Decoding method, device, equipment and medium

Abstract

The invention discloses a decoding method, a decoding device, decoding equipment and a medium. The method comprises the following steps: acquiring a current log-likelihood ratio corresponding to a to-be-decoded signal and a corresponding historical log-likelihood ratio; comparing the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio, and determining a corresponding first quantization factor and a corresponding second quantization factor for the current log-likelihood ratio and the historical log-likelihood ratio according to the reliability difference; and quantizing the current log-likelihood ratio by using the first quantization factor, quantizing the historical log-likelihood ratio by using the second quantization factor, and performing merging and decoding processing based on the two quantized log-likelihood ratios. According to the method provided by the invention, the adverse effect of the log-likelihood ratio with relatively low reliability on the decoding performance can be avoided, and the improvement effect of the log-likelihood ratio with relatively high reliability on the decoding performance is improved.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a decoding method, device, equipment and medium. Background Art

[0002] Hybrid Automatic Repeat Request (HARQ) technology is a technology that combines Automatic Repeat Request (ARQ) and Forward Error Correction (FEC). HARQ merging is to merge soft bit data of different redundancy versions (RV), specifically, to merge and decode the historical LLR (Log-Likelihood Ratio) and the current LLR, which can increase the incremental redundant information of the data to improve the success rate of data decoding.

[0003] However, in high-speed scenarios, there may be a large difference in reliability between the historical LLR and the current LLR. At this time, using the historical LLR and the current LLR for combined decoding will increase the adverse impact of the LLR with higher reliability on the decoding performance and reduce the improvement effect of the LLR with higher reliability on the decoding performance. Therefore, a decoding method that improves reliability and ensures decoding performance is urgently needed. Summary of the invention

[0004] The present invention provides a decoding method, device, equipment and medium, aiming to effectively solve the above technical problems.

[0005] According to a first aspect of the present invention, the present invention provides a decoding method, comprising:

[0006] Obtaining a current log-likelihood ratio and a corresponding historical log-likelihood ratio corresponding to the signal to be decoded;

[0007] Comparing the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio, and determining a corresponding first quantization factor and a second quantization factor for the current log-likelihood ratio and the historical log-likelihood ratio according to the reliability difference;

[0008] The current log-likelihood ratio is quantized using the first quantization factor, the historical log-likelihood ratio is quantized using the second quantization factor, and the two quantized log-likelihood ratios are combined and decoded.

[0009] According to a second aspect of the present invention, the present invention further provides a decoding device, comprising:

[0010] An information acquisition unit, used to acquire a current log-likelihood ratio and a corresponding historical log-likelihood ratio corresponding to a signal to be decoded;

[0011] a quantization factor determination unit, configured to compare the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio, and determine a first quantization factor and a second quantization factor corresponding to the current log-likelihood ratio and the historical log-likelihood ratio according to the reliability difference;

[0012] A decoding unit is used to quantize the current log-likelihood ratio using the first quantization factor, quantize the historical log-likelihood ratio using the second quantization factor, and combine and decode the two quantized log-likelihood ratios.

[0013] According to a third aspect of the present invention, the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the steps of any one of the above methods.

[0014] According to a fourth aspect of the present invention, the present invention further provides a storage medium, in which a plurality of instructions are stored, and the instructions are suitable for being loaded by a processor to execute the steps of any of the above methods.

[0015] Through one or more of the above embodiments of the present invention, at least the following technical effects can be achieved: by comparing the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio, the corresponding first quantization factor and the second quantization factor are determined, and the current log-likelihood ratio and the historical log-likelihood ratio are quantized based on the first quantization factor and the second quantization factor to reduce the reliability difference between the two. Then, the quantized log-likelihood ratios are combined and decoded. By reducing the reliability difference, the adverse effects of the log-likelihood ratio with lower reliability on the decoding performance are avoided, and the improvement effect of the log-likelihood ratio with higher reliability on the decoding performance is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The technical solutions and other beneficial effects of the present invention will be made apparent by describing in detail the specific embodiments of the present invention in conjunction with the accompanying drawings.

[0017] Figure 1 One of the flowcharts of the decoding method provided by the embodiment of the present invention;

[0018] Figure 2 The second flowchart of the decoding method provided by the embodiment of the present invention;

[0019] Figure 3 A schematic diagram of the sub-step flow chart of step S25 provided in an embodiment of the present invention;

[0020] Figure 4 A schematic diagram of the structure of a decoding device provided by an embodiment of the present invention;

[0021] Figure 5 A schematic diagram of the structure of a decoding system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The technical scheme in the embodiment of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. Based on the embodiment of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0023] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the term "and / or" herein is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " herein, unless otherwise specified, generally indicates that the associated objects before and after are in an "or" relationship.

[0024] The HARQ protocol is divided into two categories in the time domain: synchronous and asynchronous; in the frequency domain, it is divided into two categories: adaptive and non-adaptive. Among them, asynchronous HARQ (Asynchronous HARQ) means that retransmission can occur at any time, and it also means that the HARQ process can be used in any order. Asynchronous HARQ makes retransmission scheduling more flexible. Adaptive HARQ (Adaptive HARQ) means that the physical resource block (PRB) resources and the modulation and coding scheme (MCS) used for retransmission can be changed.

[0025] In mobile communication systems, especially in scenarios where trains move at high speeds (500km / h), there is an obvious Doppler effect in the information transmission process. The Doppler effect refers to the phenomenon that the frequency or wavelength of a wave changes for an observer moving relative to the wave source, and this frequency change is called Doppler frequency offset (Frequency Offset, referred to as FO). FO affects the uplink access success rate, the switching success rate, and has an impact on the capacity and coverage of the communication system. To address this problem, the base station uses the uplink pilot signal to estimate the frequency offset value, and then achieves frequency synchronization through two processes of inter-symbol and intra-symbol frequency offset compensation. However, decoding and compensation are more or less inaccurate, which is more likely to cause decoding errors in high-speed scenarios. The uplink channel needs to be combined with HARQ technology to request retransmission of signals with decoding errors in high-speed scenarios, and merge soft bit data of different RVs to improve the success rate of data decoding.

[0026] At present, HARQ merging is to directly store the merged LLR into the HARQ buffer for each subframe, and directly read the historical LLR from the HARQ buffer after the last update before decoding the next retransmitted subframe, and send it to the LDEC (Low Density Parity Check Decoder) for merged decoding together with the LLR output by the current subframe processing. In this process, the historical LLR and the current LLR adopt the same level of quantization. However, there is a large difference in the reliability of LLR after decoding and compensation in high-speed scenarios. When HARQ merge decoding is performed, if the same level of quantization is still used for merging, the adverse effect of low-reliability LLR on decoding performance will be increased, and the improvement effect of high-reliability LLR on decoding performance will be reduced. Therefore, the present invention provides a decoding method, device, equipment and medium.

[0027] First, the nouns in the present invention are defined as follows:

[0028] Channel Estimation (CE) is used to estimate frequency offset FO and signal-to-noise ratio SNR.

[0029] Frequency Domain Equalization (FDEQ) is a signal processing technique that suppresses additive noise and transmission distortion by processing the signal in the frequency domain. By applying an equalizer in the frequency domain, the impact of multipath effects on the signal can be reduced and the reception quality of the signal can be improved.

[0030] The demultiplexing demodulator and descrambler DMAP (Demapper) can perform demodulation, descrambling and demultiplexing processing on the equalized data to obtain the soft bit information required for decoding.

[0031] Log-Likelihood Ratio (LLR) is a statistic used in signal detection and decoding to measure the similarity between the received signal and the hypothesized signal. In channel coding and decoding, LLR is often used in soft decision decoding algorithms to help improve decoding accuracy.

[0032] Doppler frequency offset (FO) refers to the change in frequency or wavelength of a wave relative to an observer moving relative to the wave source.

