in the future. ) and other standard protocols, and become one of the indispensable technologies in the process of formulating future wireless communication standards. MIMO technology can provide diversity gain and spatial multiplexing gain. Under the premise of constant bandwidth and transmit power, the use of MIMO diversity gain can effectively improve the reliability of information and data transmission, and the use of MIMO spatial multiplexing gain can effectively increase the data transmission rate.

LDPC-MIMO通信系统是指采用LDPC编码和MIMO技术的通信系统。在LDPC-MIMO通信系统中，提高接收机检测性能是LDPC-MIMO系统接收机设计的重要内容。The LDPC-MIMO communication system refers to a communication system using LDPC coding and MIMO technology. In LDPC-MIMO communication system, improving receiver detection performance is an important content of receiver design for LDPC-MIMO system.

为了在传统接收算法基础上进一步提高接收机检测性能，在MIMO检测和LDPC译码之间采用Turbo迭代是一种可行的方法，能够显著提高接收机检测性能。然而与LDPC迭代译码类似，随着Turbo迭代次数的增加，在达到一定程度后，接收机检测性能不再随着Turbo迭代次数的增加而改善，同时接收机复杂度与Turbo迭代次数成正比关系。为此，需要提出一种用于Turbo迭代的停止准则，在Turbo迭代接收机检测性能基本不变的前提下，减少不必要的Turbo迭代，进而减少译码时延。In order to further improve the detection performance of the receiver based on the traditional receiving algorithm, it is a feasible method to use Turbo iteration between MIMO detection and LDPC decoding, which can significantly improve the detection performance of the receiver. However, similar to LDPC iterative decoding, with the increase of the number of Turbo iterations, after reaching a certain level, the receiver detection performance will no longer improve with the increase of the number of Turbo iterations, and the complexity of the receiver is proportional to the number of Turbo iterations . Therefore, it is necessary to propose a stopping criterion for Turbo iteration, which can reduce unnecessary Turbo iteration and further reduce the decoding delay under the premise that the detection performance of the Turbo iteration receiver is basically unchanged.

发明内容Contents of the invention

本发明所要解决的技术问题，在于需要提供一种LDPC-MIMO通信系统的迭代接收方法，减少接收机算法复杂度。The technical problem to be solved by the present invention is to provide an iterative receiving method for an LDPC-MIMO communication system to reduce the algorithm complexity of the receiver.

为了解决上述技术问题，本发明提供了一种低密度奇偶校验(LDPC)-多输入多输出(MIMO)通信系统的迭代接收方法，包括：In order to solve the above-mentioned technical problems, the present invention provides a low-density parity check (LDPC)-multiple-input multiple-output (MIMO) communication system iterative receiving method, comprising:

对接收信号进行MIMO检测，得到估计符号；Perform MIMO detection on the received signal to obtain estimated symbols;

对所述估计符号进行软解调，对所述软解调产生的软比特信息进行硬判决，获得硬判决码字；Performing soft demodulation on the estimated symbol, performing hard decision on the soft bit information generated by the soft demodulation, to obtain a hard decision codeword;

进行一次LDPC译码迭代即内迭代，产生译码信息序列、译码码字及码元对数似然比，更新内迭代次数；Perform an LDPC decoding iteration, that is, an inner iteration, generate a decoding information sequence, a decoding codeword and a symbol logarithmic likelihood ratio, and update the number of inner iterations;

对所述译码码字进行校验，校验成功则输出所述译码信息序列，LDPC-MIMO迭代译码结束，否则判断所述内迭代是否满足预设的内迭代停止准则；Verifying the decoded codeword, if the verification is successful, the decoding information sequence is output, and the LDPC-MIMO iterative decoding ends, otherwise, it is judged whether the inner iteration meets the preset inner iteration stop criterion;

所述内迭代满足所述内迭代停止准则时，输出所述译码信息序列、译码码字及码元对数似然比；否则进行下一次所述内迭代；When the inner iteration meets the inner iteration stop criterion, output the decoding information sequence, decoding codeword and symbol logarithmic likelihood ratio; otherwise, perform the next inner iteration;

根据所述硬判决码字及所述译码码字，判断MIMO检测与LDPC译码之间的迭代即外迭代是否满足预设的外迭代停止准则，满足则所述LDPC-MIMO迭代译码结束，否则更新外迭代次数，对所述码元对数似然比进行软映射，获得软映射调制符号反馈至所述MIMO检测。According to the hard decision codeword and the decoding codeword, it is judged whether the iteration between MIMO detection and LDPC decoding, that is, the outer iteration, satisfies the preset outer iteration stop criterion, and if it is satisfied, the LDPC-MIMO iterative decoding ends , otherwise update the number of outer iterations, perform soft mapping on the log likelihood ratio of the symbol, and obtain a soft mapped modulation symbol to feed back to the MIMO detection.

优选地，判断所述内迭代是否满足所述内迭代停止准则的步骤，包括：Preferably, the step of judging whether the inner iteration meets the inner iteration stop criterion includes:

计算码元对数似然比绝对值的方差；Calculate the variance of the absolute value of the symbol logarithmic likelihood ratio;

检查内迭代次数；Check inner iteration count;

判断所述方差是否大于预设的方差门限且所述内迭代次数小于预设的最大内迭代次数，是则不满足所述内迭代停止准则，否则满足所述内迭代停止准则。Judging whether the variance is greater than a preset variance threshold and the number of inner iterations is less than a preset maximum number of inner iterations, if yes, the inner iteration stop criterion is not satisfied, otherwise the inner iteration stop criterion is satisfied.

