A multi-user joint decoding method for Internet of Things applicable to downlink channels

Abstract

The present invention belongs to the field of wireless communication technology, and specifically discloses an Internet of Things multi-user joint decoding method applicable to a downlink channel, comprising the following steps: based on Polar‑NOMA System model, polarization code construction in multi-user scenario, and polarization code sequence are obtained; polarization code iterative decoder is designed according to polarization code sequence; iterative receiver is designed according to polarization code iterative decoder; based on Polar‑NOMA System Model and Iteration Receiver, Design Polar‑ NOMA The system receiver realizes multi-user joint decoding. The present invention adopts a soft information iterative decoding method, solves the problem that the existing iterative decoding scheme cannot retain and utilize the information from the channel as much as possible, and has a lower bit error rate.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communications, and in particular relates to a design of an Internet of Things multi-user joint decoding method suitable for a downlink channel. Background Art

[0002] With the development of wireless communication technology in the fields of mobile Internet, smart home, industrial control, deep space communication, etc., new application scenarios and demands continue to emerge. This puts higher requirements on the effectiveness and reliability of communication systems. 物联网 , 物联网 ) scenarios, the communication system needs to have low latency, high reliability, and massive access, and the realization of these features requires the support of well-designed channel coding, multiple access and other technologies. As the fifth generation mobile communication (5 第二代移动通信技术 , 5 G ) was selected as the technical solution, polarization code ( 极化码 ), non-orthogonal multiple access ( 非正交多址接入 , 非正交多址接入 ) and other technologies have attracted widespread attention from domestic and foreign academic circles.

[0003] In a broad sense, polar code construction mainly includes sequence generation, rate matching (puncturing or shortening) strategy, etc. In a narrow sense, polar code construction is a sequence generation algorithm. From the perspective of whether it depends on channel conditions such as erasure probability and transition probability, polar code construction algorithms can be mainly divided into two categories. One category is channel-related construction algorithms, such as the construction algorithm for binary erasure channels ( 二进制删除信道 , 二进制删除信道 ) and the extension of polar code construction to 双向离散无记忆信源信道 Channel degradation based on channel 信道降级 ) and the construction method based on 加性高斯白噪声 Gaussian approximation of the channel ( 高斯近似 , 高斯近似 ) construction algorithms, etc.

[0004] The hard-input and hard-output decoding algorithm of polar codes mainly includes 逐次消除 Decoding, 堆栈逐次消除 Decoding algorithm and 堆栈逐次消除 Serial offset stack decoding algorithm for decoding simplification scheme ( 连续消除堆栈 , 连续消除堆栈 ) etc. The soft input soft output decoding algorithm mainly includes the confidence propagation decoding algorithm ( 置信传播 , 置信传播 ) and soft offset decoding algorithm ( 软消除 , 连续消除和累加网络 )wait. Summary of the invention

[0005] The purpose of the present invention is to solve the problem that the existing iterative decoding scheme cannot retain and utilize the information from the channel as much as possible and has a high bit error rate. A multi-user joint decoding method for the Internet of Things suitable for the downlink channel is proposed.

[0006] The technical solution of the present invention is: a multi-user joint decoding method for the Internet of Things applicable to a downlink channel, comprising the following steps:

[0007] S1. Based on 极化 - 非正交多址接入 System model, polarization code construction in multi-user scenario, and polarization code sequence;

[0008] S2. Design a polar code iterative decoder based on the polar code sequence;

[0009] S3. Design an iterative receiver based on the polar code iterative decoder;

[0010] S4. Based on 极化 - 非正交多址接入 System Model and Iteration Receiver, Design 极化 - 非正交多址接入 System receiver, realizing multi-user joint decoding.

[0011] The beneficial effects of the present invention are:

[0012] The present invention adopts a soft information iterative decoding method, which can retain and utilize information from the channel as much as possible and has a lower bit error rate compared to hard decision.

[0013] Preferably, the polarization code construction in step S1 adopts a Gaussian approximation construction algorithm or a polarization weight construction algorithm.

[0014] Preferably, the Gaussian approximation construction algorithm is specifically: calculating the mathematical expectation of the log-likelihood ratio of the original coding channel and the mathematical expectation of the log-likelihood ratio of each sub-channel after channel polarization; and sorting the mathematical expectation values ​​to complete the sequence generation of the polarization code.