[0033] Signal-to-Noise Ratio (SNR) is a measure of signal strength relative to the background noise level. In communication systems, a high SNR usually means better signal quality and higher reliability of data transmission. SNR is one of the key factors that determine the performance of wireless communication systems, especially in adaptive systems. Transmission parameters such as modulation mode and coding rate can be adjusted according to the SNR to optimize transmission efficiency and reliability.

[0034] Low Density Parity Check Decoder (LDEC) is an efficient energy error correction decoder LDEC, which works based on low-density parity check codes (LDPC codes). LDPC codes are linear block codes defined by a sparse parity check matrix, and their error correction capability is very close to the theoretical maximum, i.e., the Shannon limit. LDPC decoder LDEC minimizes transmission errors and optimizes the reliability of data transmission through iterative algorithms such as the Belief Propagation algorithm or its variants.

[0035] A decoding method, device, equipment and medium provided by the present invention are described below in conjunction with the accompanying drawings.

[0036] Figure 1 One of the flowcharts of the decoding method provided by the embodiment of the present invention is as follows Figure 1 As shown, a decoding method comprises the following steps:

[0037] S11, obtaining a current log-likelihood ratio and a corresponding historical log-likelihood ratio corresponding to the signal to be decoded.

[0038] In this step, the current log-likelihood ratio is calculated based on the soft bit information to be decoded obtained after channel estimation of the uplink channel in the wireless communication and equalization, demodulation and descrambling. In addition, the historical decoded signal is determined from the HARQ information buffer area according to the user index number, system frame number, subframe number and other parameters of the signal to be decoded, and the log-likelihood ratio of the historical decoded signal is used as the historical log-likelihood ratio.

[0039] S12, comparing the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio, and determining corresponding first quantization factors and second quantization factors for the current log-likelihood ratio and the historical log-likelihood ratio according to the reliability difference.

[0040] In this step, the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio can be evaluated from the dimensions of modulation order, frequency deviation, signal-to-noise ratio, channel estimation pulse diagram, equalization constellation diagram, etc.

[0041] Among them, the modulation order is determined based on the modulation method of the signal to be decoded. The higher the modulation order, the higher the decoding difficulty, and the worse the corresponding reliability of the log-likelihood ratio; conversely, the lower the modulation order, the lower the decoding difficulty, and the better the corresponding reliability of the log-likelihood ratio.

[0042] The larger the frequency deviation, the more difficult it is to correct the signal to be decoded after estimation and compensation, and accordingly, the worse the reliability of the log-likelihood ratio. Conversely, the smaller the frequency deviation, the easier it is to correct the signal to be decoded through estimation and compensation, and the better the reliability of the log-likelihood ratio.

[0043] The larger the signal-to-noise ratio, the higher the quality of the signal to be decoded, the smaller the impact of the background noise, and the better the reliability of the log-likelihood ratio; conversely, the smaller the signal-to-noise ratio, the lower the quality of the signal to be decoded, the greater the impact of the background noise, and the worse the reliability of the log-likelihood ratio.

[0044] When the reliability difference is large, the log-likelihood ratio with poor reliability is quantized and attenuated to a greater extent, that is, more low-order bit information is discarded in the log-likelihood ratio with poor reliability; the log-likelihood ratio with good reliability can be kept without quantization and retain the original value, or fewer low-order bits can be discarded to reduce the reliability difference between the two quantized log-likelihood ratios.

[0045] The numerical values ​​of the first quantization factor and the second quantization factor are determined according to the reliability difference. If the current log-likelihood ratio is more reliable, the first quantization factor is greater than the second quantization factor; if the historical log-likelihood ratio is more reliable, the first quantization factor is less than the second quantization factor, so that the log-likelihood ratio with poor reliability is attenuated to a greater extent. It should be noted that if the log-likelihood ratio is an N-bit binary number, the log-likelihood ratio includes 2 N The quantization factor is the value of 2 N The larger the quantization factor, the higher the retention of the value. NThe more bits are retained in the value, the greater the impact of the log-likelihood ratio after quantization by the quantization factor on the merged decoding; correspondingly, the smaller the quantization factor, the fewer bits are retained, and the smaller the impact on the merged decoding. Schematically, taking the 8-bit log-likelihood ratio of 127 as an example, the binary form of the log-likelihood ratio is "01111111". If the quantization factor is 1, all bits need to be retained, and after multiplying the quantization factor and the log-likelihood ratio, it is still "01111111". When the quantization factor is , it means that the low 3 bits of "01111111" need to be discarded, and finally it becomes "01111", which is also equivalent to

[0046] S13, quantizing the current log-likelihood ratio using the first quantization factor, quantizing the historical log-likelihood ratio using the second quantization factor, and combining and decoding the two quantized log-likelihood ratios.

[0047] In this step, the current log-likelihood ratio is quantized using the first quantization factor to obtain the quantized current log-likelihood ratio, and the historical log-likelihood ratio is quantized using the second quantization factor to obtain the quantized historical log-likelihood ratio. The quantized current log-likelihood ratio and the quantized historical log-likelihood ratio are combined, and the combined log-likelihood ratio is decoded to obtain a decoding result.

[0048] After decoding is completed, the current log-likelihood ratio and its related information are also stored in the buffer as a new historical log-likelihood ratio to facilitate decoding processing at the next moment.

[0049] The decoding method provided by the embodiment of the present invention determines the corresponding first quantization factor and second quantization factor by comparing the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio, and quantizes the current log-likelihood ratio and the historical log-likelihood ratio based on the first quantization factor and the second quantization factor to reduce the reliability difference between the two, and then merges and decodes the quantized log-likelihood ratios. By reducing the reliability difference, the adverse effects of the log-likelihood ratio with lower reliability on the decoding performance are avoided, and the improvement effect of the log-likelihood ratio with higher reliability on the decoding performance is improved.

[0050] In some embodiments of the present invention, since the modulation order, frequency offset FO and signal-to-noise ratio SNR have a great influence on the decoding difficulty difference, the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio is evaluated based on the modulation order, frequency offset and signal-to-noise ratio among many factors.

[0051] Among them, the greater the difference between the modulation orders, the greater the corresponding decoding difficulty. Specifically, QPSK needs to distinguish 4 signal points, 16QAM needs to distinguish 16 signal points, 64QAM needs to distinguish 64 signal points, and 256QAM needs to distinguish 256 signal points. It can be seen that as the number of signal points that need to be distinguished increases, the corresponding decoding difficulty also increases exponentially step by step. Therefore, different modulation methods have a greater impact on the reliability of the log-likelihood ratio.

[0052] The greater the difference between SNRs, the greater the corresponding decoding difficulty. Since SNR is the ratio of signal to noise, the higher the SNR, the less likely the signal is to be interfered or masked by noise, and the easier it is to decode. Conversely, the smaller the SNR, the more difficult it is to decode.

[0053] The greater the difference between FOs, the greater the corresponding difference in decoding difficulty. This is because after the frequency offset of the signal to be decoded is estimated, the FO is obtained, and then the frequency offset can be compensated through recovery, decoding, etc. The larger the estimated FO, the larger the compensated FO, but the corresponding performance degradation is also greater, and the decoding is more prone to errors; conversely, the smaller the FO, the less likely the decoding is to make errors.

[0054] Furthermore, the current log-likelihood ratio corresponds to a first modulation order, a first frequency offset, and a first signal-to-noise ratio, and the historical log-likelihood ratio corresponds to a second modulation order, a second frequency offset, and a second signal-to-noise ratio.

[0055] Among them, the first modulation order is determined according to the modulation mode of the signal to be decoded, which corresponds to the current log-likelihood ratio, and different modulation modes correspond to a modulation order. Modulation modes include QPSK (Quadrature Phase ShiftKeying), 16QAM (Quadrature Amplitude Modulation), 64QAM, 256QAM, etc. Correspondingly, different modulation modes make the bit information carried by each symbol 2 bits, 4 bits, 6 bits, and 8 bits respectively, and the corresponding modulation order also increases successively. The specific modulation order can be set to 2, 4, 6, 8 according to the number of bit information carried, or it can be set to 1, 2, 3, 4, or it can be set to 4, 16, 64, 256, etc., and there is no specific limitation on this.