优选地，判断所述外迭代是否满足所述外迭代停止准则的步骤，包括：Preferably, the step of judging whether the outer iteration meets the outer iteration stop criterion includes:

对于首次外迭代，For the first outer iteration,

对所述硬判决码字及所述译码码字进行比较，计算对应比特位置上不相同的比特数，获得反映LDPC译码错误程度的长度比值；Comparing the hard-decision codeword and the decoded codeword, calculating the number of different bits in the corresponding bit position, and obtaining a length ratio reflecting the LDPC decoding error degree;

判断所述长度比值是否小于预设的长度比值门限且外迭代次数是否小于预设的最大外迭代次数，是则不满足所述外迭代停止准则，否则满足所述外迭代停止准则；Judging whether the length ratio is less than a preset length ratio threshold and whether the number of outer iterations is less than a preset maximum number of outer iterations, if yes, the outer iteration stop criterion is not satisfied, otherwise the outer iteration stop criterion is satisfied;

对非首次外迭代，For non-first outer iterations,

对所述硬判决码字及所述译码码字进行比较，进一步获得反映外迭代译码收敛性的比特数比值；Comparing the hard-decision codeword and the decoded codeword to further obtain a bit ratio reflecting the convergence of the outer iterative decoding;

进一步判断所述比特数比值是否小于预设的比特数比值门限，是则不满足所述外迭代停止准则，否则满足所述外迭代停止准则。It is further judged whether the bit ratio is smaller than a preset bit ratio threshold, if yes, the outer iteration stop criterion is not satisfied, otherwise the outer iteration stop criterion is satisfied.

进一步地，所述长度比值，表示所述不相同的比特数相对于所述译码码字长度的比值；Further, the length ratio indicates the ratio of the number of different bits to the length of the decoding codeword;

所述比特数比值，表示本次外迭代译码相对于上次外迭代译码不相同的比特数的比值。The ratio of the number of bits indicates the ratio of the number of bits that are different in this outer iterative decoding compared to the last outer iterative decoding.

优选地，对所述接收信号进行所述MIMO检测，包括：Preferably, performing the MIMO detection on the received signal includes:

对于首次外迭代，仅对所述接收信号进行所述MIMO检测，For the first outer iteration, the MIMO detection is only performed on the received signal,

对于非首次外迭代，对所述接收信号及所述软映射调制符号进行所述MIMO检测。For non-first outer iterations, the MIMO detection is performed on the received signal and the soft-mapped modulation symbols.

优选地，对所述估计符号进行归一化处理后，再进行所述软解调。Preferably, the soft demodulation is performed after performing normalization processing on the estimated symbols.

优选地，对所述软比特信息进行所述硬判决的步骤，包括：Preferably, the step of performing the hard decision on the soft bit information includes:

对所述软比特信息正负值进行所述硬判决。The hard decision is made on the positive and negative values of the soft bit information.

优选地，初始时，所述内迭代次数为0，所述外迭代次数为1。Preferably, initially, the number of inner iterations is 0, and the number of outer iterations is 1.

优选地，对所述译码码字进行所述校验的步骤，包括：Preferably, the step of verifying the decoded codeword includes:

采用校验方程cH^T＝0对所述译码码字进行所述校验；Performing the verification on the decoded codeword by using the verification equation cH^T =0;

其中，c为所述译码码字，H为LDPC校验矩阵。Wherein, c is the decoding codeword, and H is the LDPC check matrix.

优选地，所述外迭代，包括Turbo迭代。Preferably, the outer iteration includes Turbo iteration.

与现有技术相比，本发明提出的双迭代控制的LDPC-MIMO迭代接收方法，通过对内迭代(LDPC迭代译码)和外迭代(MIMO检测与LDPC译码之间迭代)同时引入迭代停止准则，从而在保证迭代LDPC-MIMO接收机性能基本不变的基础上，减少了内外迭代次数，进而减少了接收机算法复杂度和计算时延。Compared with the prior art, the LDPC-MIMO iterative receiving method with dual iterative control proposed by the present invention introduces an iterative stop through internal iteration (LDPC iterative decoding) and external iteration (iteration between MIMO detection and LDPC decoding) Therefore, on the basis of ensuring that the performance of the iterative LDPC-MIMO receiver is basically unchanged, the number of internal and external iterations is reduced, thereby reducing the algorithm complexity and calculation delay of the receiver.

附图说明Description of drawings

图1为本发明方法实施例的流程示意图；Fig. 1 is the schematic flow sheet of the method embodiment of the present invention;

图2为本发明方法应用实例中LDPC-MIMO通信系统发射机框图；Fig. 2 is a block diagram of an LDPC-MIMO communication system transmitter in an application example of the method of the present invention;

图3为本发明方法应用实例中LDPC-MIMO通信系统接收机框图；Fig. 3 is the receiver block diagram of LDPC-MIMO communication system in the method application example of the present invention;

图4是本发明方法应用实例流程示意图；Fig. 4 is a schematic flow chart of the application example of the method of the present invention;

图5是本发明方法应用实例中格雷映射16QAM调制星座。Fig. 5 is a Gray-mapped 16QAM modulation constellation in an application example of the method of the present invention.

具体实施方式Detailed ways

以下将结合附图及实施例来详细说明本发明的实施方式，借此对本发明如何应用技术手段来解决技术问题，并达成技术效果的实现过程能充分理解并据以实施。The implementation of the present invention will be described in detail below in conjunction with the accompanying drawings and examples, so as to fully understand and implement the process of how to apply technical means to solve technical problems and achieve technical effects in the present invention.