[0015] Preferably, the mathematical expectation of the original coding channel log-likelihood ratio is calculated as follows:

[0016]

[0017] The calculation formula of the mathematical expectation of the log-likelihood ratio of each sub-channel after channel polarization is:

[0018]

[0019] in, represents the mathematical expectation of the log-likelihood ratio; Indicates the original coding channel bits of log-likelihood ratio; Indicates the original coded channel bits; represents the noise variance; Represents the polarization code subchannels and The log-likelihood ratio of the sub-channels; represents the code length of the polar code; represents the integral used to calculate the expected value of the square of the hyperbolic tangent function;

[0020] The integral for calculating the expected value of the square of the hyperbolic tangent function The calculation formula is:

[0021]

[0022] Function Approximate processing yields:

[0023]

[0024] in, represents the integration variable.

[0025] Preferably, the polarization weight construction algorithm is specifically as follows: obtaining the weight of each subchannel by calculating the sequence number, sorting the subchannels according to the weights, and completing the sequence generation of the polarization code;

[0026] The weight calculation formula of each sub-channel is:

[0027]

[0028] in, Indicates The weight of each sub-channel; express The maximum number of digits in binary form; express The binary form digits.

[0029] The beneficial effects of the above preferred solution are:

[0030] The polarization weight construction algorithm does not depend on channel parameters, is simple to calculate, has a nested feature, and can derive the polarization code sequence from the existing generated sequence.

[0031] Preferably, the polar code iterative decoder in step S2 adopts 连续消除和累加网络 Design the decoding algorithm;

[0032] Said 连续消除和累加网络 The specific decoding algorithm is:

[0033] S201. 连续相位调制 The external information of the demodulator is input into the polar code iterative decoder, where the external information is the log-likelihood ratio of the original coded channel to the codeword bit. ;

[0034] S202. Log-likelihood ratio of the original coding channel with respect to the codeword bits , and we get the log-likelihood ratio , the log-likelihood ratio It's about raw bits The original bit including information bits and frozen bits, wherein the information bits and frozen bits are selected according to a polar code sequence;

[0035] S203. The frozen bit is converted from the original bit prior L The value is converted to obtain the prior of the codeword bit L value , and the prior L value Initialize and get the initialized prior L value ;

[0036] S204. Based on the log-likelihood ratio and prior L value ,get 连续消除和累加网络 The decoder's final decision.

[0037] As a preference, the a priori after the initialization in step S203 L value The formula is:

[0038]

[0039] in, Indicates frozen bits, ; Represents information bits.

[0040] Preferably, the final decision in step S204 is expressed as:

[0041]

[0042] in, represents the decision value obtained by hard decision; Represents the polarization code The log-likelihood ratio of the sub-channels; Decoder representing the soft-cancellation decoding algorithm.

[0043] The beneficial effects of the above preferred solution are:

[0044] 连续消除和累加网络 The serial decoding characteristics of the decoding algorithm enable it to have a higher convergence speed and channel polarization degree. 连续消除和累加网络 The decoding algorithm transforms the frozen bit information and uses it in the entire decoding process.

[0045] Preferably, the iterative receiver in step S3 is composed of 连续相位调制 The demodulator and the polar code iterative decoder are connected in series through an interleaving operation; 连续相位调制 The demodulator uses 最大后验概率 Criteria for judging.

[0046] The beneficial effects of the above preferred solution are:

[0047] 连续相位调制 The demodulator and the polar code iterative decoder are connected in series through interleaving operations, which can approximate burst errors in the channel as random errors, reduce the correlation of external information between iterative decoding modules, and help reduce the correlation between adjacent bits for high-dimensional modulation systems. In serial mode, the receiver only needs to deploy one iterative receiver to meet the requirements.

[0048] Preferably, in step S4 极化 - 非正交多址接入 System receiver includes 线性最小均方误差 - 连续干扰消除 modules, and multiple 线性最小均方误差 - 连续干扰消除 The iterator sink to which the module is connected.

[0049] The beneficial effects of the above preferred solution are:

[0050] By setting 线性最小均方误差 - 连续干扰消除 The module can achieve a balance between interference suppression and noise amplification based on the principle of minimizing the mean square error between the output signal and the actual signal. At the same time, each time a user signal is detected, only one row of the calculation matrix needs to be updated, thereby shortening the calculation time. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 Shown is a flow chart of an IoT multi-user joint decoding method applicable to a downlink channel. DETAILED DESCRIPTION

[0052] Now, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be understood that the embodiments shown and described in the accompanying drawings are only exemplary and are intended to explain the principles and spirit of the present invention, rather than to limit the scope of the present invention.