[0056] The first frequency offset is obtained by frequency offset estimation of the pilot signal in the signal to be decoded, and the first signal-to-noise ratio is obtained after the signal to be decoded is processed by FDEQ. The second modulation order, the second frequency offset and the second signal-to-noise ratio are located in the buffer area and stored corresponding to the historical log-likelihood ratio.

[0057] Compare the reliability differences between the current log-likelihood ratio and the historical log-likelihood ratio, including:

[0058] When the first modulation order is greater than or equal to the preset order and the first frequency deviation is greater than or equal to the preset frequency deviation threshold, the first modulation order is compared with the second modulation order, the first frequency deviation, the second frequency deviation and the frequency deviation threshold to determine the type of reliability difference.

[0059] When the first modulation order is less than the preset order and the first frequency deviation is less than the frequency deviation threshold, compare the first modulation order with the second modulation order, the first frequency deviation, the second frequency deviation and the frequency deviation threshold, the first signal-to-noise ratio with the second signal-to-noise ratio to determine the type of reliability difference.

[0060] In this embodiment, since among the modulation order, frequency deviation and signal-to-noise ratio, when the modulation order is higher, the performance difference is more obvious, but in the context of large modulation order differences, in order to ensure that the performance difference is more obvious, it is necessary to add frequency deviation to determine the reliability difference between the log-likelihood ratios from the two dimensions of modulation order and frequency deviation. When the modulation order is lower, the performance difference is less obvious, even if the frequency deviation is considered, it is not obvious enough, so it is necessary to add the signal-to-noise ratio to further determine the reliability difference between the log-likelihood ratios.

[0061] In addition, the type of reliability difference refers to the degree of reliability difference. The larger the difference, the more bits are needed to quantize the attenuation. Conversely, the smaller the difference, the smaller the number of bits are needed to quantize the attenuation. The frequency deviation threshold can be set based on experience, actual conditions, etc. Similarly, the order threshold can also be adjusted based on simulation experience, measured performance in actual scenarios, etc., and there is no limitation on this.

[0062] Accordingly, determining a first quantization factor and a second quantization factor corresponding to the current log-likelihood ratio and the historical log-likelihood ratio according to the reliability difference includes:

[0063] According to the type of reliability difference, the first quantization factor is determined by the current log-likelihood ratio, and the second quantization factor is determined by the historical log-likelihood ratio.

[0064] Illustratively, case 1: if the reliability corresponding to the current log-likelihood ratio is much better than that of the historical log-likelihood ratio, then the historical log-likelihood ratio needs more bits to be quantized and attenuated than the current log-likelihood ratio, and the first quantization factor is greater than the second quantization factor. Case 2: if the reliability corresponding to the current log-likelihood ratio is slightly better than that of the historical log-likelihood ratio, then the historical log-likelihood ratio needs more bits to be quantized and attenuated than the current log-likelihood ratio, but less than the historical log-likelihood ratio in case 1, that is, the second quantization factor in case 2 is greater than the second quantization factor in case 1.

[0065] The decoding method provided by the embodiment of the present invention subdivides the reliability difference by comparing the modulation order, frequency deviation and signal-to-noise ratio, and determines different quantization factors according to the different reliability differences, thereby reducing the reliability difference between the current log-likelihood ratio and the historical likelihood ratio. In addition, different quantization factors are used to avoid the adverse effects of the log-likelihood ratio with poor reliability on decoding, and to enhance the improvement effect of the log-likelihood ratio with high reliability on decoding performance.

[0066] In some embodiments of the present invention, under the condition that the first modulation order is greater than or equal to a preset order and the first frequency offset is greater than or equal to a preset frequency offset threshold, comparing the first modulation order with the second modulation order, the first frequency offset, the second frequency offset and the frequency offset threshold to determine the type of reliability difference includes:

[0067] When the second modulation order is greater than or equal to the preset order and the second frequency deviation is greater than or equal to the frequency deviation threshold, the reliability difference is determined to be the first difference. At this time, it means that the current log-likelihood ratio and the historical log-likelihood ratio are both high-order modulation and high-frequency deviation. Although the reliability of both is poor, the difference between the two is small, so the number of bits required for quantization is also small, so the current reliability difference is determined as the first difference with a smaller number of quantization bits.

[0068] When the second modulation order is greater than or equal to the preset order and the second frequency deviation is less than the frequency deviation threshold, the reliability difference is determined to be the second difference. At this time, the current log-likelihood ratio is a high-order high-frequency deviation, and the historical log-likelihood ratio is a high-order low-frequency deviation, then the historical log-likelihood ratio is more reliable than the current log-likelihood ratio, and the difference between the two is larger than the above-mentioned difference of both being high-order high-frequency deviations, and the number of bits required to be quantized is larger, therefore, the current reliability difference is determined to be the second difference.

[0069] When the second modulation order is less than the preset order, the reliability difference is determined to be the third difference. At this time, the current log-likelihood ratio is a high-order high-frequency deviation, and the historical log-likelihood ratio is a low-order, then regardless of whether the second frequency deviation is greater than the frequency deviation threshold, the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio is large, and the current reliability difference is determined to be the third difference.

[0070] It should be noted that the first difference, the second difference and the third difference increase successively, and the number of bits required for quantization also increases successively. In addition, in the embodiment of the present invention, the reliability difference is divided into the first difference, the second difference and the third difference. In other embodiments, the reliability difference can be divided into 4 levels, 5 levels, etc. by setting the first frequency deviation threshold, the second frequency deviation threshold, the third frequency deviation threshold, etc. to further refine the number of bits required for quantization attenuation, and there is no limitation on this.

[0071] In some embodiments of the present invention, under the condition that the first modulation order is less than a preset order and the first frequency offset is less than a frequency offset threshold, comparing the first modulation order with the second modulation order, the first frequency offset, the second frequency offset and the frequency offset threshold, and the first signal-to-noise ratio with the second signal-to-noise ratio to determine the type of reliability difference includes:

[0072] When the second modulation order is greater than or equal to the preset order and the second frequency deviation is greater than or equal to the frequency deviation threshold, the reliability difference is determined to be the third difference. At this time, the current log-likelihood ratio is a low order and a low frequency deviation, while the historical log-likelihood ratio is a high order and a high frequency deviation, then the reliability difference is large, and the current reliability difference is determined to be the third difference.

[0073] When the second modulation order is less than the preset order and the second frequency deviation is less than the frequency deviation threshold, the signal-to-noise ratio difference between the signal-to-noise ratio and the second signal-to-noise ratio is calculated. When the current log-likelihood ratio and the historical log-likelihood ratio are both low-order and low-frequency deviation, the reliability of both is high. In order to further improve the improvement effect of the log-likelihood ratio with better reliability on decoding, it is necessary to introduce the signal-to-noise ratio to more accurately determine the number of bits to be quantized, so as to make full use of the log-likelihood ratio with better reliability. Specifically:

[0074] When the signal-to-noise ratio difference is within a first SNR threshold range, the reliability difference is determined to be a first difference.

[0075] When the signal-to-noise ratio difference is within a second SNR threshold range, the reliability difference is determined to be a second difference.

[0076] When the signal-to-noise ratio difference is within a third SNR threshold range, the reliability difference is determined to be a third difference.

[0077] When the signal-to-noise ratio difference is within a fourth SNR threshold range, the reliability difference is determined to be a fourth difference.