本发明提供一种双迭代控制的LDPC-MIMO迭代接收方法，通过设定内迭代停止准则和外迭代停止准则，来控制LDPC迭代译码次数和MIMO检测与LDPC译码之间Turbo迭代次数，从而在保持迭代LDPC-MIMO接收机性能基本不变的基础上，减少LDPC迭代译码次数以及MIMO检测与LDPC译码之间Turbo迭代次数，进而减少接收机计算复杂度和计算时延。其中，本发明将LDPC迭代称之为内迭代，将MIMO检测与LDPC译码之间的迭代称之为外迭代，以区分这两种迭代。The present invention provides an LDPC-MIMO iterative receiving method with dual iterative control, which controls the number of LDPC iterative decoding and the number of Turbo iterations between MIMO detection and LDPC decoding by setting the inner iterative stop criterion and the outer iterative stop criterion, thereby On the basis of keeping the performance of the iterative LDPC-MIMO receiver basically unchanged, the number of LDPC iterative decoding and the number of Turbo iterations between MIMO detection and LDPC decoding are reduced, thereby reducing the computational complexity and delay of the receiver. Wherein, in the present invention, the LDPC iteration is called inner iteration, and the iteration between MIMO detection and LDPC decoding is called outer iteration, so as to distinguish these two iterations.

图1为本发明方法实施例的流程示意图。如图1所示，该方法实施例主要包括如下步骤：Fig. 1 is a schematic flow chart of a method embodiment of the present invention. As shown in Figure 1, the method embodiment mainly includes the following steps:

步骤S110，对接收信号进行MIMO检测处理，得到估计符号；Step S110, performing MIMO detection processing on the received signal to obtain estimated symbols;

步骤S120，对MIMO检测得到的估计符号进行软解调，并对软解调后产生的软比特信息进行硬判决，获得硬判决码字；Step S120, performing soft demodulation on the estimated symbol obtained by MIMO detection, and performing hard decision on the soft bit information generated after soft demodulation, to obtain a hard decision codeword;

步骤S130，进行一次LDPC译码迭代，产生译码信息序列、译码码字及其码元对数似然比(LLR)，更新LDPC译码迭代次数，其中该LDPC译码迭代称之为内迭代，该码元表示译码码字中的码元；Step S130, perform an iteration of LDPC decoding, generate a decoding information sequence, a decoding codeword and its symbol logarithmic likelihood ratio (LLR), and update the number of iterations of LDPC decoding, wherein the iteration of LDPC decoding is called an internal Iterating, the symbol represents a symbol in the decoded codeword;

步骤S140，对LDPC译码产生的译码码字采用校验方程进行校验，若校验成功，则转步骤S145，否则转步骤S146；Step S140, the decoding code word that LDPC decoding produces adopts check equation to check, if check is successful, then turn to step S145, otherwise turn to step S146;

步骤S145，LDPC译码成功，译码信息序列即为发送端所发送的信息序列，输出该译码信息序列，LDPC-MIMO迭代译码过程结束；Step S145, the LDPC decoding is successful, the decoding information sequence is the information sequence sent by the sending end, the decoding information sequence is output, and the LDPC-MIMO iterative decoding process ends;

步骤S146，计算码元对数似然比绝对值的方差，检查LDPC迭代译码次数；Step S146, calculating the variance of the absolute value of the symbol logarithmic likelihood ratio, checking the number of times of LDPC iterative decoding;

步骤S147，判断是否同时满足方差大于预先设定的方差门限并且LDPC迭代译码次数小于预设的最大LDPC译码迭代次数即最大内迭代次数，是则转入步骤S130进行下一次LDPC译码，否则转步骤S148；Step S147, judging whether the variance is greater than the preset variance threshold and the number of LDPC iterative decoding is less than the preset maximum number of LDPC decoding iterations, i.e. the maximum number of inner iterations, if so, proceed to step S130 for the next LDPC decoding, Otherwise go to step S148;

步骤S148，LDPC译码失败，输出译码信息序列、译码码字以及码元对数似然比，然后转步骤S150；Step S148, LDPC decoding fails, output decoding information sequence, decoding code word and symbol logarithmic likelihood ratio, then go to step S150;

步骤S150，对前述步骤S120中的硬判决码字与步骤S 130中LDPC译码产生的译码码字进行比较，计算对应比特位置上不相同的比特数，由此获得外迭代控制参数；Step S150, comparing the hard-decision codeword in the aforementioned step S120 with the decoded codeword produced by LDPC decoding in the step S130, calculating the number of bits that are different in the corresponding bit positions, thereby obtaining the outer iteration control parameter;

步骤S160，将外迭代控制参数与预先设定的门限值进行比较，当外迭代控制参数小于预设的门限值且外迭代次数小于预设的最大外迭代次数时，更新外迭代次数，转入步骤S 170，否则LDPC-MIMO迭代译码失败，LDPC-MIMO迭代译码过程结束；Step S160, comparing the outer iteration control parameter with a preset threshold value, and updating the outer iteration number when the outer iteration control parameter is smaller than the preset threshold value and the number of outer iterations is less than the preset maximum number of outer iterations, Go to step S170, otherwise the LDPC-MIMO iterative decoding fails, and the LDPC-MIMO iterative decoding process ends;

步骤S170，对码元对数似然比进行软映射，获得软映射调制符号，将该软映射调制符号反馈至步骤S110中的MIMO检测，继续进行下一次外迭代处理。Step S170, perform soft mapping on the symbol log likelihood ratio to obtain a soft mapped modulation symbol, feed back the soft mapped modulation symbol to the MIMO detection in step S110, and continue to perform the next outer iteration process.

其中，步骤S110中，对于首次外迭代，仅对接收信号进行MIMO检测；对于非首次外迭代，同时对接收信号和步骤S170反馈的软映射调制符号进行MIMO检测。Wherein, in step S110, for the first outer iteration, MIMO detection is only performed on the received signal; for non-first outer iterations, MIMO detection is performed on the received signal and the soft-mapped modulation symbol fed back in step S170.

其中，步骤S120中，根据所采用的用于软解调的调制方式，将MIMO检测获取的估计符号软解调为软比特信息，并对软比特信息正负值进行硬判决处理，产生硬判决码字。Wherein, in step S120, according to the modulation method used for soft demodulation, the estimated symbol obtained by MIMO detection is soft demodulated into soft bit information, and the positive and negative values of the soft bit information are subjected to hard decision processing to generate hard decision Codeword.