[0053] likeFigure 1 As shown, a multi-user joint decoding method for an Internet of Things downlink channel includes the following steps:

[0054] S1. Based on polarization code-non-orthogonal multiple access ( 极化 - 非正交多址接入 ) system model, construct polarization codes in multi-user scenarios, and obtain polarization code sequences;

[0055] S2. Design a polar code iterative decoder based on the polar code sequence;

[0056] S3. Design an iterative receiver based on the polar code iterative decoder;

[0057] S4. Based on 极化 - 非正交多址接入 System Model and Iteration Receiver, Design 极化 - 非正交多址接入 System receiver, realizing multi-user joint decoding.

[0058] In this embodiment, the polar code construction in step S1 adopts a Gaussian approximation construction algorithm or a polarization weight construction algorithm.

[0059] In this embodiment, the Gaussian approximation construction algorithm is specifically: calculating the mathematical expectation of the log-likelihood ratio of the original coding channel and the mathematical expectation of the log-likelihood ratio of each sub-channel after channel polarization; and sorting the mathematical expectation values ​​to complete the sequence generation of the polarization code.

[0060] In this embodiment, the change of the log-likelihood ratio during the channel polarization process is considered, and the polarization code with a code length of 2 is taken as an example. The corresponding log-likelihood ratio is , we can get:

[0061]

[0062] Extending it to the polar code with code length N, we can get:

[0063]

[0064] Among them, Indicates the original coding channel No. Bits The log-likelihood ratio of , we can get: .

[0065] default ,right The mathematical expectation is obtained, and the calculation formula for the mathematical expectation of the original coding channel log likelihood ratio is obtained as follows:

[0066]

[0067] The calculation formula of the mathematical expectation of the log-likelihood ratio of each sub-channel after channel polarization is:

[0068]

[0069] in, represents the mathematical expectation of the log-likelihood ratio; Indicates the original coding channel bits of log-likelihood ratio; Indicates the original coded channel bits; represents the noise variance; Represents the polarization code subchannels and The log-likelihood ratio of the subchannels; represents the code length of the polar code; represents the integral used to calculate the expected value of the square of the hyperbolic tangent function;

[0070] The integral for calculating the expected value of the square of the hyperbolic tangent function The calculation formula is:

[0071]

[0072] Function Approximate processing yields:

[0073]

[0074] in, represents the integration variable.

[0075] In this embodiment, the polarization weight construction algorithm is specifically as follows: the weight of each subchannel is obtained by calculating the sequence number, and the subchannels are sorted according to the weights to complete the sequence generation of the polarization code; the weight calculation formula of each subchannel is:

[0076]

[0077] in, Indicates The weight of each sub-channel; express The maximum number of digits in binary form, ; express The binary form digits.

[0078] In this embodiment, the polar code iterative decoder in step S2 adopts 连续消除和累加网络 Decoding algorithm is designed; 连续消除和累加网络 The specific decoding algorithm is:

[0079] S201. 连续相位调制 The external information of the demodulator is input into the polar code iterative decoder, where the external information is the log-likelihood ratio of the original coded channel to the codeword bit. ;

[0080] S202. Log-likelihood ratio of the original coding channel with respect to the codeword bits , and we get the log-likelihood ratio , the log-likelihood ratio It's about raw bits The original bit including information bits and frozen bits, wherein the information bits and frozen bits are selected according to a polar code sequence;

[0081] S203. The frozen bit is converted from the original bit prior L The value is converted to obtain the prior of the codeword bit L value , and the prior L value Initialize and get the initialized prior L value ;

[0082] S204. Based on the log-likelihood ratio and prior L value ,get 连续消除和累加网络 The decoder's final decision.

[0083] In this embodiment, the a priori after the initialization in step S203 L value The formula is:

[0084]

[0085] in, Indicates frozen bits, ; Represents information bits.

[0086] In this embodiment, the expression formula of the final decision in step S204 is:

[0087]

[0088] in, represents the decision value obtained by hard decision; Represents the polarization code The log-likelihood ratio of the subchannels; Decoder representing the soft-cancellation decoding algorithm.