[0078] Among them, the first SNR threshold range, the second SNR threshold range, the third SNR threshold range and the fourth SNR threshold range are adjacent to each other in sequence and their respective maximum critical values ​​increase in sequence. Schematically, the first SNR threshold range, the second SNR threshold range, the third SNR threshold range and the fourth SNR threshold range are: (3,6], (6,9], (9,12], (12,+∞), and can also be [2,4), [4,6), [6,8), [8,+∞), etc. The specific SNR threshold range can be adjusted and determined based on experience or decoding effect, and there is no limitation on this. In addition, in this embodiment, the signal-to-noise ratio difference is divided into 4 SNR threshold ranges. In other embodiments of the present invention, it can also be divided more finely, divided into 5, 6, or only divided into 3, and the present invention does not limit this.

[0079] The first difference, the second difference, the third difference, and the fourth difference increase in sequence, which means that the number of bits required for quantization attenuation also increases in sequence. The specific number of bits can be set according to the actual situation. Considering that the number of bits of the log-likelihood ratio is 8, and too much quantization also affects the decoding effect, therefore, preferably, the number of bits required for quantization of the first difference, the second difference, the third difference, and the fourth difference are 1, 2, 3, and 4, respectively. The following examples are given by taking the number of bits required for quantization of the first difference, the second difference, the third difference, and the fourth difference as 1, 2, 3, and 4, respectively. In other embodiments of the present invention, the corresponding number of bits can also be 2, 3, 4, 5, and other different combinations, which are not limited to this. In addition, in the embodiment of the present invention, the reliability difference is divided into 4 types, and in other embodiments of the present invention, it can also be divided into 3, 5, and other different situations according to the amount of SNR threshold range. In the case of 3, the number of bits required for quantization of the first difference, the second difference, and the third difference can be 2, 4, 6, or 1, 2, 3, and other combinations, which are not limited to this.

[0080] In some embodiments of the present invention, determining the first quantization factor based on the current log-likelihood ratio and determining the second quantization factor based on the historical log-likelihood ratio according to the type of the reliability difference includes:

[0081] When the second modulation order is greater than or equal to the preset order and the second frequency deviation is greater than or equal to the frequency deviation threshold:

[0082] 1. If the first frequency offset is greater than the second frequency offset, it is determined that the reliability of the current log-likelihood ratio is poorer than the historical log-likelihood ratio and more quantization is required. Therefore, the first difference is used to calculate the first quantization factor. Based on the above-mentioned indication that the number of quantization bits corresponding to the first difference is 1, the first quantization factor is To achieve a quantization attenuation of one bit of the current log-likelihood ratio. At the same time, the second quantization factor is 1, so that the historical log-likelihood ratio remains the same. In other embodiments, if the number of bits to be quantized is 2, the first quantization factor is If the number of bits to be quantized is 3, then the first quantization factor is The value of the first quantization factor under different bit numbers is obtained in this way.

[0083] 2. If the first frequency offset is less than the second frequency offset, it is determined that the reliability of the historical log-likelihood ratio is poorer than the current log-likelihood ratio, and more quantitative attenuation is required. Therefore, the second quantization factor is obtained by using the second difference calculation. Similarly, the second quantization factor is At the same time, the first quantization factor is determined to be 1 so that the current log-likelihood ratio remains unchanged.

[0084] It should be noted that when the first frequency offset is equal to the second frequency offset, it means that there is not much reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio. Therefore, it can be determined that the first quantization factor and the second quantization factor are both 1 to retain all bit information in the current log-likelihood ratio and the historical log-likelihood ratio.

[0085] When the second modulation order is greater than or equal to the preset order and the second frequency deviation is less than the frequency deviation threshold, the reliability of the current log-likelihood ratio is poorer than the historical log-likelihood ratio and more quantization attenuation is required. Therefore, the first quantization factor is obtained by using the second difference calculation. The first quantization factor is And determine the second quantization factor to be 1.

[0086] When the second modulation order is less than the preset order, the reliability of the current log-likelihood ratio is poorer than the historical log-likelihood ratio, and more quantization attenuation is required. Therefore, the first quantization factor is obtained by using the third difference calculation. The first quantization factor is And determine the second quantization factor to be 1.

[0087] The decoding method provided by the embodiment of the present invention determines which of the two log-likelihood ratios has a relatively better reliability by comparing the modulation order and frequency deviation with the historical log-likelihood ratios when the current log-likelihood ratio is of high order and high frequency deviation, and determines quantization factors of different sizes by subdividing the reliability difference. The reliability difference is reduced by quantizing the attenuation of the two quantization factors on the respective log-likelihood ratios, thereby increasing the improvement effect of the log-likelihood ratio with high reliability on the decoding performance and reducing the adverse effect of the log-likelihood ratio with low reliability on the decoding performance, thereby ultimately improving the reliability of the communication system.

[0088] In some embodiments of the present invention, determining the first quantization factor based on the current log-likelihood ratio and determining the second quantization factor based on the historical log-likelihood ratio according to the type of the reliability difference includes:

[0089] When the second modulation order is greater than or equal to the preset order and the second frequency deviation is greater than or equal to the frequency deviation threshold, the reliability of the historical log-likelihood ratio is poorer than the current log-likelihood ratio and more quantization attenuation is required. Therefore, the second quantization factor is obtained by using the third difference calculation. The second quantization factor is And determine the first quantization factor to be 1.

[0090] When the signal-to-noise ratio difference is within the first SNR threshold range, if the first signal-to-noise ratio is greater than the second signal-to-noise ratio, it is determined that the historical log-likelihood ratio needs more quantization attenuation, and the second quantization factor is obtained by using the first difference calculation, and the second quantization factor is And determine the first quantization factor to be 1. If the first signal-to-noise ratio is less than the second signal-to-noise ratio, it is determined that the current log-likelihood ratio needs more quantization attenuation, and the first quantization factor is obtained by using the first difference calculation. The first quantization factor is And determine the second quantization factor to be 1.

[0091] When the signal-to-noise ratio difference is within the second SNR threshold range, if the first signal-to-noise ratio is greater than the second signal-to-noise ratio, it is determined that the historical log-likelihood ratio needs more quantization attenuation, and the second quantization factor is obtained by using the second difference calculation, and the second quantization factor is And determine the first quantization factor to be 1. If the first signal-to-noise ratio is less than the second signal-to-noise ratio, it is determined that the current log-likelihood ratio needs more quantization attenuation, and the first quantization factor is obtained by using the second difference calculation. The first quantization factor is And determine the second quantization factor to be 1.

[0092] When the signal-to-noise ratio difference is within the third SNR threshold range, if the first signal-to-noise ratio is greater than the second signal-to-noise ratio, it is determined that the historical log-likelihood ratio requires more quantitative attenuation, and the second quantization factor is obtained by using the third difference calculation, and the second quantization factor is And determine the first quantization factor to be 1. If the first signal-to-noise ratio is less than the second signal-to-noise ratio, it is determined that the current log-likelihood ratio needs more quantization attenuation, and the first quantization factor is obtained by using the third difference calculation. The first quantization factor is And determine the second quantization factor to be 1.

[0093] When the signal-to-noise ratio difference is within the fourth SNR threshold range, if the first signal-to-noise ratio is greater than the second signal-to-noise ratio, it is determined that the historical log-likelihood ratio needs more quantization attenuation, and the second quantization factor is obtained by using the fourth difference calculation, and the second quantization factor is And determine the first quantization factor to be 1. If the first signal-to-noise ratio is less than the second signal-to-noise ratio, it is determined that the current log-likelihood ratio needs more quantization attenuation, and the first quantization factor is obtained by using the fourth difference calculation. The first quantization factor is And determine the second quantization factor to be 1.

[0094] In addition, when the signal-to-noise ratio difference is 0, that is, when the first signal-to-noise ratio is equal to the second signal-to-noise ratio, it means that there is not much reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio, and therefore, the first quantization factor and the second quantization factor are both 1. In addition, if the minimum value of the first SNR threshold range is greater than 0, then when the signal-to-noise ratio difference is less than the minimum value of the first SNR threshold range and greater than 0, it can also be considered that there is not much reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio, and it is also determined that the first quantization factor and the second quantization factor are both 1.