其中，步骤S130中，选用LDPC译码算法进行一次LDPC译码，产生译码信息序列(即译码结果)、译码码字及其码元对数似然比，然后更新LDPC译码迭代次数，即将LDPC译码迭代次数加一，初始时LDPC译码迭代次数为0。Wherein, in step S130, select LDPC decoding algorithm to carry out LDPC decoding once, produce decoding information sequence (i.e. decoding result), decoding code word and symbol logarithm likelihood ratio thereof, then update LDPC decoding iteration number , that is, the number of iterations of LDPC decoding is increased by one, and the number of iterations of LDPC decoding is 0 initially.

其中，步骤S140中，采用校验方程cH^T＝0对LDPC译码产生的译码码字进行校验。其中，c为译码码字，H为LDPC校验矩阵。若校验方程成立，则LDPC译码成功，译码信息序列即为发送端所发送信息序列，输出译码信息序列，LDPC-MIMO迭代译码过程结束；若校验方程不成立，计算码元对数似然比绝对值的方差，当同时满足方差大于预先设定的方差门限和LDPC迭代译码次数小于预先设定的最大LDPC迭代译码次数时，转入步骤S130继续进行LDPC迭代译码，否则LDPC迭代译码失败，输出译码信息序列、译码码字以及码元对数似然比。Wherein, in step S140, the decoded codeword generated by LDPC decoding is verified by using the verification equation cH^T =0. Among them, c is the decoding codeword, and H is the LDPC check matrix. If the check equation is established, the LDPC decoding is successful, and the decoded information sequence is the information sequence sent by the sender, and the decoded information sequence is output, and the LDPC-MIMO iterative decoding process ends; if the check equation is not established, the symbol pairs are calculated The variance of the absolute value of the likelihood ratio, when the variance is greater than the preset variance threshold and the LDPC iterative decoding number of times is less than the preset maximum LDPC iterative decoding number of times, then proceed to step S130 to continue the LDPC iterative decoding, Otherwise, the LDPC iterative decoding fails, and the decoding information sequence, decoding codeword and symbol logarithmic likelihood ratio are output.

其中，步骤S147中，方差小于等于预先设定的方差门限，和/或LDPC迭代译码次数大于等于预设的最大LDPC译码迭代次数，则认为满足前述的内迭代停止准则。满足该内迭代停止准则，则转步骤S148，否则转步骤S130进行一下次LDPC译码。Wherein, in step S147, if the variance is less than or equal to the preset variance threshold, and/or the number of LDPC iterative decoding is greater than or equal to the preset maximum number of LDPC decoding iterations, it is considered that the aforementioned internal iteration stop criterion is satisfied. If the internal iteration stop criterion is met, go to step S148, otherwise go to step S130 to perform LDPC decoding next time.

其中，步骤S150中，对步骤S120中的硬判决码字与步骤S130中LDPC译码产生的译码码字进行比较，计算对应比特位置上不相同的比特数，由此计算外迭代控制参数，外迭代控制参数主要有两个：Wherein, in step S150, the hard-decision codeword in step S120 is compared with the decoded codeword produced by LDPC decoding in step S130, and the number of bits that are different in corresponding bit positions is calculated, thereby calculating the outer iteration control parameter, There are two main parameters for outer iteration control:

长度比值P₁，表示不相同比特数相对于译码码字长度的比值，反映LDPC译码错误程度；The length ratio P₁ represents the ratio of the number of different bits to the length of the decoded codeword, reflecting the degree of LDPC decoding error;

比特数比值P₁，本次外迭代译码相对于上次外迭代译码不相同比特数的比值，反映外迭代译码收敛性；Bit number ratio P₁ , the ratio of different bit numbers in this outer iterative decoding compared to the previous outer iterative decoding, reflecting the convergence of outer iterative decoding;

对于首次外迭代，只比较长度比值P₁与其设定的长度比值门限T_P1之间的关系，当长度比值P₁小于长度比值门限T_P1，且外迭代次数小于预设的最大外迭代次数时，在更新外迭代次数后，转入步骤S170处理，否则LDPC-MIMO迭代译码失败，LDPC-MIMO迭代译码过程结束；For the first outer iteration, only compare the relationship between the length ratio P₁ and its set length ratio threshold T_P1 , when the length ratio P₁ is smaller than the length ratio threshold T_P1 , and the number of outer iterations is less than the preset maximum number of outer iterations , after updating the number of outer iterations, turn to step S170 for processing, otherwise the LDPC-MIMO iterative decoding fails, and the LDPC-MIMO iterative decoding process ends;

对于非首次外迭代，需要比较长度比值P₁与长度比值门限T_P1之间的关系，以及比特数比值P₁与比特数比值门限T_P2之间关系，当长度比值P₁小于与长度比值门限T_P1，比特数比值P₂小于比特数比值门限T_P2且外迭代次数小于预设的最大外迭代次数时，在更新外迭代次数后，转入步骤S170处理，否则LDPC-MIMO迭代译码失败，LDPC-MIMO迭代译码过程结束；For the non-first outer iteration, it is necessary to compare the relationship between the length ratio P₁ and the length ratio threshold T_P1 , and the relationship between the bit number ratio P₁ and the bit number ratio threshold T_P2 , when the length ratio P₁ is less than the length ratio threshold T_P1 , when the bit ratio P₂ is less than the bit ratio threshold T_P2 and the number of outer iterations is less than the preset maximum number of outer iterations, after updating the number of outer iterations, go to step S170 for processing; otherwise, LDPC-MIMO iterative decoding fails , the LDPC-MIMO iterative decoding process ends;

初始时，外迭代次数为1。Initially, the number of outer iterations is 1.