[0089] In this embodiment, the iterative receiver in step S3 is composed of 连续相位调制 The demodulator and the polar code iterative decoder are connected in series through an interleaving operation; 连续相位调制 The demodulator uses 最大后验概率 Decision criteria. In serial mode, after completing the user k After processing the signal, the iterative receiver will send 连续干扰消除 Module feedback updated Signal parameters to complete the 线性最小均方误差 The calculation of the matrix finally completes the detection of the next user signal and enters a new cycle. Therefore, in the serial mode, the receiving end only needs to deploy one iterative receiver to meet the demand.

[0090] Considering a single bit, the probability of its value being 0 or 1 can be directly calculated by or To facilitate representation and calculation, we can directly use and The logarithmic form of the ratio The distribution of is represented by the log probability ratio, and the log probability ratio is The corresponding L Values ​​are:

[0091] .

[0092] 极化 - 非正交多址接入 The system adopts a serial iterative decoding design. Respectively 连续相位调制 The demodulator obtains the information input from the channel and the preset frozen bit information known by the polar code decoder. Respectively 连续相位调制 The demodulator and polar code decoder receive information input from each other. Respectively 连续相位调制 Decision information of demodulator and polar code decoder, Respectively 连续相位调制 The demodulator and polar code decoder output the external information obtained from each other.

[0093] The information iteration process in the receiver is as follows: 连续相位调制 Both the demodulator and polar code decoder use 最大后验概率 Criteria for judging, definition For the posterior L Decision information expressed in value form ,because 最小移频键控 and 高斯最小移频键控All belong to binary modulation (i.e., a single symbol carries a 1-bit codeword), with:

[0094]

[0095] Since at the receiving end The value of has been determined, and it is decomposed by the Bayesian formula to obtain:

[0096]

[0097] in, Respectively and of L Value form. Correspondingly, the polar code decoder decision information It can be expressed as:

[0098] .

[0099] Consider the initial stage of the iteration, when 连续相位调制 The demodulator receives only information from the channel , The form is about On the other hand, the polar code decoder does not get the log-likelihood ratio from 连续相位调制 The demodulator only has information about the preset frozen bits. ,and The information is in the form of Prior L During the iterative decoding process, the CPM demodulator and polar code decoder need to share the information they have mastered with each other.

[0100] In this embodiment, the step S4 极化 - 非正交多址接入 System receiver includes 线性最小均方误差 - 连续干扰消除 modules, and multiple 线性最小均方误差 - 连续干扰消除 The receiving end performs an iterative receiver connected to the module. 线性最小均方误差 - 连续干扰消除 Processing, you can get the user k Estimated value of the signal ,in, Indicates user k No. The estimated value of a signal.

[0101] In the detection After that, the iterative receiver will receive the estimated value as input, perform multiple iterative calculations until the preset number of iterations is reached, and finally output the decision value.

[0102] The present invention is simulated and verified. and Gaussian constructed sequence with a length of 1024 codes under the condition. The larger the value, the stronger the noise impact, and the lower the channel reliability; The smaller the value, the weaker the noise impact, and the higher the channel reliability. , under the condition of code rate 0.5, due to the short code length and insufficient channel polarization of the 1024-code length polar code, some information bits still need to be transmitted through non-ideal sub-channels; with the increase of code length, the channel polarization phenomenon is stronger, so all information bits of the 32768-code length polar code can be transmitted through relatively ideal sub-channels.

[0103] By simulating the polarization weight construction algorithm of 1024 code length, the nested characteristics of the polarization weight construction sequence can be obtained, that is, the short code length polarization weight sequence can be regarded as being included in the long code length polarization weight sequence.

[0104] By using polarization code 连续消除和累加网络 The algorithm is compared with the polar code SC decoding algorithm. Under the condition of code length 8192, 逐次消除 and 连续消除和累加网络 The frame error rate curves of the decoding algorithms are very close. On the other hand, 连续消除和累加网络 The decoding only requires one iteration to obtain a close 逐次消除 The performance of the decoding algorithm shows 连续消除和累加网络 The decoding algorithm has the advantage of fast convergence. Under the shorter 1024 code length, 连续消除和累加网络 Decoding 逐次消除 The decoding algorithm has certain performance improvements. 连续消除和累加网络 Decoding algorithm 逐次消除 The biggest advantage of the decoding algorithm is that it can output soft information, thereby providing conditions for information exchange and iterative reception.