[0095] In addition, after obtaining the first quantization factor and the second quantization factor, the product of the first quantization factor and the current log-likelihood ratio is used as the current log-likelihood ratio after quantization, and the product of the second quantization factor and the historical log-likelihood ratio is used as the historical log-likelihood ratio after quantization. The current log-likelihood ratio after quantization and the historical log-likelihood ratio after quantization are input into LDEC for combined decoding processing to obtain a decoding result.

[0096] The decoding method provided by the embodiment of the present invention determines which of the two log-likelihood ratios has better reliability by comparing the modulation order, frequency deviation, and signal-to-noise ratio with those of the historical log-likelihood ratios when the current log-likelihood ratio is low-order and low-frequency offset, and determines quantization factors of different sizes by subdividing the reliability difference. The reliability difference is reduced by quantizing the attenuation of the two quantization factors on the respective log-likelihood ratios, thereby increasing the improvement effect of the log-likelihood ratio with high reliability on the decoding performance and reducing the adverse effect of the log-likelihood ratio with low reliability on the decoding performance, thereby ultimately improving the reliability of the communication system.

[0097] It should be noted that the above-mentioned preset order can be a higher order among the modulation orders 2, 4, 6, and 8, such as 6 or 8, and can be specifically determined based on actual measurement results, etc. Preferably, the preset order is the highest order 8.

[0098] In some embodiments of the present invention, obtaining the corresponding historical log-likelihood ratio includes:

[0099] Determine whether the signal to be decoded is a retransmitted signal based on the upload indication information of the signal to be decoded. That is, it is necessary to first determine whether the decoded signal is a retransmitted signal or a newly transmitted signal. If it is a newly transmitted signal, there is no corresponding historical log-likelihood ratio that can be used for merging, and there will be no impact on the decoding performance caused by the reliability difference during merging. The current log-likelihood ratio can be directly decoded to obtain the decoding result. If it is a retransmitted signal, the relevant historical log-likelihood ratio and related information can be obtained from the cache area, so as to be used for quantization and merging. For the specific quantization process, see the above description.

[0100] When the signal to be decoded is a retransmitted signal, the current HARQ state value of the signal to be decoded is used to perform matching among multiple cached historical HARQ state values.

[0101] If the matching result is characterized as a historical HARQ state value obtained by matching, then whether the matched historical HARQ state value is reliable is determined based on the time interval between the current HARQ state value and the matched historical HARQ state value. If no corresponding historical HARQ state value is matched, the current log-likelihood ratio is also processed like the above-mentioned new transmission signal, and is directly input into LDEC for decoding processing to obtain a decoding result.

[0102] If the matched historical HARQ state value is reliable, the historical log-likelihood ratio corresponding to the matched historical HARQ state value is used as the corresponding historical log-likelihood ratio. If the matched historical HARQ state value is unreliable, the current log-likelihood ratio is also processed like the above-mentioned new transmission signal, and is directly input into LDEC for decoding to obtain a decoding result.

[0103] In other embodiments of the present invention, the log-likelihood ratio of the current user at the last moment may be directly used as the historical log-likelihood ratio, or a historical log-likelihood ratio with higher reliability may be obtained in the cache area according to the actual test performance. For example, if all 8 bits of the historical log-likelihood ratio are 0, it is determined to be unreliable; if there are more 1 values ​​in the 8 bits, it is determined to be highly reliable.

[0104] The decoding method provided in the embodiment of the present invention determines a more reliable historical log-likelihood ratio for the current log-likelihood ratio by judging whether the signal to be decoded is a retransmission, matching the HARQ state value, and judging the time interval of the HARQ state value, thereby improving the quantization of the log-likelihood ratio to be more reliable and accurate, and further improving the reliability of the communication system.

[0105] In some embodiments of the present invention, whether the matched historical HARQ state value is reliable can be determined in the following manner: calculate the time interval between the current HARQ state value and the matched historical HARQ state value, and determine whether the time interval is less than or equal to the interval threshold; if so, the matched historical HARQ state value is determined to be reliable; otherwise, the matched historical HARQ state value is determined to be unreliable.

[0106] In another embodiment of the present invention, the reliability difference is divided into a first difference, a second difference, a third difference, and a fourth difference, the number of bits required to be quantized corresponding to the reliability difference is 1, 2, 3, and 4, the first SNR threshold range, the second SNR threshold range, the third SNR threshold range, and the fourth SNR threshold range are: (3, 6], (6, 9], (9, 12], (12, +∞), and the preset order is the highest order 8. For example, the decoding method is described. Specifically, Figure 2 As shown, the decoding method comprises the following steps:

[0107] S21, designing a HARQ state management structure Buffer, which is used to cache historical log-likelihood ratios and related information, and then entering S22.

[0108] Among them, the parameters of the HARQ buffer may include the maximum number of activated users of the uplink channel MAX_ACT_USERS, the maximum number of HARQ processes MAX_UL_HARQ_PROC, the value range of the system frame number and subframe number SfnSf of the signal to be decoded, the value range of the user index number UEID of the signal to be decoded, the value range of the HARQ process number HarqProcId, the value range of the modulation order ModOrder, the unit of frequency deviation Hz, the unit of signal-to-noise ratio dB, etc.

[0109] The value range of MAX_ACT_USERS, UEID, and HarqProcId can be set based on experience. MAX_UL_HARQ_PROC is determined according to the communication protocol. If it is specified by the 3GPP protocol, MAX_UL_HARQ_PROC is 16. SfnSf is specifically from the low bit, the 1st to 4th bits are the subframe number, representing the range of 0 to 9, the 5th to 14th bits are the system frame number, representing the range of 0 to 1023, and 1 system frame contains 10 subframes. The value range of ModOrder is determined according to the modulation mode, which can be [2, 4, 6, 8], and the corresponding modulation modes are QPSK, 16QAM, 64QAM, and 256QA respectively.

[0110] S22, obtaining the scheduling information of the signal to be decoded, and then entering S22.

[0111] Specifically, the system frame number, subframe number, user index number, HARQ process number, modulation order, number of code blocks, new transmission or retransmission indication parameter and other information that is convenient for historical log-likelihood ratio are obtained. Among them, the data type of the new transmission or retransmission indication parameter is uint8, the value 1 represents new transmission, and the value 0 represents retransmission.

[0112] S23, judging whether the new transmission or retransmission indication parameter indicates retransmission or new transmission, and proceeding to step S24 in case of new transmission, and proceeding to step S25 in case of retransmission.

[0113] S24, input the current log-likelihood ratio LLR-c of the newly transmitted signal to be decoded into LDEC for decoding processing, and then enter the end state.

[0114] Specifically, the absolute subframe number AbsSfIdx of the current signal to be decoded can be obtained by converting the system frame number and the subframe number Sfnsf, and a new HARQ buffer is allocated to the new transmission user according to the user index number UEID and the HARQ process number HarqProcId.

[0115] It should be noted that the current log-likelihood ratio LLR-c is directly input to LDEC for decoding processing, which is equivalent to the first quantization factor being 1. After being processed by the first quantization factor, the current log-likelihood ratio LLR-c remains unchanged.

[0116] In addition, after decoding is completed, the current log-likelihood ratio LLR-c is stored as a new historical log-likelihood ratio in the HARQ buffer. At the same time, the information related to this new historical log-likelihood ratio is updated and stored in the HARQ state management structure Buffer, including the system frame number and subframe number Sfnsf, the user index number UEID and the HARQ process number HarqProcId, the modulation order ModOrder, the frequency offset FO, the signal-to-noise ratio SNR, etc.