其中，步骤S160中，当外迭代控制参数大于等于预设的门限值，和/或外迭代次数大于等于预设的最大外迭代次数时，认为满足前述的外迭代停止准则。满足该外迭代停止准则，则LDPC-MIMO迭代译码过程结束，否则更新外迭代次数并转步骤S170。Wherein, in step S160, when the outer iteration control parameter is greater than or equal to the preset threshold value, and/or the number of outer iterations is greater than or equal to the preset maximum number of outer iterations, it is considered that the aforementioned outer iteration stop criterion is satisfied. If the outer iteration stop criterion is satisfied, the LDPC-MIMO iterative decoding process ends, otherwise, update the number of outer iterations and go to step S170.

其中，步骤S170中，将步骤S130产生的码元对数似然比软映射为调制符号，反馈回步骤S110进行MIMO检测，从而进行下一次外迭代操作。Wherein, in step S170, the symbol log-likelihood ratio generated in step S130 is soft-mapped into modulation symbols, and fed back to step S110 for MIMO detection, so as to perform the next outer iteration operation.

以下借助一具体的应用实例对本发明的上述实施例进行进一步的详细说明。The above-mentioned embodiment of the present invention will be further described in detail below with the help of a specific application example.

如图2所示，在LDPC-MIMO通信系统发射侧，二进制信息比特序列i＝[i，i₂，…，i_K]输入LDPC编码器，进行LDPC编码。LDPC编码后二进制LDPC码字比特序列为c＝[c₁，c₂，…，c_N]，LDPC编码码率为R＝K/N。随后，LDPC码字比特序列输入调制器，根据系统确定的调制方式进行调制，获得符号序列s＝[s₁，s₂，…，s_L]，调制阶数为M＝N/L。调制后符号序列s经串并转换后，分别送入多个发射天线进行发射。某时刻，复数发送信号向量$x={(x_1,x_2,\·\·\·,x_{N_T})}^T,$其中N_T为发射天线数。As shown in FIG. 2 , at the transmitting side of the LDPC-MIMO communication system, the binary information bit sequence i=[i, i₂ , . . . , i_K ] is input to the LDPC encoder for LDPC encoding. The binary LDPC codeword bit sequence after LDPC encoding is c=[c₁ , c₂ ,...,c_N ], and the LDPC encoding code rate is R=K/N. Subsequently, the LDPC codeword bit sequence is input into the modulator, and modulated according to the modulation method determined by the system to obtain the symbol sequence s=[s₁ , s₂ ,...,s_L ], and the modulation order is M=N/L. The modulated symbol sequence s is sent to multiple transmitting antennas for transmission after serial-to-parallel conversion. At some point, the complex sending signal vector$x={(x_1,x_2,\&Center\; Dot;\&Center\; Dot;\·,x_{N_T})}^T,$ where_NT is the number of transmit antennas.

如图3所示，在LDPC-MIMO通信系统接收侧，某时刻，复数接收信号向量$y={(y_1,y_2,\·\·\·,y_{N_R})}^T,$其中N_R为接收天线数。收发端之间信号可以建模为：y＝Hx+n，其中，复数信道矩阵$H={\left[h_{\mathrm{ij}}\right]}_{N_R\×N_T}$是N_R×N_T维平坦Rayleigh衰落信道系数矩阵，其中元素h_ij代表从第j根发送天线到第i根接收天线的信道衰落系数，为均值0、方差1的独立同分布复高斯变量；$n={(n_1,n_2,\·\·\·,n_{N_R})}^T$为高斯白噪声分量，其元素为均值0，方差σ²的独立同分布复高斯变量。As shown in Figure 3, at the receiving side of the LDPC-MIMO communication system, at a certain moment, the complex received signal vector$\mathrm{the\; y}={({\mathrm{the\; y}}_1,{\mathrm{the\; y}}_2,\·\&Center\; Dot;\·,{\mathrm{the\; y}}_{N_R})}^T,$ Where_NR is the number of receiving antennas. The signal between the transmitting and receiving ends can be modeled as: y=Hx+n, where the complex channel matrix$h={\left[h_{\mathrm{ij}}\right]}_{N_R\×N_T}$ is the N_R × N_T dimensional flat Rayleigh fading channel coefficient matrix, in which the element h_ij represents the channel fading coefficient from the jth transmitting antenna to the ith receiving antenna, which is an independent and identically distributed complex Gaussian variable with a mean value of 0 and a variance of 1;$\mathrm{no}={({\mathrm{no}}_1,{\mathrm{no}}_2,\&Center\; Dot;\&Center\; Dot;\&Center\; Dot;,{\mathrm{no}}_{N_R})}^T$ is a Gaussian white noise component whose elements are independent and identically distributed complex Gaussian variables with mean 0 and variance σ² .

图4为本发明双迭代控制的LDPC-MIMO迭代接收方法应用实例流程示意图。本应用实例中MIMO检测算法采用并行软干扰消除-最小均方误差MIMO检测算法，调制方式采用M-QAM星座图调制。如图4所示，本发明方法应用实例包括以下步骤：Fig. 4 is a schematic flow chart of an application example of the LDPC-MIMO iterative receiving method with dual iterative control according to the present invention. In this application example, the MIMO detection algorithm adopts the parallel soft interference cancellation-minimum mean square error MIMO detection algorithm, and the modulation mode adopts M-QAM constellation diagram modulation. As shown in Figure 4, the application example of the method of the present invention comprises the following steps:

步骤S410，对接收信号进行MIMO检测，获得估计符号；Step S410, performing MIMO detection on the received signal to obtain estimated symbols;

对于首次外迭代，对接收信号进行MIMO检测，MIMO检测输出估计符号向量为$\tilde{x}={[H^{H}H+\frac{1}{\mathrm{SNR}}I]}^{-1}{H}^{H}y;$For the first outer iteration, MIMO detection is performed on the received signal, and the MIMO detection output estimates the symbol vector as$\tilde{x}={[h^{h}h+\frac{1}{\mathrm{SNR}}I]}^{-1}{h}^h\mathrm{the\; y};$