[0105] The total number of iterations between the demodulator and decoder in the iterative receiver is shown as I With the increase of , the system’s bit error rate performance has been significantly improved. Through 16 soft information iterations, iterative decoding is better than hard decision decoding scheme in 误码率 For 10 -4 Under these conditions, a gain of about 1.1dB was obtained.

[0106] For code length 512, number of access users is 10, number of receiving antennas is 6, frame error rates of different iteration numbers under code rate 0.5 and code rate 0.75 are compared. Compared with one iteration, the frame error rate curve can enter the waterfall area faster under the conditions of 5 and 10 iterations.

[0107] Through simulation results, it can be concluded that compared with hard decision, the iterative decoding method has a lower bit error rate. Through the mutual transmission of soft information between modules, the iterative decoding algorithm retains and utilizes the information from the channel as much as possible.

[0108] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A multi-user joint decoding method for the Internet of Things applicable to downlink channels, characterized in that: The following steps are involved: S1. Based on Polar - NOMA A system model is provided, and polarization codes are constructed in a multi-user scenario to obtain a polarization code sequence; wherein the polarization code is constructed by using a Gaussian approximation construction algorithm or a polarization weight construction algorithm; S2. According to the polar code sequence, a polar code iterative decoder is designed; the polar code iterative decoder adopts SCAN Design the decoding algorithm; Said SCAN The specific decoding algorithm is: S201. CPM The external information of the demodulator is input into the polar code iterative decoder, where the external information is the log-likelihood ratio of the original coded channel to the codeword bit. ; S202. Log-likelihood ratio of the original coding channel with respect to the codeword bits , and we get the log-likelihood ratio , the log-likelihood ratio It's about raw bits The original bit including information bits and frozen bits, wherein the information bits and frozen bits are selected according to a polar code sequence; S203. The frozen bit is converted from the original bit prior L The value is converted to obtain the prior of the codeword bit L value , and the prior L value Initialize and get the initialized prior L value ; S204. Based on the log-likelihood ratio and prior L value ,get SCAN The decoder's final verdict; The priori after initialization in step S203 L value The formula is: in, Indicates frozen bits, ; Represents information bits; The expression formula of the final judgment in step S204 is: in, represents the decision value obtained by hard decision; Represents the polarization code The log-likelihood ratio of the subchannels; A decoder representing a soft-offset decoding algorithm; S3. Design an iterative receiver according to the polar code iterative decoder; the iterative receiver is composed of CPM The demodulator and the polar code iterative decoder are connected in series through an interleaving operation; CPM The demodulator uses MAP Criteria for sentencing; S4. Based on Polar - NOMA System Model and Iteration Receiver, Design Polar - NOMA The system receiver realizes multi-user joint decoding; Polar - NOMA System receiver includes LMMSE - SIC modules, and multiple LMMSE - SIC The iterator sink to which the module is connected.

2. The IoT multi-user joint decoding method applicable to downlink channels according to claim 1 is characterized in that: The Gaussian approximation construction algorithm specifically comprises: calculating the mathematical expectation of the log-likelihood ratio of the initial coding channel and the mathematical expectation of the log-likelihood ratio of each sub-channel after channel polarization; and sorting the mathematical expectation values ​​to complete the sequence generation of the polarization code.

3. The IoT multi-user joint decoding method applicable to downlink channels according to claim 2 is characterized in that: The calculation formula of the mathematical expectation of the initial coding channel log-likelihood ratio is: The calculation formula of the mathematical expectation of the log-likelihood ratio of each sub-channel after channel polarization is: in, represents the mathematical expectation of the log-likelihood ratio; Indicates the original coding channel bits of log-likelihood ratio; Indicates the original coded channel bits; represents the noise variance; Represents the polarization code subchannels and The log-likelihood ratio of the subchannels; represents the code length of the polar code; Represents the integral used to compute the expected value of the square of the hyperbolic tangent function.

4. The IoT multi-user joint decoding method applicable to downlink channels according to claim 1, characterized in that: The polarization weight construction algorithm is specifically as follows: obtaining the weight of each subchannel through sequence number calculation, sorting the subchannels according to the weights, and completing the sequence generation of the polarization code; The weight calculation formula of each sub-channel is: in, Indicates The weight of each sub-channel; express The maximum number of digits in binary form; express The binary form digits.

Images