[0117] S25, determine the corresponding historical log-likelihood ratio LLR-h for the current log-likelihood ratio LLR-c according to the scheduling information of the signal to be decoded, and determine the first quantization factor LLRFactor-c and the second quantization factor LLRFactor-h by comparing the reliability difference between LLR-c and LLR-h, quantize LLR-c by LLRFactor-c, quantize LLR-h by LLRFactor-h, input the two quantized LLRs into LDEC for merging and decoding, obtain the decoding result, and then enter the end state. Among them, quantizing LLR-c by LLRFactor-c refers to LLRFactor-c×LLR-c, and quantizing LLR-h by LLRFactor-h refers to LLRFactor-h×LLR-h.

[0118] Specifically, if Figure 3 As shown, S25 includes the following sub-steps:

[0119] S251, determine whether the current UEID can be compared with all the historical UEIDs stored in the HARQ state management structure buffer. If not, it means that there is no historical HARQ state value that matches the current UEID, and the current decoded signal is processed as a new transmission signal, and enter S24. If there is a historical UEID that matches, it means that there is a historical HARQ state value that matches the current UEID, and enter S252.

[0120] S252, obtaining historical information related to the matched historical UEID, and then proceeding to S253.

[0121] The relevant historical information specifically includes historical SfnSf, historical HarqProcId, historical ModOrder, historical FO, historical SNR, LLR-h, etc. The historical SfnSf and the current SfnSf are converted into time units to convert into the current absolute subframe number AbsSfIdx and the historical absolute subframe number AbsSfIdx in seconds. The time interval SfDiff is obtained by calculating the difference between the current absolute subframe number AbsSfIdx and the historical absolute subframe number AbsSfIdx.

[0122] S253, determine whether the time interval SfDiff is greater than the interval threshold SfDiffThre. If SfDiff>SfDiffThre, it means that the time interval between the matched LLR-h and LLR-c is too large, the historical HARQ state value is judged to be invalid, and the signal to be decoded is used as a new transmission signal to enter S24. If SfDiff≤SfDiffThre, it means that the time interval between the matched LLR-h and LLR-c is normal, the historical HARQ state value is judged to be valid, and enter S254. Among them, the interval threshold SfDiffThre is set according to experience, and is set to 2000 in this embodiment, and can also be set to 1500, 2500, etc. in other embodiments.

[0123] S254, after multiple judgments are made based on the modulation order, FO value, and SNR value of LLR-h and LLR-c, the first quantization factor LLRFactor-c and the second quantization factor LLRFactor-h are determined, and then the process goes to S255.

[0124] In the communication performance simulation experiment of the high-speed scenario, it is found that when the 256QAM high-order modulation mode is used and the frequency deviation is greater than or equal to 800Hz, the decoding performance is significantly worse than other scenarios. The highest order is the modulation order corresponding to the 256QAM modulation mode, that is, 8, and the frequency deviation threshold is set to 800Hz. Based on this, S254 includes the following sub-steps:

[0125] S25401, determine whether the first modulation order is equal to 8 and whether the first frequency deviation is greater than or equal to 800Hz. If the first modulation order is 8 and the first frequency deviation is greater than or equal to 800Hz, enter S25402; if the first modulation order is not 8 and the first frequency deviation is less than 800Hz, enter S25409.

[0126] S25402, determine whether the first modulation order is greater than the second modulation order. When the first modulation order is greater than the second modulation order, enter S25403; when the first modulation order is equal to the second modulation order (that is, the second modulation order is also the highest order), enter S25404.

[0127] S25403, determine that the reliability difference between LLR-h and LLR-c is the third difference, the number of bits that need to be quantized and attenuated is 3, determine that LLRFactor-h=1, Then enter S255.

[0128] S25404, determine whether the second frequency deviation is greater than or equal to 800 Hz. If the second frequency deviation is greater than or equal to 800 Hz, proceed to S25405; if the second frequency deviation is less than 800 Hz, proceed to S25408.

[0129] S25405, determine whether the first frequency offset is greater than the second frequency offset. If the first frequency offset is greater than the second frequency offset, proceed to S25406; if the first frequency offset is less than the second frequency offset, proceed to S25407.

[0130] S25406, determining that the reliability difference between LLR-h and LLR-c is the first difference, the number of bits that need to be quantized and attenuated is 1, and determining that LLRFactor-h=1, Then enter S255.

[0131] S25407, determine that the reliability difference between LLR-h and LLR-c is the first difference, the number of bits to be quantized and attenuated is 1, and determine LLRFactor-c=1, then enter S255.

[0132] S25408, determining that the reliability difference between LLR-h and LLR-c is the second difference, the number of bits that need to be quantized and attenuated is 2, and determining that LLRFactor-h=1, Then enter S255.

[0133] S25409, determine whether the second modulation order is the highest order 8 and whether the second frequency deviation is greater than or equal to 800Hz. If the second modulation order is 8 and the second frequency deviation is greater than or equal to 800Hz, enter S25410; if the second modulation order is not 8 and the second frequency deviation is less than 800Hz, enter S25411.

[0134] S25410, determine that the reliability difference between LLR-h and LLR-c is the third difference, the number of bits required for quantization attenuation is 3, and determine LLRFactor-c=1, then enter S255.

[0135] S25411, determine whether the first signal-to-noise ratio is greater than the second signal-to-noise ratio. If the first signal-to-noise ratio is greater than the second signal-to-noise ratio, proceed to S25412; if the first signal-to-noise ratio is less than the second signal-to-noise ratio, proceed to S25419.

[0136] S25412, determine whether the difference between the first signal-to-noise ratio and the second signal-to-noise ratio is within the first SNR threshold range. If so, proceed to S25413; if not, proceed to S25414.

[0137] S25413, determine that the reliability difference between LLR-h and LLR-c is the first difference, the number of bits to be quantized and attenuated is 1, and determine LLRFactor-c=1, then enter S255.

[0138] S25414, determine whether the difference between the first signal-to-noise ratio and the second signal-to-noise ratio is within the second SNR threshold range. If so, proceed to S25415; if not, proceed to S25416.

[0139] S25415, determine that the reliability difference between LLR-h and LLR-c is the second difference, the number of bits required for quantization attenuation is 2, and determine LLRFactor-c=1, then enter S255.

[0140] S25416, determine whether the difference between the first signal-to-noise ratio and the second signal-to-noise ratio is within a third SNR threshold range. If so, proceed to S25417; if not, proceed to S25418.

[0141] S25417, determine that the reliability difference between LLR-h and LLR-c is the third difference, the number of bits required for quantization attenuation is 3, and determine LLRFactor-c=1, then enter S255.

[0142] S25418, determine that the reliability difference between LLR-h and LLR-c is the fourth difference, the number of bits required for quantization attenuation is 4, and determine LLRFactor-c=1, then enter S255.

[0143] S25419, determine whether the difference between the second signal-to-noise ratio and the first signal-to-noise ratio is within the first SNR threshold range. If so, proceed to S25420; if not, proceed to S25421.

[0144] S25420, determining that the reliability difference between LLR-h and LLR-c is the first difference, the number of bits that need to be quantized and attenuated is 1, and determining that LLRFactor-h=1, Then enter S255.

[0145] S25421, determine whether the difference between the second signal-to-noise ratio and the first signal-to-noise ratio is within the second SNR threshold range. If so, proceed to S25422; if not, proceed to S25423.

[0146] S25422, determine that the reliability difference between LLR-h and LLR-c is the second difference, the number of bits that need to be quantized and attenuated is 2, and determine that LLRFactor-h=1, Then enter S255.

[0147] S25423, determine whether the difference between the second signal-to-noise ratio and the first signal-to-noise ratio is within a third SNR threshold range. If so, proceed to S25424; if not, proceed to S25425.

[0148] S25424, determine that the reliability difference between LLR-h and LLR-c is the third difference, the number of bits that need to be quantized and attenuated is 3, determine that LLRFactor-h=1, Then enter S255.