对于非首次外迭代，同时对接收信号和反馈的软映射调制符号进行MIMO检测，MIMO检测输出估计符号向量为For non-first outer iterations, MIMO detection is performed on the received signal and the feedback soft-mapped modulation symbols at the same time, and the MIMO detection output estimated symbol vector is

$\stackrel{<span><span>~</span>~</span>}{\mathrm{<span><span>x</span>x</span>}}<span><span>=</span>=</span>{<span><span>[</span>[</span>\mathrm{<span><span>diag</span>diag</span>}<span><span>(</span>(</span>{\mathrm{<span><span>H</span>h</span>}}^{\mathrm{<span><span>H</span>h</span>}}\mathrm{<span><span>H</span>h</span>}<span><span>)</span>)</span><span><span>+</span>+</span>\frac{\mathrm{<span><span>1</span>1</span>}}{\mathrm{<span><span>SNR</span>SNR</span>}}\mathrm{<span><span>I</span>I</span>}<span><span>]</span>]</span>}^{<span><span>-</span>-</span>\mathrm{<span><span>1</span>1</span>}}<span><span>\&CenterDot;</span>\&CenterDot;</span><span><span>[</span>[</span>{\mathrm{<span><span>H</span>h</span>}}^{\mathrm{<span><span>H</span>h</span>}}\mathrm{<span><span>y</span>the\; y</span>}<span><span>-</span>-</span>\mathrm{<span><span>ddiag</span>ddiag</span>}<span><span>(</span>(</span>{\mathrm{<span><span>H</span>h</span>}}^{\mathrm{<span><span>H</span>h</span>}}\mathrm{<span><span>H</span>h</span>}<span><span>)</span>)</span>\stackrel{<span><span>^</span>^</span>}{\mathrm{<span><span>x</span>x</span>}}<span><span>]</span>]</span><span><span>;</span>;</span>$

其中，表示反馈的软映射调制符号向量，diag(H^HH)和ddiag(H^HH)分别表示包含矩阵H^HH的对角和非对角元素矩阵；y表示复数接收信号向量，H^H表示复数信道矩阵H的共轭转置；SNR表示信噪比；I表示N_T×N_T的单位矩阵。in, Represents the feedback soft-mapped modulation symbol vector, diag(H^H H) and ddiag(H^H H) represent the diagonal and off-diagonal element matrices containing the matrix H^H H, respectively; y represents the complex received signal vector, H^H represents the complex The conjugate transpose of the channel matrix H; SNR means the signal-to-noise ratio; I means the identity matrix of N_T × N_T.

步骤S420，将MIMO检测获取的估计符号软解调为比特对数似然比序列1＝[l₁，l₂，…，l_N]；Step S420, soft-demodulate the estimated symbols acquired by MIMO detection into a bit log likelihood ratio sequence 1=[l₁ , l₂ ,...,l_N ];

本应用实例中采用格雷映射16QAM调制，其映射生成关系如图5所示，且MIMO检测输出估计符号经归一化处理；In this application example, Gray mapping 16QAM modulation is used, and the mapping generation relationship is shown in Figure 5, and the MIMO detection output estimated symbols are normalized;

设l_kⁱ为估计符号软解调后的第i个比特的对数似然比，则$l_k^1=\mathrm{Re}\left({\tilde{x}}_k\right),$$l_k^2=\mathrm{Im}\left({\tilde{x}}_k\right),$$l_k^3=C_1-\left|l_k^1\right|,$$l_k^4=C_1-\left|l_k^2\right|,$其中$C_1=\frac{2}{\sqrt{10}};$Let l_kⁱ be the estimated symbol The logarithmic likelihood ratio of the i-th bit after soft demodulation, then$l_k^1=\mathrm{Re}\left({\tilde{x}}_k\right),$$l_k^2=\mathrm{Im}\left({\tilde{x}}_k\right),$$l_k^3=C_1-\left|l_k^1\right|,$$l_k^4=C_1-\left|l_k^2\right|,$ in$C_1=\frac{2}{\sqrt{10}};$

软解调后，对获取的比特对数似然比正负号硬判决为0或1的硬判决码字序列c′＝[c₁′，c₂′，…，c_N′]；After soft demodulation, the hard-decision codeword sequence c′=[c₁ ′, c₂ ′, ..., c_N ′] for the obtained bit log-likelihood ratio sign is hard-judged as 0 or 1;

步骤S430，选用LDPC译码算法，例如反向传播(BP)算法或者Log-BP算法等，本应用实例选用BP算法，进行一次LDPC译码迭代，产生译码信息序列i′＝[i₁′，i₂′，…，i_K′](关于二进制比特序列i的LDPC译码结果)、译码码字c″＝[c₁″，c₂″，…，c_N″](关于码字比特序列c的LDPC译码判决)及其对应的码元对数似然比序列L＝[L₁，L₂，…，L_N]；Step S430, select the LDPC decoding algorithm, such as the backpropagation (BP) algorithm or the Log-BP algorithm, etc., this application example selects the BP algorithm, performs an LDPC decoding iteration, and generates the decoding information sequence i'=[i₁ ' , i₂ ′,..., i_K ′] (with respect to the LDPC decoding result of the binary bit sequence i), the decoded code word c″=[c₁ ″, c₂ ″,…, c_N ″] (with respect to the code word LDPC decoding decision of bit sequence c) and its corresponding symbol logarithmic likelihood ratio sequence L=[L₁ , L₂ ,...,L_N ];

更新LDPC译码迭代次数I_LDPC，即将LDPC译码迭代次数加1，I_LDPC＝I_LDPC+1；初始时，LDPC译码迭代次数I_LDPC为0；Update the number of iterations I_LDPC of LDPC decoding, that is, add 1 to the number of iterations of LDPC decoding, I_LDPC =I_LDPC +1; initially, the number of iterations I_LDPC of LDPC decoding is 0;

步骤S440，对LDPC译码产生的译码码字c″采用校验方程c″H^T＝0进行校验，若校验方程成立，则LDPC译码成功，转步骤S445，若校验方程不成立，则转步骤S446；Step S440, the decoding code word c " that LDPC decoding produces adopts verification equation c " H^T =0 to check, if verification equation is established, then LDPC decoding is successful, turn to step S445, if verification equation is not established , turn to step S446;

步骤S445，译码信息序列i′即为发送端所发送的信息序列i，输出译码信息序列i′，LDPC-MIMO迭代译码过程结束；Step S445, the decoded information sequence i' is the information sequence i sent by the sender, and the decoded information sequence i' is output, and the LDPC-MIMO iterative decoding process ends;

步骤S446，计算对数似然比序列L绝对值的方差σ_L²；Step S446, calculating the variance σ_L² of the absolute value of the log-likelihood ratio sequence L;

步骤S447，当${\mathrm{\&sigma;}}_L^2>T_L$且I_LDPC＜Iter_{LDPC_MAX}，转入步骤S430继续进行LDPC迭代译码，否则转步骤S448；Step S447, when${\mathrm{\&sigma;}}_L^2>T_L$ And I_LDPC <Iter_{LDPC_MAX} , go to step S430 to continue LDPC iterative decoding, otherwise go to step S448;

步骤S448，LDPC迭代译码失败，输出译码信息序列i′、译码码字c″以及对应的码元对数似然比序列L，然后执行步骤S450；Step S448, LDPC iterative decoding fails, output decoding information sequence i', decoding code word c" and corresponding symbol logarithmic likelihood ratio sequence L, and then execute step S450;

其中，T_L为预先设定的方差门限，I_LDPC为迭代译码次数，Iter_{LDPC_MAX}为预设的LDPC最大迭代译码次数；Wherein,_TL is a preset variance threshold, I_LDPC is the number of iterative decoding, and Iter_{LDPC_MAX} is the maximum number of iterative decoding of preset LDPC;

步骤S450，对步骤S420中的硬判决码字c′与步骤S430中LDPC译码产生的译码码字c″进行比较，计算对应比特位置不相同的比特数N_e；Step S450, compare the hard-decision codeword c' in step S420 with the decoded codeword c" produced by LDPC decoding in step S430, and calculate the number of bits N_e whose corresponding bit positions are different;

对于首次外迭代，计算外迭代控制参数中的长度比值P₁，P₁＝N_e/N；For the first outer iteration, calculate the length ratio P₁ in the outer iteration control parameters, P₁ =N_e /N;

对于非首次外迭代，计算外迭代控制参数长度比值P₁和比特数比值P₂，其中P₁＝N_e/N，$P_2=N_e^{\mathrm{cur}}/N_e^{\mathrm{last}};$For non-first outer iteration, calculate outer iteration control parameter length ratio P₁ and bit number ratio P₂ , where P₁ =N_e /N,$P_2=N_e^{\mathrm{cur}}/N_e^{\mathrm{last}};$

其中，in,

N为译码码字c″的长度；N is the length of decoding code word c ";

N_e^cur和N_e^last分别表示本次与上次外迭代中N₃的取值；N_e^cur and N_e^last represent the value of N₃ in this and last outer iteration respectively;

步骤S460，对于首次外迭代，当P₁＜T_P1且I_Outer＜Iter_{Outer_MAX}时，在外迭代次数I_Outer加1后，转入步骤S470处理，否则，LDPC-MIMO迭代译码失败，LDPC-MIMO迭代译码过程结束；Step S460, for the first outer iteration, when P₁ <T_P1 and I_Outer <Iter_{Outer_MAX} , after the number of outer iterations I_Outer is increased by 1, turn to step S470 for processing, otherwise, LDPC-MIMO iterative decoding fails, LDPC-MIMO The iterative decoding process ends;

对于非首次外迭代，当P₁＜T_P1、P₂＜T_P2且I_Outer＜Iter_{Outer_MAX}时，在外迭代次数I_Outer加1后，转入步骤S470处理，否则LDPC-MIMO迭代译码失败，LDPC-MIMO迭代译码过程结束；For non-first outer iterations, when P₁ <T_P1 , P₂ <T_P2 and I_Outer <Iter_{Outer_MAX} , after adding 1 to the number of outer iterations I_Outer , turn to step S470 for processing, otherwise LDPC-MIMO iterative decoding fails, The LDPC-MIMO iterative decoding process ends;

其中，长度比值门限T_P1和比特数比值门限T_P2，分别为长度比值P₁和比特数比值P₂预先设定的门限；Wherein, the length ratio threshold T_P1 and the bit number ratio threshold T_P2 are the preset thresholds of the length ratio P₁ and the bit number ratio P₂ respectively;

初始时，外迭代次数I_Outer为1；Initially, the number of iterations I_Outer is 1;

步骤S470，对步骤S430产生的码元对数似然比序列进行软映射处理，得到软映射调制符号，将该软映射调制符号反馈回步骤S410继续进行下一次外迭代处理；Step S470, performing soft mapping processing on the symbol log likelihood ratio sequence generated in step S430 to obtain a soft mapping modulation symbol, and feeding back the soft mapping modulation symbol to step S410 to continue the next outer iteration process;

软映射后的调制符号设为则：The modulation symbol after soft mapping is set to but:

其中，$P_{i,k}^{c^i}=\frac{\exp \left(c^{i}L\left(c_{k,i}\right)\right)}{1+\exp \left(c^{i}L\left(c_{k,i}\right)\right)},i=\mathrm{1,2,3,4};$in,$P_{i,k}^{c^i}=\frac{\exp \left(c^{i}L\left(c_{k,i}\right)\right)}{1+\exp \left(c^{i}L\left(c_{k,i}\right)\right)},i=\mathrm{1,2,3,4};$

L(c_k，i)表示对数似然比序列L中对应码元c_k，i的对数似然比取值，cⁱ表示码元c_k，i的取值。L(c_{k, i} ) represents the value of the log likelihood ratio corresponding to the symbol c_k,i in the log likelihood ratio sequence L, and cⁱ represents the value of the symbol c_k,i .