[0149] S25425, determining that the reliability difference between LLR-h and LLR-c is the fourth difference, the number of bits that need to be quantized and attenuated is 4, and determining that LLRFactor-h=1, Then enter S255.

[0150] S255, quantize LLR-h and LLR-c using the first quantization factor LLRFactor-c and the second quantization factor LLRFactor-h, and then enter S256:

[0151] After quantization, LLR-h = LLRFactor-h × LLR-h;

[0152] After quantization, LLR-c=LLRFactor-c×LLR-c.

[0153] S256, input the quantized LLR-h and the quantized LLR-c into LDEC for merging and decoding, obtain a decoding result, and then enter S257.

[0154] S257, store the combined LLR as the new historical LLR in the HARQ buffer, and store the LLR-c related information corresponding to the new historical LLR, and then enter the end state. The related information includes SfnSf, UEID, HarqProcId, ModOrder, FO, SNR, etc.

[0155] Figure 4A schematic diagram of the structure of a decoding device provided by an embodiment of the present invention is shown in FIG. Figure 4 As shown, a decoding device, the device comprises:

[0156] The information acquisition unit 401 is used to acquire the current log-likelihood ratio and the corresponding historical log-likelihood ratio corresponding to the signal to be decoded.

[0157] The quantization factor determination unit 402 is used to compare the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio, and determine the corresponding first quantization factor and second quantization factor for the current log-likelihood ratio and the historical log-likelihood ratio according to the reliability difference.

[0158] The decoding unit 403 is used to quantize the current log-likelihood ratio using the first quantization factor, quantize the historical log-likelihood ratio using the second quantization factor, and combine and decode the two quantized log-likelihood ratios.

[0159] The decoding device provided in this embodiment corresponds to the decoding method provided above, and will not be described in detail here.

[0160] Figure 5 A schematic diagram of the structure of a decoding system provided by an embodiment of the present invention is shown in FIG. Figure 5 As shown, a decoding system includes an information management module, a channel estimation module CE, a frequency domain equalization module FDEQ, a demultiplexing demodulator and a demodulator DMAP, a quantization module, a cache module HARQ buffer, and a decoder LDEC.

[0161] Among them, the information management module is connected to the channel estimation module and the quantization module respectively. The information management module can provide the channel estimation module with various parameters, signals and other information related to the HARQ process related to the signal to be decoded, and it provides the quantization module with the modulation order corresponding to the current LLR.

[0162] The channel estimation module CE is connected to the frequency domain equalization module FDEQ and the quantization module respectively, and calculates the frequency offset FO of the signal to be decoded according to various parameters, signals and other information related to the HARQ process related to the signal to be decoded, and transmits the frequency offset FO to the quantization module.

[0163] The frequency domain equalization module FDEQ is connected to the demultiplexing demodulator DMAP and the quantization module respectively. It calculates the signal-to-noise ratio SNR of the signal to be decoded according to various parameters, signals and other information related to the HARQ process related to the signal to be decoded, and transmits the signal-to-noise ratio SNR to the quantization module.

[0164] The demultiplexer demodulator DMAP is connected to the quantization module, which converts the signal to be decoded into a log-likelihood ratio and transmits it to the quantization module as a current pair of LLRs.

[0165] The cache module HARQ buffer is connected to the quantization module, and stores a plurality of historical LLRs and related historical data, wherein the related historical data includes historical modulation orders, historical frequency deviation values, and historical signal-to-noise ratios.

[0166] The quantization module determines the first quantization factor of the current log-likelihood ratio and the second quantization factor of the historical log-likelihood ratio according to the modulation order, frequency deviation, signal-to-noise ratio, current LLR, corresponding historical LLR and related historical data, and quantizes the current log-likelihood ratio using the first quantization factor and quantizes the historical log-likelihood ratio using the second quantization factor, and obtains the quantized current LLR and the quantized historical LLR, and inputs the two quantized log-likelihood ratios into the decoder LDEC. The specific quantization factor determination process is described above and will not be repeated here.

[0167] The decoder LDEC combines the two quantized log-likelihood ratios, and then performs decoding processing on the combined LLR to obtain a decoding result, and stores the combined LLR as a new historical log-likelihood ratio in the cache module HARQbuffer.

[0168] In the decoding system provided by the embodiment of the present invention, the quantization module determines the first quantization factor of the current log-likelihood ratio and the second quantization factor of the historical log-likelihood ratio according to the modulation order, frequency deviation, signal-to-noise ratio, current log-likelihood ratio, corresponding historical log-likelihood ratio and related historical data, and quantizes the current log-likelihood ratio using the first quantization factor and quantizes the historical log-likelihood ratio using the second quantization factor. The decoder LDEC combines the two quantized log-likelihood ratios, and then performs decoding processing on the combined log-likelihood ratio to obtain a decoding result. This can avoid the adverse effects of log-likelihood ratios with lower reliability on decoding performance, and enhance the improvement effect of log-likelihood ratios with higher reliability on decoding performance.

[0169] Based on any of the above embodiments, another embodiment of the present invention further provides an electronic device, which may include: a processor, a communications interface, a memory, and a communication bus, wherein the processor, the communications interface, and the memory communicate with each other via the communication bus. The processor may call the logic instructions in the memory to execute the above decoding method.

[0170] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0171] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0172] On the other hand, an embodiment of the present invention further provides a storage medium on which a plurality of instructions are stored, and the instructions are suitable for being loaded by a processor to execute the decoding methods provided in the above embodiments.

[0173] In addition, the logic instructions in the above-mentioned storage medium can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0174] On the other hand, an embodiment of the present invention further provides a computer program product. When the program code contained in the computer program product is executed by a processor in an electronic device, the decoding method provided in the above embodiments is implemented.

[0175] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.

[0176] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment.

[0177] In summary, although the present invention has been disclosed as above in terms of preferred embodiments, the above preferred embodiments are not intended to limit the present invention. A person skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the scope defined in the claims.

Claims

1. A decoding method, characterized in that: include: Obtaining a current log-likelihood ratio and a corresponding historical log-likelihood ratio corresponding to the signal to be decoded; Comparing the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio, and determining a corresponding first quantization factor and a second quantization factor for the current log-likelihood ratio and the historical log-likelihood ratio according to the reliability difference; The current log-likelihood ratio is quantized using the first quantization factor, the historical log-likelihood ratio is quantized using the second quantization factor, and the two quantized log-likelihood ratios are combined and decoded.

2. The method according to claim 1, characterized in that The comparing the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio includes: The reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio is compared according to their respective modulation orders, frequency offsets and signal-to-noise ratios.

3. The method according to claim 2, characterized in that The current log-likelihood ratio corresponds to a first modulation order, a first frequency offset, and a first signal-to-noise ratio, and the historical log-likelihood ratio corresponds to a second modulation order, a second frequency offset, and a second signal-to-noise ratio; The comparing the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio includes: When the first modulation order is greater than or equal to a preset order and the first frequency offset is greater than or equal to a preset frequency offset threshold, compare the first modulation order with the second modulation order, the first frequency offset, the second frequency offset and the frequency offset threshold to determine the type of reliability difference; When the first modulation order is less than a preset order and the first frequency offset is less than the frequency offset threshold, comparing the first modulation order with the second modulation order, the first frequency offset, the second frequency offset and the frequency offset threshold, and the first signal-to-noise ratio with the second signal-to-noise ratio to determine the type of reliability difference; The determining, according to the reliability difference, a first quantization factor and a second quantization factor corresponding to the current log-likelihood ratio and the historical log-likelihood ratio comprises: A first quantization factor is determined for the current log-likelihood ratio and a second quantization factor is determined for the historical log-likelihood ratio according to the type of the reliability difference.