综上所述，本发明提出的双迭代控制的LDPC-MIMO迭代接收方法，通过设置LDPC译码迭代(内迭代)停止准则和MIMO检测与LDPC译码之间迭代(外迭代)停止准则，在保持LDPC-MIMO迭代接收机性能基本不变的基础上，减少了不必要的内外迭代次数，从而减少了LDPC-MIMO迭代接收机算法复杂度和处理时延。In summary, the LDPC-MIMO iterative reception method with dual iterative control proposed by the present invention, by setting the LDPC decoding iteration (inner iteration) stop criterion and the iteration (outer iteration) stop criterion between MIMO detection and LDPC decoding, in On the basis of keeping the performance of the LDPC-MIMO iterative receiver basically unchanged, the unnecessary number of internal and external iterations is reduced, thereby reducing the algorithm complexity and processing delay of the LDPC-MIMO iterative receiver.

虽然本发明所揭露的实施方式如上，但所述的内容只是为了便于理解本发明而采用的实施方式，并非用以限定本发明。任何本发明所属技术领域内的技术人员，在不脱离本发明所揭露的精神和范围的前提下，可以在实施的形式上及细节上作任何的修改与变化，但本发明的专利保护范围，仍须以所附的权利要求书所界定的范围为准。Although the embodiments disclosed in the present invention are as above, the described content is only an embodiment adopted for the convenience of understanding the present invention, and is not intended to limit the present invention. Anyone skilled in the technical field to which the present invention belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed by the present invention, but the patent protection scope of the present invention, The scope defined by the appended claims must still prevail.

Claims (8)

1. An iterative receiving method for a low density parity check, LDPC, multiple-input multiple-output, MIMO, communication system, comprising:

carrying out MIMO detection on the received signal to obtain an estimated symbol;

performing soft demodulation on the estimated symbol, and performing hard decision on soft bit information generated by the soft demodulation to obtain a hard decision codeword;

performing LDPC decoding iteration once, namely internal iteration, generating a decoding information sequence, a decoding code word and a code element log-likelihood ratio, and updating the internal iteration times;

checking the decoded code word, if the check is successful, outputting the decoded information sequence, and ending LDPC-MIMO iterative decoding, otherwise, judging whether the internal iteration meets a preset internal iteration stop criterion;

when the internal iteration meets the internal iteration stop criterion, outputting the decoding information sequence, the decoding code word and the code element log-likelihood ratio; otherwise, carrying out the next internal iteration;

judging whether iteration between MIMO detection and LDPC decoding, namely outer iteration, meets a preset outer iteration stop criterion or not according to the hard decision codeword and the decoding codeword, if so, finishing LDPC-MIMO iterative decoding, otherwise, updating the outer iteration times, performing soft mapping on the code element log-likelihood ratio, and obtaining a soft mapping modulation symbol to feed back to the MIMO detection;

the step of judging whether the inner iteration meets the inner iteration stop criterion comprises the following steps:

calculating the variance of the absolute value of the code element log-likelihood ratio;

checking the number of internal iterations;

judging whether the variance is greater than a preset variance threshold or not and the internal iteration frequency is less than a preset maximum internal iteration frequency, if so, not meeting the internal iteration stop criterion, otherwise, meeting the internal iteration stop criterion;

the step of determining whether the outer iteration meets the outer iteration stop criterion includes:

for the first outer iteration of the process,

comparing the hard decision code word with the decoding code word, calculating different bit numbers on corresponding bit positions, and obtaining a length ratio reflecting the error degree of LDPC decoding;

judging whether the length ratio is smaller than a preset length ratio threshold or not and whether the external iteration frequency is smaller than a preset maximum external iteration frequency or not, if so, not meeting the external iteration stop criterion, otherwise, meeting the external iteration stop criterion;

for the non-first outer iteration,

comparing the hard decision code word with the decoding code word to further obtain a bit number ratio reflecting the convergence of the external iterative decoding;

and further judging whether the bit number ratio is smaller than a preset bit number ratio threshold, if so, not meeting the external iteration stop criterion, and otherwise, meeting the external iteration stop criterion.

2. The method of claim 1, wherein:

the length ratio represents the ratio of the different bit numbers to the length of the decoded code word;

the bit number ratio represents the ratio of the different bit numbers of the current external iterative decoding relative to the last external iterative decoding.

3. The method of claim 1, wherein performing the MIMO detection on the received signal comprises:

for a first outer iteration, performing the MIMO detection only on the received signals,

for non-first outer iterations, performing the MIMO detection on the received signal and the soft mapped modulation symbols.

4. The method of claim 1, wherein:

and after normalization processing is carried out on the estimated symbols, the soft demodulation is carried out.

5. The method of claim 1, wherein the step of making the hard decision on the soft bit information comprises:

and performing the hard decision on the positive and negative values of the soft bit information.

6. The method of claim 1, wherein:

initially, the number of the inner iterations is 0, and the number of the outer iterations is 1.

7. The method of claim 1, wherein said step of performing said check on said decoded codeword comprises:

using a check equation cH^TPerforming the check on the decoded codeword as 0;

wherein c is the decoding code word, and H is an LDPC check matrix.

8. The method of claim 1, wherein:

the outer iteration comprises a Turbo iteration.