4. The method according to claim 3, characterized in that The step of comparing the first modulation order with the second modulation order, the first frequency offset, the second frequency offset, and the frequency offset threshold to determine the type of reliability difference when the first modulation order is greater than or equal to a preset order and the first frequency offset is greater than or equal to a preset frequency offset threshold comprises: When the second modulation order is greater than or equal to a preset order and the second frequency deviation is greater than or equal to the frequency deviation threshold, determining that the reliability difference is a first difference; When the second modulation order is greater than or equal to a preset order and the second frequency deviation is less than the frequency deviation threshold, determining the reliability difference to be a second difference; When the second modulation order is less than a preset order, determining the reliability difference to be a third difference; Among them, the first difference, the second difference and the third difference increase in sequence.

5. The method according to claim 3, characterized in that The step of comparing the first modulation order with the second modulation order, the first frequency offset, the second frequency offset and the frequency offset threshold, and the first signal-to-noise ratio with the second signal-to-noise ratio to determine the type of reliability difference when the first modulation order is less than a preset order and the first frequency offset is less than the frequency offset threshold, includes: When the second modulation order is greater than or equal to a preset order and the second frequency deviation is greater than or equal to the frequency deviation threshold, determining that the reliability difference is a third difference; When the second modulation order is less than a preset order and the second frequency offset is less than the frequency offset threshold, calculating a signal-to-noise ratio difference between the signal-to-noise ratio and the second signal-to-noise ratio; When the signal-to-noise ratio difference is within a first SNR threshold range, determining the reliability difference to be a first difference; When the signal-to-noise ratio difference is within a second SNR threshold range, determining the reliability difference to be a second difference; When the signal-to-noise ratio difference is within a third SNR threshold range, determining the reliability difference to be a third difference; When the signal-to-noise ratio difference is within a fourth SNR threshold range, determining the reliability difference to be a fourth difference; Among them, the first SNR threshold range, the second SNR threshold range, the third SNR threshold range and the fourth SNR threshold range are adjacent to each other in sequence and their respective maximum critical values ​​increase in sequence, and the first difference, the second difference, the third difference and the fourth difference increase in sequence.

6. The method according to claim 5, characterized in that The first difference, the second difference, the third difference or the fourth difference is characterized by different numbers of difference bits, the first quantization factor is characterized by the number of low-order bits that need to be discarded for the current log-likelihood ratio, and the second quantization factor is characterized by the number of low-order bits that need to be discarded for the historical log-likelihood ratio.

7. The method according to claim 4, characterized in that The step of determining a first quantization factor based on the current log-likelihood ratio and a second quantization factor based on the historical log-likelihood ratio according to the type of the reliability difference includes: When the second modulation order is greater than or equal to the preset order and the second frequency offset is greater than or equal to the frequency offset threshold, if the first frequency offset is greater than the second frequency offset, the first difference is used to calculate the first quantization factor and the second quantization factor is determined to be 1; if the first frequency offset is less than the second frequency offset, the second difference is used to calculate the second quantization factor and the first quantization factor is determined to be 1; When the second modulation order is greater than or equal to a preset order and the second frequency offset is less than the frequency offset threshold, obtaining a first quantization factor by using the second difference calculation and determining the second quantization factor to be 1; When the second modulation order is less than a preset order, the first quantization factor is obtained by using the third difference calculation and the second quantization factor is determined to be 1.

8. The method according to claim 7, characterized in that When the second modulation order is greater than or equal to the preset order and the second frequency offset is greater than or equal to the frequency offset threshold, if the first frequency offset is equal to the second frequency offset, it is determined that the first quantization factor and the second quantization factor are both 1.

9. The method according to claim 5, characterized in that The step of determining a first quantization factor based on the current log-likelihood ratio and a second quantization factor based on the historical log-likelihood ratio according to the type of the reliability difference includes: When the second modulation order is greater than or equal to a preset order and the second frequency offset is greater than or equal to the frequency offset threshold, obtaining a second quantization factor by using the third difference calculation and determining the first quantization factor to be 1; When the signal-to-noise ratio difference is within a first SNR threshold range, if the first signal-to-noise ratio is greater than the second signal-to-noise ratio, the first difference is used to calculate the second quantization factor and the first quantization factor is determined to be 1; if the first signal-to-noise ratio is less than the second signal-to-noise ratio, the first difference is used to calculate the first quantization factor and the second quantization factor is determined to be 1; When the signal-to-noise ratio difference is within a second SNR threshold range, if the first signal-to-noise ratio is greater than the second signal-to-noise ratio, the second difference is used to calculate the second quantization factor and the first quantization factor is determined to be 1; if the first signal-to-noise ratio is less than the second signal-to-noise ratio, the second difference is used to calculate the first quantization factor and the second quantization factor is determined to be 1; When the signal-to-noise ratio difference is within a third SNR threshold range, if the first signal-to-noise ratio is greater than the second signal-to-noise ratio, the second quantization factor is obtained by using the third difference calculation and the first quantization factor is determined to be 1; if the first signal-to-noise ratio is less than the second signal-to-noise ratio, the first quantization factor is obtained by using the third difference calculation and the second quantization factor is determined to be 1; When the signal-to-noise ratio difference is within a fourth SNR threshold range, if the first signal-to-noise ratio is greater than the second signal-to-noise ratio, the fourth difference is used to calculate the second quantization factor and the first quantization factor is determined to be 1; if the first signal-to-noise ratio is less than the second signal-to-noise ratio, the fourth difference is used to calculate the first quantization factor and the second quantization factor is determined to be 1.

10. The method according to claim 9, characterized in that If the first signal-to-noise ratio is equal to the second signal-to-noise ratio, it is determined that the first quantization factor and the second quantization factor are both 1.

11. The method according to any one of claims 3 to 10, characterized in that: The preset order is the highest modulation order.

12. The method according to claim 1, characterized in that The obtaining of the corresponding historical log-likelihood ratio comprises: Determining whether the signal to be decoded is a retransmission signal according to the upload indication information of the signal to be decoded; When the signal to be decoded is a retransmitted signal, matching the current HARQ state value of the signal to be decoded among multiple cached historical HARQ state values; If the matching result is characterized as a historical HARQ state value obtained by matching, determining whether the matched historical HARQ state value is reliable according to a time interval between the current HARQ state value and the matched historical HARQ state value; In the case that the matched historical HARQ state value is reliable, the historical log-likelihood ratio corresponding to the matched historical HARQ state value is used as the corresponding historical log-likelihood ratio.

13. The method according to claim 12, characterized in that After determining whether the signal to be decoded is a retransmission signal according to the upload indication information of the signal to be decoded, the method further includes: The new transmission step is to decode the current log-likelihood ratio.

14. The method according to claim 13, characterized in that After matching the current HARQ state value of the signal to be decoded among the cached multiple historical HARQ state values, the method further includes: If the matching result indicates that the historical HARQ state value is not matched, the new transmission step is performed; After determining whether the matched historical HARQ state value is reliable according to the time interval between the current HARQ state value and the matched historical HARQ state value, the method further includes: If the matched historical HARQ state value is unreliable, the new transmission step is performed.

15. A decoding device, characterized in that: include: An information acquisition unit, used to acquire a current log-likelihood ratio and a corresponding historical log-likelihood ratio corresponding to a signal to be decoded; a quantization factor determination unit, configured to compare the reliability difference between the current log-likelihood ratio and the historical log-likelihood ratio, and determine a first quantization factor and a second quantization factor corresponding to the current log-likelihood ratio and the historical log-likelihood ratio according to the reliability difference; A decoding unit is used to quantize the current log-likelihood ratio using the first quantization factor, quantize the historical log-likelihood ratio using the second quantization factor, and combine and decode the two quantized log-likelihood ratios.

16. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the method according to any one of claims 1 to 14 are implemented.

17. A storage medium, characterized in that: The storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the steps of the method according to any one of claims 1 to 14.

Images